Anyone have experience adding custom boorus? I'm trying to add ddlc.booru.org. I noticed that furry.booru.org and rule43.xxx (listed as rule34@booru.org) have the same structure, so I basically copied their formatting into the "manage url classes" dialog. I then added it to the "manage boorus" dialog, again basically copying furry booru verbatim except swapping to ddlc.booru.org, of course. ddlc is on my boorus list, and it apparently connects/reads because the tag search lists back results, but unfortunately the downloader seems to be stuck trying to find a download link. Could not parse a download link for http://ddlc.booru.org/index.php?id=2590&page=post&s=view!… (Copy note to see full error)
Traceback (most recent call last):
File "include\ClientImporting.py", line 628, in _WorkOnFiles
gallery.GetFile( temp_path, url )
File "include\ClientDownloading.py", line 690, in GetFile
( file_url, tags ) = self._GetFileURLAndTags( url )
File "include\ClientDownloading.py", line 685, in _GetFileURLAndTags
return self._ParseImagePage( html, url )
File "include\ClientDownloading.py", line 620, in _ParseImagePage
raise HydrusExceptions.DataMissing( 'Could not parse a download link for ' + url_base + '!' )
DataMissing: Could not parse a download link for http://ddlc.booru.org/index.php?id=2590&page=post&s=view! Is there some other change I should be making or is this something that could only be remedied internally?

Does anyone use Hydrus in Manjaro? I was using Ubuntu some days ago and now that I changed to Manjaro the function of downloading an image by just grabbing it from the browser and dropping it into the Hydrus GUI doesn't seem to work. I'm a techlet so I haven't been able to put it to work again.

Been having a minor inconvenience since my last update, no idea what caused it but i did a jump from v304 to v312 Basically every time i open hydrus and loads the last session, it always focuses to a page in the middle instead of the last one opened to the far right like it did before As i said it's minor, but every time i open it to do a quick search i'd rather have it focus on the last one opened (that probably was another quickly opened page) instead of one in the middle that i keep as static As it is every time i open it i have to move away from that middle page instead of just being able to quickly search tags I even tried to save and reload session or start them from scratch, but it always pick up a middle page as the remembered position

>>9338 >i did a jump from v304 to v312 i found your problem

Is there a way to add a parent to a whole namespace:*anything*? It's easy enough to select them all and add a parent, but if it was possible to do it automatically i wouldn't have to remember to add the parent every time i added a new tag with that namespace

>>9338 Nvm i got it, i understand that now the focus is always on the first page opened in a session and not the last one, and that middle page was actually the first one that i moved around at some point. so it was only a matter of rearranging the pages

Is there a booru for regular images? Most of the boorus I've seen are for porn, but if I want to find, for example, a bunch of :thinking: variants where would I go?

What happens when I interrupt the initial check from my subscriptions? Does it continue the check later or am I fucked and will never get the rest of it? Does it remember the page it stopped and doesn't shit its bed when a booru added 20 more files in the mean time? In the same vein, does the "on first check, get at most this many files" value prevent me from getting everything? The questions seem kind of retarded but the wording of the options don't exactly make it clear what i can expect from it.

What exactly does "dead" mean in the subscriptions? I have several subs that display dead even though the site shows me the tag just fine. How can make it undead?

>>9347 I'm pretty sure that just means that there hasn't been a new image posted with that tag in a while. It's not necessarily something to fix, but you could try changing the subscription type to something slower.

how do I get past the "bandwidth free in .." if I changed the rules after hitting the limit?

>>9327 (OP) Can I remove all URLs from selected files? Alternatively, can I remove all URLs from a certain domain? I don't want my source URL to have an imageboard since the file will go dead after a few days to weeks.

You can sort the file import status by source time. Is there a way to sort a search page by source time rather than import time?

>>9327 (OP) Can I do something to automatically give page numbers to different pools based on time imported? I have tons of pools that are in reverse order, and I'd like to give page numbers to them, but there are way too many to do manually. Alternatively, is there a way to add page numbers manually that doesn't involve F3>page:#>enter?

>>9363 Just came here to ask the exact same question. You going through old imports sorting and hoping to catch series of images too?

>>9376 Funny you say that… while I am not doing so atm, I thought to myself that I might want to do that in the future. Sorting by source time seems to be one of the best solutions to locate series of images that aren't tagged as such/pooled/etc.. since they're likely to have source times close to each other even if they aren't imported at the same time. Anyways, glad I am not the only one that thought that it might come in rather handy. If there isn't a way to achieve this, I think it could be fairly easy to add in… unless that information appears only temporarily during importing and isn't saved anywhere. Should that be the case: OOF

There should be a way to apply URLs to multiple images at once. I guess it seems unusual, but sometimes a single URL will list multiple images and I would want to link to that URL in some cases. Applying notes to multiple images at once would also be useful.

Are the intentions to support webp? My entire collection is in webp nowadays, which makes the current hydrus unusable for me.

>>9381 see >>9325 I myself await the day FLIF takes over never ever

>>9378 The information is preserved and I've used it before to tell what things are for, but I've somehow never needed this until now. >>9383 Seconding FLIF support, nothanks Google

>>9383 >>9385 >FLIF good lord, where is the psa for this? I've never heard about it but it looks like a god send

If I download the PTR on a hydrus, can I use it on a different DB or is it tied to the DB you download it on? >>9358 Click the gear "overwrite each 5 seconds"

I have tons of pictures that need tags however i'm too lazy for doing it manually. What is the best way to add gelbooru or rule 34 hentai/palheal so the tags are grabbed from there and added automatically to each file? sorry new too this and i was not able to make the the sync tag repository on the help site to work properly

>>9413 As far as I know, you can only reverse search one image at a time for now. The best way to do a lot of images is probably copying all the hashes, adding md5: to the front of each in notepad or something, and then opening a download page for a booru that supports md5 searching and pasting all of them (make sure to enable getting tags for images already in db). If the image is on the booru, then it should get all the tags for you.

What do you guys use when you have a set of images but you don't know what the actual title is? I don't want to just do 'series:that one where this thing happens' but I do want to be able to group them

Can I tell Hydrus to download a single Tumblr post? The simple downloader doesn't have an option, and the tumblr downloader only works on whole accounts.

Is any way to sort your gallery by the pics that does not have a specific tag?

>>9417 Try to find out what it's called. Got an example? >>9418 Not that I'm aware. I vaguely recall being able to use "/page/XX" to get in the ballpark of posts I was looking for but it wasn't too reliable. >>9419 add "-tag" to your search? Same as any booru

>>9417 I tag them all:series:[some keyboard mashing here, just to serve a a unique ID for the set]
meta:need qaAlso I apply tags that I might associate with it and search for in the meantime, unnamespaced. Then as I'm going through meta:* when I have time, if I come across one like this I just search up the group in a new tab and figure out the group or ask someone online for help figuring it out. My other common meta: tags are either content notes or meta:need crop, meta:need edit, meta:need qa, etc. For the latter, I add a qa:purified tag to everything I don't delete out of every given meta:need qa batch, and remove the meta:need qa tag. Not strictly needed but less confusing later than if I have some meta:need qa and some just suspiciously blank with common tags, this way I know I didn't just forget about the meta tag.

Is there a way to tag a bunch of files at once with known tags like from the hydrus tag repository? Have like 1500 images now but none of them have any tags. And are there any other repositories?

What is the correct way to retry a gallery download after changing the tag options, to tag already downloaded files? I checked the "fetch tags even…" in the tag options, but all the files were being skipped until I right-clicked the status log button and "delete successful file import items from queue". Is there a better way to do it?

>>9381 WebP is a good idea. >>9388 FLIF is a pretty dead project, HEIF is better and alive.

>>9450 >HEIF I had heard of it, but I didn't know that it was better. I just did a test and it made smaller files than FLIF did, and was faster (I made sure it was in lossless mode). Is there any software out there that supports HEIF yet? I saw a GIMP plugin, but is there any sort of regular image viewer that can see them?

>>9451 I did some more tests and found that HEIF is hit-or-miss when it comes to beating FLIF at lossless. One 6.47MB PNG made a 6.31MB HEIF, but FLIF got it down to 4.15MB. It also doesn't seem to work well on indexed or grayscale image, size-wise.

I'm having some trouble with the Pixiv downloader. I was trying to get knpf's stuff (20778542), but Hydrus isn't finding any URLs from that.

>>9457 Nevermind guess I just needed to restart the client after adding the login

>>9451 Windows 10 itself in a few months. https://blogs.windows.com/windowsexperience/2018/03/16/announcing-windows-10-insider-preview-build-17623-for-skip-ahead/

I want to add a namespace of pixivid to my local tags so I can keep track of my pixiv downloads even when artists change their display name (like some often do to advertise upcoming events or releases). Since their numerical ID doesn't change (except for the very rare time when they create an entire new account), I was hoping to use this as a unique identifier. The way I handled it was simply duplicating the creator tag parser and changing the search tag from userName to userId. It works fine in the test parser, and it seems to save properly. However, as soon as I run a Pixiv downloader or subscription, the new tag disappears from the parser. Is there a way to make the parser retain the additional tag?

Is there any way for this program to import existing file metadata (particularly tags/keywords), whether from the file or from something like an XMP sidecar?

When searching by number of tags, is there any way to only apply that to a specific namespace? For example, searching for images that are tagged with multiple creators, or images that have exactly three character tags. If not, is there some other way to search for something like this?

>>9488 No, but the dev has said it is something he wants himself so expect it sooner rather than later.

Can I change the default duplicate filter settings (I want to copy ratings by default)? I can change it per duplicate filter session but I can't figure out how to make it apply everywhere permanently.

Is there a way to set vim shourtcuts to navigate in the main gui?

Does anyone have a good script for treating mega directories like subscriptions yet? Ideally that downloads from the base directory and applies its name and the names of subdirectories each file comes from as tags, and checks all the subdirectories periodically for more files?

What do you do when you need to replace some images because you got them in higher res? I feel like with hydrus it's a bit awkward and not just a matter of delete and replace because of tags and favorite\like status Not sure what's the best\fastest\less messy course of actions is, for example in a situation where you have to replace 50+ images Do i just upload all the new tagless ones, and then work my way between the two pages to copy tags and delete the old images? Do i instead upload them on another hydrus client first (in order to be able to have a side by side comparison instead of having to switch pages), copy the tags over, then export them from that client and upload them in the right one? Or do i export the older ones with tags, then delete them after i gave the .txt files to the newer images? The first method is the only one that allows me to copy the rating of images since i'm copying one by one, but it's also the only one that doesn't allow me to have a side by side comparison, which is kinda needed when i have to locate the duplicates in what could be a very large page The other two methods are better for the comparing view, but i'm just copying the tags without the rating, so chances are that when i upload the new ones i'll either forget to rate them correctly (if i already deleted the older ones) or have to resort in tabbing between two pages to rate them

>>9496 You absolutely MUST learn to use the duplicate system, Hydrus without it is a real pain and you can save yourself like 90% of the time either of those methods would take. It's weird to set up but you'll get used to it quick and you can just save your session/continue from last time and leave a duplicates tab open. Honestly we're probably at the point where I should just write a shitty PDF of some of the necessary basics of Hydrus use, upload it, and let others berate me and tell me about the stuff I'm doing the old wrong way and things I don't know about that I should. But for now, here you go: >Import all of your copies of the available images with all of the tags you happen to already have on that copy. >Under new page, pick new page > special > duplicates processing >hit the play button under the 1-preparation box on the side - if you have lots of files this could take a while, otherwise it's cake, but after that whenever you want to update your duplicates list just hit the rebalance play button once every month or two and it likely won't take long at all >under 2-discovery, leave it on exact match (0 distance) for now, this isn't strictly true as it's just based on a quick but limited look at the files, so you will still have to look out for alternates while you're filtering >under 3-processing, probably leave as my files, ignore all of the "quick and dirty filtering" stuff in the box below, and hit Launch the Filter >after a small delay a fullscreen window should pop up showing an image >scroll wheel or left/right keys to switch between images, be careful if using keys because up/down goes forward/back a pair >you can use right or left click or maybe keys or programmable buttons to make content decisions, but for now I vastly prefer mousing over the top center of the screen which pops up info about the image and also button choices (probably there's now a way to get it to always show, but I missed a lot of updates) Your choices at this point are: >not duplicates - if it's genuinely not anything like a duplicate, false positive, on exact match have had this maybe 3 times, all screenshots of different text from same websites >custom action - might be super useful but I'd stay away >same quality - no idea what this does, last I knew it just passed over the pair and kept both copies in the database uselessly, which doesn't help you given they're exact matches (visually and filesize wise, if they're true dupe dupes Hydrus is meant to filter them out in download/import…sometimes…I don't really know how it works) >alternates - I use this one mostly for different images in an image set, I don't know if it makes it easier to find sets without tagging them or not, but it does define a relationship while keeping them both >this is better - click while on the one you like better, best for last. This commutes all tags over to the "better" file before deleting the inferior file. No takebacks, but it doesn't actually commit changes until you hit your 50 or 250 or whatever file limit and it asks you if you want to save changes, or if you Esc out of the filter in which case it also asks if you want to save changes…there are buttons to skip pairs or skip back to pairs and redo your choice, I think mouse back/forward buttons also work for that If you realize both images are trash, just hit Del, left key, Enter to send them to trash. Do it twice to delete them entirely. However, if you have two terrible res versions of a pic you like, just pick the "better one" whether visually or based on the helpful hints in the top bar. Speaking of them, it's easier to see than to have explained to you, but they are very useful especially when dupe sorting on a small screen like a laptop…just remember, # of tags doesn't really matter since the tags will all merge to the better one. Back to keeping the better of two bad dupes, there's a chance you don't actually have a higher res version of the image like you think you do, so just keep the "best" and eventually you will be left with only the truly best (and all the tags) through dupe sorting. As of yet you can only compare two images at once, so something with five or six dupes will take a few goes (though it does a sort of tree elimination to save time). It also sometimes stops asking about a particular file before all dupes are sorted to go to some other seemingly randomly-picked dupes. I have written some ideas recently for a simple to use and hopefully simple to code optional multi-dupe-filter, maybe after the massive undertaking of the new download/login engines are done dev will have time to think about adding in some extras like that. Last warning: Once you realize how easy dupes are, you will start downloading more volume and more carelessly and even with the new code recognizing already used download locations you will still end up with more dupes than you ever actually sort.

>>9497 Using the duplicate filter is easy enough, but if it's not up to date it can't help if you want to immediately replace a bunch of images you have For example, let's say i downloaded a bunch of new images and i know that among these there are some of which i have lower res duplicates in hydrus that now i want to get rid of Now let's add that i have slacked in my dupe filtering process, this would mean that those old images i want to replace "right now" will probably have to wait before they are presented to me by the dupe filter Also, probably it's just me, but even if i was up to date with the filtering, i'd worry that the filter somehow didn't pick up on some of the new images being duplicates Basically i rely on the dupe filter for "accidental" duplicates, when i truly didn't know i was downloading similar images, i was asking for some suggestion in the cases when you are 100% sure you have duplicate at hand and want to quickly swap them with tags\ratings before waiting on the dupe filter

>>9498 Yes, they will have to wait, and we're all chronically behind on dupe filtering. Don't get me wrong I would love some more control over which dupes or to dupe filter specifically within a local context like an import page or search, but probably won't be available for a while. You're still just losing time by dupe sorting manually, so in your case I'd say yeah, if you know what your two groups are (say anything tagged "wallpaper" and also your newly downloaded batch of wallpapers), export the former with tags in filename, import them to another hydrus install, import the new ones, and filter there. Probably the ultimate solution from dev's pov is multi-db sessions where you can generate temporary separate dbs, perhaps even virtual dbs within your db using some internal tag to label them as part of it, which a new tab-set of Hydrus then acts upon. Might be quicker than trying to add more functionality to everything else. I won't ask why you're not immediately downloading everything to Hydrus (other than lack of downloader scripts I guess)

>>9498 Oh, I forgot! There's a free software called Visipics that I used to use before Hydrus which isn't perfect but does a lot of the same stuff…viewing panel, actually does let you view multiples at once, but you have to manually click for each one you want deleted in a list, then run the batch at the end. It MIGHT have a crude system where it auto-applies delete filter for dupes, I really can't remember. I think AntiDupe only works if file md5 is the same, so won't help.

>>9500 Basically in my most recent case, i came across some 50+ higher res image of an artist of which i had the lower pixiv res counterpart I wanted to replace them asap, but i was behind on dupe filtering so i had to do it manually Unfortunately they weren't part of a pool or anything, just random images so i had to manually locate each one of them in the whole artist search page, copy over the tags\rating and then delete the old ones But yeah, it was pretty slow and also a bit annoying because hydrus doesn't let you have 2 pages side by side for an easier time seeing and locating them like you could with windows folders (similar thing would be to use 2 hydrus clients, but then the rating doesn't carry over), the only positive thing is that i'm sure to have replaced them all correctly since i was doing it one by one I'll try to be up to date with the dupes so next time i can use it, but i'd probably still wonder if it managed to find every duplicate pair or if it missed something Being able to search for duplicates in the context of a single page would be amazing, hope we get that someday >>9501 I used visipics as well before hydrus, but not anymore since it has the advantage to not show you previous pairs anymore (unless i was misusing it and that too had a way to set alternates and stuff, but for now hydrus is ok)

>>9498 >>9502 There's a system tag for "num duplicate relationships" with a "unknown relationship" option to search for things that haven't been sorted yet. You could use that with the artist tag to easily find all duplicate images by that artist instead of searching manually. Alternatively, you can right click on each file and then click one of the "find similar files" options.

On the subject of dupe filtering, is it possible to make the better\worse decision carry over the rating as well? Mine only carries over the tags and sometimes i forget to check it while i'm filtering

>>9502 For now, the only way to make sure Hydrus doesn't miss any dupes is probably to set distance to 1 instead of 0, and sort through that massive pile of dupes. I would love to just right-click and search all dupes at x default distance, but I'm sure it will need significant new code, and depending on how the db is set up, might be a processing-expensive operation (though I'd think it could just do a lookup in the same tree the duplicate detection already makes) For Visipics, I'm talking about just using it for your specific higher/lower res pairs you want done, so you likely wouldn't need to run it more than once on the same group unless I'm missing something here.

I have some images that don't get picked up by the duplicate filter, but are clearly alternates, is there a way i can force the pair in the filter or set up their relationship outside of it?

Is there some functionality missing for linux? I can't import webms or import tags when downloading from a gallery. I'm using xubuntu LTS .

Is there a consistent guideline out there for when it's safe to force close the client? Every so often it freezes for possibly hours and I always want to force close it but I'm too scared to do so. Currently I'd just imported 280 images at once, but as soon as it opened the tab for it the tab was broken, as in the after-image of the previously-viewed tab is sort of glitched into view, and none of the contents of the import tab is visible. Well I let the number finish counting up to import it all, then restarted the client, but after showering I noticed the client had disappeared from my taskbar but was frozen at like 192k in task manager. It's been over four hours and it's now at 167,008K and still frozen. I don't want to lose my database by forcing this closed, and the manual only said like "it's usually safe to force close it after a very, very long time." That doesn't give me any context whatsoever. Has anyone ever made a list of hypotheticals for where the client would not fuck you by forcing it closed and if so please show it to me, thanks.

Decided to finally give this program a go to replace my massive clusterfuck of folders of over 50k images but there are some problems. Currently I sort images in windows explorer by time created (aka time saved), newest first. This has the convenient side effect of keeping related images grouped (for example, when I save a pixiv gallery they will all show up together in the folder, in the order they were in the gallery) but there doesn't seem to be a way to this with hydrus. When I import one of my folders it just seems to import the files with a completely random order, and the sorting options feel extremely limited. Is there any way to do this at all? Related: is there any way to preserve the image's original metadata like filename and time created? I want to know when I first saved my images.

>>9590 When you import, if you're on v315+ hydrus import order will be forced by name instead of whatever order you dragged them in (i'm staying on 314 because i prefer being able to import them by time created) Hydrus in general is kind of annoying to deal with regarding images that go next to each other, even if you import them in the correct order and then sort by time imported, they may get swapped around (i can only assume this is because they get the same time by the seconds and hydrus can't remember the exact order), not to mention if you sort by newest it would mean seeing them backwards Only "solution" would be to give those images a namespaced tag and sort by that, i personally don't like this solution as it creates other problems but it may work for you If you want to preserve the filename and maybe use it as an ordering tag, just open "apply tags based on filename" and check "add filename" As for preserving the time created, no, i'm not even sure it's possible as hydrus doesn't put the original image in the db, but creates a copy and deletes the original if you allow it to

>>9590 bulk rename utility can add the metada to your filenames, and you can either do some regex wizardry, add the tags to a text file (no idea how) or just rename them and re-import without deleting them.

Did Tumblr finally fuck itself over? I can't open any blogs without logging in (and when I do it's a dashboard blog) and Hydrus can't download anything even with the clickthrough thing.

Is there a way to get ratings (safe/questionable/explicit) as tags when downloading from gelbooru (and other boorus)?

>>9600 JetBoom's booru tag parser

Any way to set the external viewer as the default one? I want to use Hydrus just as a DB and use MPV and feh to view files. I just found out about hydrus and so far has been great. Great work dev

>>9602 Afaik no, you could just make an habit of opening the files with the shortcut (ctrl+e i think), but with the external viewer you'd just browse between the files in the db folder instead of the hydrus page, just to open the next image you'd have to tab back to open that one externally as well

>>9605 >Afaik no, you could just make an habit of opening the files with the shortcut (ctrl+e i think) Somehow I thought the right click option would tell you of this shortcut, but it didn't so I figured it didnt exist. But this helps me, thanks.

>>9605 Okay, thanks. If I'm taggin or just browsing files I will use the hydrus browser but specificaly for videos I wanted to open them externall because no sound in hydrus. The shortcut helps a lot, right clicking is hell

So I'm trying to set up an instance of hydrus-server. So I'm wondering if is there an easy way to convert/import stuff from a client db to a server db? Also stupid(er) question, but how do I actually add files to a instance of hydrus-server? I have an admin server, a file repo, and a tag repo set up. I can get tags into the tag repo, but I can't find a way to move/import a file into the file repo.

>add tumblr blog to downloader >66 images detected >every one fails >error message in log basically says "access denied" >the source URL that Hydrus tries doesn't work in the first place fuck

How do I make it scan my files and import their tags from boorus? Can I give it a priority list like check gelbooru first, if it's not found check danbooru, and so on?

>>9656 No need to scan or anything. If the hash matches the one on the booru, the tags will 99% be automatically added as long as someone has uploaded them to the PTR. Otherwise, you can get the md5 and do a md5: on gelbooru.

>>9619 Same here. Downgrading to 315 didn't help. Tumblr probably changed something internally. Thats something i would love to see fixed! Many thanks!

>>9553 Select both of them, right click > duplicates > set duplicate relationships You can also do more than two images this way and it'll set the relationship for every image in relation to every other selected image, so it's useful if you have a large set that are all variations of each other.

Is there any way to open the File Import Status window on a multiple watcher tab? It kind of sucks to see that a bunch of files were ignored without being able to see which files were ignored or check why they were ignored or even to do literally anything about it at all without just giving up on multiwatchers and just re-adding the thread on a normal watcher tab.

>Otherwise, you can get the md5 and do a md5: on gelbooru I have lots of images from a few boorus that aren't auto tagged by the PTR. Is there a built in way or plugin that does an md5 search of all my files for me? Doing it manually would take forever.

>>9672 Meant for >>9662

I use these addons to one-click download from pixiv: https://addons.mozilla.org/en-US/firefox/addon/ank-pixiv-tool/ https://addons.mozilla.org/en-US/firefox/addon/px-downloader/ and they (after configuring) produce filenames of: AnkPixiv/?site-name?/(?member-id?) ?memoized-name?/(?illust-id?)?illust-year?-?illust-month?-?illust-day?[?illust-hour??illust-minute?] ?title? ie: PxDownloader\椅子(20663)\(18182022) 2011-04-15[2044] ももこ Tbh the above formatting is a clusterfuck, just ignore it (the first addon stops working a lot, so I use the second one, which randomly fails to fetch images and never tells you so it sucks, but the first is still not working so I had to use+mention it). Can Hydrus match ripping this information? I made the pixiv illustration id the first thing in the filename, so basically if Hydrus can lift that and tag them as such I wouldn't really need the information following, though I guess it'd be nice.

>>9667 >>9619 For what it's worth I once found a random site that shows me NSFW-flagged tumblrs without logging in, and just now I searched it in my history to simply reload the same page that worked for me just a few months ago, but now it doesn't (it returns "Hrmm, couldn't find tumblr user 'Hecchidechu'. Showing twitter search."). You don't want to visit that tumblr btw since it's a western artist who loves armpit hair. Tumblr definitely broke something, probably on purpose. I wouldn't be surprised if tumblr-hosted images no longer showed up in reverse image searches. Considering they blocked NSFW tumblrs from being searched for or viewed without an account, they're fuckers and I don't have faith they'll lift this, so hoping it can be circumvented.

Does Hydrus not read deviantart tags or am I screwing something up

How do you download from 8ch threads? I'm on the latest version as of this moment. The thread watcher works fine for 4chan, but just returns 404 for 8ch (the thread still exists, I checked). I think in a previous version it worked for me, but it hasn't recently.

Is there a way for me to see a list of my local tags?

I've found this program very useful, and I've found myself archiving and taggin a lot of files that are not image files itself. Mostly PDFs, I would love to know if there is a way that a thumbnail preview of the first page form a pdf could be displayed in the thumbnail preview. I've found Hydrus Network far beyond in usefulness than others tag-based programs. It's nice to have a database for each need: images, pdf books, editable files, etc. Thanks :)

>>9682 pages > new download page > simple downloader paste thread link into 1 hit enter if it doesn't work, your shit is seriously fucked

Is there a way to make the url downloader automatically paste whatever url i just copied? Kinda like how hdoujin downloader does Otherwise it doesn't seem very fast if i have to paste it myself one by one everytime

Could hydrus import files in the reverse order? As of now, it works like this: let's say you download some artist's artwork, it will start to download from newest image to oldest one, which means if you mass download some images and use sort by the imported option, new uploads will appear first under oldest first choice and vice versa. And this means if you download some images later, unrelated to the mass upload, file organization by date will be completely fucked up.

>>9715 You could add order:id_asc to download them in reverse order, although i'm not sure on how many boorus it works on (i just use it on sankaku, no idea on how to do it for pixiv) The last resort would be to manually download and drag them into the correct time order (i do this with the booru tag parser)

Been a while since i used the gallery downloader for Sankaku, so now there's a limit of 50 images per day? That's a bit low to be useful, any way to work around it?

>>9716 Thanks, that works for Danbooru which is the only place I ever download from, I'd rather upload from Pixiv to Danbooru instead of directly downloading from there. >drag them into the correct time order Not sure if I understand that, you mean changing the time order manually?

I can use a Firefox addon called "Imagus" to expand images upon hover, and I noticed it tells me its timestamp. I don't even see how to get the timestamp otherwise. How may I tag instagram images with their date all the way down, automatically?

>>9718 I meant downloading them without hydrus' downloader, then sorting by time created asc\desc on explorer depending if you started on the newest or oldest images and then dragging them into hydrus Since i use the booru tag parser to download stuff i do it like this

Is there any way to select an image and say "find similar images" ?

I made a new pixiv account and tried to download a gallery using the pixiv artist id box, but the status will not switch away from "stopped". When I add an artist id, it says checking next page and then immediately switches to stopped status. I'm assuming the id to be the last bit of the url of an artist's main page. But since it behaves the exact same with a non-existing id, I'm assuming I'm doing it wrong. Anyone know what the issue could be?

>>9747 Turn on help->advanced mode.

How can I view my alternate pairs in the duplicates tab? Can it automatically tag alternates as siblings for me?

>>9748 wtf i love hydrus now

How do I import tags?

I'm pretty new to this. I need to be able to search derpibooru, how do I add it to the list of boorus? I've tried messing around with the URL classes, but I can't seem to get anywhere at all.

Can you batch rename tags? There is this xkcd comic (https://xkcd.com/1179/) that enlightened me of "ISO 8601". Basically it's a standard to format dates in the format of YYYY-MM-DD. To me this appeals to my biases, so I wish to format dates this way when necessary. However recently I discovered an artist who only posts his art to his blog (as far as I know), and his filenames aren't entirely uniform, so feeding it to Hydrus straight leads to a clusterfuck. But I did that anyway. I'm going back to tag images by their date, but the format the blog owner uses is YYYY/MM/DD. The order is in conformity with ISO 8601, but the forward slashes aren't. Is there a way I can just highlight multiple tags, and change the same character for all of them? This would help me in other ways, too, since I use an addon to one-click download from pixiv, which generates filenames with a lot of information. So maybe not batch renaming in that case, but to isolate the information and produce a tag from it to the specific image it's from, that would be nice. Also I guess I'll link the guy's blog: http://medakaken.blog87.fc2.com/ I found a page with dead links, so I archived it on the wayback machine, and it kinda freaked me out that someone archived the same page two years prior, with the same links dead.

Does hydrus still support importing tags from https://github.com/JetBoom/boorutagparser-server because it's not working for me anymore. I haven't used it for over a year. Is there a new setting I have to change to make hydrus read the txt files when importing?

>>9754 Yeah it works, plenty of people use it including me (no idea how i could without) Anyway, when importing just click on "add tags based on filename" and then on the right check "try to load tags from neighboring txt file" In case you forgot, remember that the .txt has to reflect both name and extension of the image, as in "name.ext.txt"

>>9754 >>9756 A better solution for the boorutagparser server is going to file > import and export folders > manage import folders. You can then click "add" in tag options and that option is there. This way, everything can be imported automatically as files are downloaded and you don't have to worry about settings again. The tagparser server describes how to do this, but not very well.

>>9756 >Yeah it works, plenty of people use it including me (no idea how i could without) I've never used it- partially from not knowing what it was- but my interests are very limited. Like I know everyone has experienced scrolling through pages on a porn site, only to eventually open one video due to the thumbnail, the scene of which turns out to only occur for like 5 seconds before being forfeit completely for the rest of the vid. But the information tags provide, like blush, bracelets, shorts, these do nothing for me since I don't sort that way or pursue these things. Some tags share parallels to what I like, but even then just because something is tagged thick thighs doesn't mean I could even like how it was done in the pic. So basically everything I don't care about is forfeit. I use the tags on boorus to find pics, but I don't get off to the tags I use to find them. Maybe I'm misunderstanding but at this point I barely tag anything anymore, and instead I tag a single pic by an artist, if that, and only tag the rest with the artist tag. In this scenario if I ripped booru tags, searching booru tags I don't care for will provide a coincidental selection of pics I've all cherrypicked. But I don't care for the tag so why would I care- it does nothing for me. I can only see this working if we were several generations into the future and humans had developed a way to extrapolate from your own mind to tag shit in exactly a way that appeals to your biases. But today I can get bitches automatically tagged with red hair if they have red hair. But I would never start a "red-haired bitches only night" in my lifetime. So I don't see the point in immortalizing that distinction. Even if I were showing someone else my Hydrus, I simply don't care for those things, so if they want to see how many coincidental red-haired bitches I've saved they're free to clone my Hydrus and rip those tags themselves, since it would serve myself no purpose. If I'm misunderstanding then sorry but redpill me on it please.

Hello, is it me or has the right click -> share -> copy -> hashes -> md5 function disappeared? If so, is there any other ways to copy the md5 hashes for one or a selection of pictures? Sorry, haven't used Hydrus in a pretty long time

>>9767 I feel like I encountered that once, but I don't remember the context, sorry. I think it wasn't working in the thumbnail viewer, but it worked in the image viewer. But both have that functionality fro me right now and I can't seem to recreate it disappearing. …….. actually, I say that, but I had "advanced mode" enabled under the "help" tab of my Hydrus. After disabling it the client only offers copying hash (not limited to just md5) from the media viewer. I dunno why it'd purposely limit this functionality to the image viewer unless advanced mode is on. I guess for new users it's to simplify what they look at when they first play around with it?

>>9767 >>9769 yeah you probably need to enable advanced mode because I have those options.

>>9769 >>9770 Yep, I'm dumb. I just needed to turn on advanced options to have it pop up again. Thank you !

>>9766 I get what you're saying, it's hard to see yourself specifically starting a search for tags that you don't care about, but i see it like this: First of all, since i mainly download with the parser, all the useless tags come with the important ones, it's true that i'd never do a bracelet only search, but as you said it may come in handy when looking for a specific pic, and also at that point it's more work to remove them than to just import them all Second, in my case there actually are tags for which i'd do a search, or a combo of them (for example i'm into contrapposto+arms up or on side + hand on hip), also funny that you mentioned it but i'd actually do a red haired bitches only night Basically i get everything because it's faster than to tag everything myself and among all the useless tags there are actually some i search for actual browsing When i actually have to tag stuff myself that has no booru to pull them from, i obviously don't bother with every damn detail, depending on my mood i might go as far as to tag the hair color or the style if i like it, but otherwise it's just creator\copyright\character

>>9772 To be honest I started writing a reply to this earlier this morning and it was mad autism. Like during it I was like I don't needa be doin this. So I tabbed out then just closed the tab after a while. But basically, porn to me is only one thing. By which I mean, the best porn is one thing, like, it never changes. Everything else is just less literally representative of what I think is the best, and I can entertain the thought of some of it. I was going to explain how I manage to get reimu's detached sleeves, but not bracelets, but it's awful so forget it. I just at best don't have any bias to these tags (nor the subject matter), and at worst I would never willingly create/immortalize it. If I could feel anything about them at all, it'd be pain, so to me it sucks. So they still exist, and I could still use them to find an image, but it's exclusively inadvertent if ever I happen to have memory of the tags I don't get off to. I'd rather sort things using my small scale of what I can get. Even where tags coincide with the limited amount of stuff I can get, I may not think they're done right, so I would never go to the pic to indulge in that subject matter. So I have trouble indiscriminately eating conventional tags when I 99% of the time won't care for the distinction of the tag, but even if I do I might wish the tag removed provided I don't think it's worth turning to the pic for that subject matter. This was hard to type and it still feels like I failed, even though I purposely didn't specify anything. Anyway I don't care for most tags. I cherrypick everything I put in my database, so I know what I'd turn to the pic for, and how well I believe it to be done, even if I'm too lazy to tag it appropriately immediately. The only things in my db I might not necessarily like at all are stuff from twitter, since there's so much garbage there, I just rip twitter accounts and feed it to my Hydrus. Even then I wouldn't use pleb tags, since they go over my head.

Is there any way to set up the Gallery Downloader to download queries one at a time in the order that they're added, the way that it used to work? Running related queries simultaneously destroys any hope of maintaining a sane import order, and pausing every single queue so that I can manually unpause one, watch for it to finish, and then unpause the next one, is not exactly an efficient work flow, especially if there are a large amount to go through.

Babby's first problem: on 3.19 released 3 days ago, I can't see the list of tag parents/tag siblings I've previously added. In fact, the window that opens doesn't look like it is shown in 'help>advanced>tag parents' and the like. The behavior does not change whether the 'show all pairs' is checked or not. Tested in both the extract-only version, and then in windows-install. Question: is it a bug, or is the function borked for now? **** MFW I found the source of the problem seconds before posting - wrapping doesn't work correctly. Unless the window that opens is over ~860px high, the 'child<>parent' pairs won't show and *there is no indication whatsoever that the option is there*. Even the text is partially cut off; by default the opened window is tiny, and if you use a laptop with low resolution the function cannot be used at all.

Is there something I'm missing to access a db on a NTFS drive from Hydrus running on Linux? I can't get it to work.

How can I reset my client options while keeping the database?

How do I raise the sync limit of a subscription? I'd like if it synched all the matching files in the hydrus, instead of checking 20 and waiting for the next check time.

I download something from a gallery and it says 66 items, but it only downloads 62, it says that the other 4 are already in the db. I manually check these files and they are very similar to previous images, but not identical (one has English text, the other has Japanese text). I believe the client is assuming these very similar images are identical. I manually download these extra 4 images by URL. They download successfully. I then do "find similar files->exact match". Sure enough, the similar but not identical image comes up. So it seems that some very similar images are being mistakenly called an exact match. So is there a way to override this, or alter the tolerances?

So is there something special going on with Sankaku? I've been given the "clicking too fast, slow down" message before while browsing manually before, but Hydrus displays the recent errors time limit in downloaders after around 20 files. I have my timer set to a healthy 30 seconds per file.

>>9820 I think I've found the relevant info for this issue, for those interested. >If you are interested, the current version of this system uses a phash (a 64-bit binary string 'perceptual hash' based on whether the values of an 8x8 DCT of a 32x32 greyscale version of the image are above or below the average value) to represent the image shape and a VPTree to search different files' phashes' relative hamming distance. I expect to extend it in future with multiple phash generation (flips, rotations, crops on interesting parts of the image) and most-common colour comparisons.

how do i add additional local tag archives? hydrus repository is all network, not local, and double clicking "local tag service" only looks to give a renaming option

>>9824 For those interested, I think I figured two things out. It looks like Sankaku really doesn't like it if you're using an adblocker. I don't know why something in Firefox would affect Hydrus, but disabling it makes everything smoother. It also doesn't like it when I have a Sankaku tab opened in Firefox. So far, closing these things has helped.

>>9824 >>9858 Well, never mind. Throttled me again right after I posted. Sankaku, you bastard.

In the new multi downloader, is the auto override bandwidth rules after 5 seconds in the cog menu only apply to the current highlight search, or all searches in the multi downloader list?

What does it mean when a tag suggestion has a range (89-206) ? I only have 89 Hatsuzuki images.

>>9871 Usually due to tag siblings, which I cannot count accurately and quickly for tag autocomplete results. I expect many/all of those 89 files also have the unnamespaced version of the tag, and maybe a '(kancolle)' vs '(kantai collection)' version thrown in there a few times as well. You can inspect for sure if you ctrl+a->f3 the results.

>>9824 >>9858 >>9859 Sank have had bandwidth problems for a long time, and they have a history of unusual bandwidth filters to try to keep it under control. I don't know anything for sure, but I wouldn't be surprised if they have put special per-IP rules on or something. An old rule used to be something like '5rqs per 8s', but I can't remember for sure, and they may have added new rules since then. It could also be that your 30s rule is firing wrongly on the hydrus side. If you like, try turning on help->debug->report modes->network report mode and then do some sank stuff. It'll spam your popups with init/start/done popups for each job. If you discover gallery hits are spamming way too fast for you or something, let me know!

>>9874 Awesome thanks, I'll try it after my 8.5 hr recent error/delay period is up. There might be some kind of script that can help. Does sankaku have some kind of anti-robot "features?" I occasionally get the "slow down" page, but nothing else.

>>9873 Oh, okay, thanks. Yeah, the number on the right does seem to be roughly equal to occurrences + siblings, though it's since dropped from 206 to 118. I am using the PTR, I assume it's something to do with that.

Is there a "do maintenance" button somewhere? I'd rather not have hydrus decide to lock up by itself when using the idle timers and I rarely remember to exit the program early enough for it to still be a decent time to do maintenance then. I do however have periods during the day when I know for sure I won't be using it for however many minutes.

>>9875 They've had different stuff over time. Beyond the basic 'slow down' responses if you hit too hard, they used some very strict Cloudflare javascript testing for a while, and at other times some pseudeo-browser-detection that did some header User-Agent stuff. The strange thing was these rules only kicked in in certain classes of page on their site (I think gallery pages, mostly, although I think they do anti-leeching image responses if you try to get a File URL without the right referral header? Not sure). I don't know about now. They contacted me some time ago, asking me to stop support completely because bandwidth from automatic sources was so much of a problem. After talking about it a bit, we agreed that I would make the current restrictive default sank bandwidth rule of I think 64MB/day. (EDIT: I just realised this may be what you are seeing–the restrictive default client-side bandwidth limit causing the delays, rather than specific import errors, as I first read. You can check your current bandwidth usage and rules under network->data->review bandwidth usage.) Afaik there is still no way to paypal/patreon/whatever them to help with bandwidth costs, but last I heard, they had a proper API in the works they were thinking of granting some kind of tiered access to. I generally don't recommend getting stuff from them if you can find it elsewhere easier. The big downloader overhaul is also hitting a big milestone tomorrow, so our download options for smaller boorus and so on will only open up from here on out.

>>9878 Alright, well I'll pause my downloaders for now. It just seems like they have the most pictures, and other boorus ban anything western. Thanks, chief!

hello frens I just got hydrus server and client system working on my local fileserver >CPU: Intel Xeon E5540 @ 16x 2.528GHz is there any way to make the client.py process more than 2 threads? I am only loading 2 threads of 16 and could definitely be using this more efficiently during downtime. Any ideas on tweaking this?

>>9848 services->review services->advanced service-wide update->one-time mass import od delete using hydrus tag archive

>>9877 Not that I'm aware, but you can disable idle maintenance in the settings and limit it to when closing the program. You can do the maintenance manually in database > maintain. If you're connected to the ptr check in services > review services > remote > ptr, process now button in repository sync. This tends to freeze the client until it's done, though. Maybe it's better if, when you're ready, you close the client and let it do it's shutdown maintenance.

New to this, sorry if it's a stupid question I'm organizing a collection of +10K images for a friend. Is there a way to give him just the tags, since we know both our collections include the same files/folders, and also the same file path of 'C://Images/. Like the 'Database backup' but not huge because it includes the images.

What's an easy way to group several images to indicate that they belong to the same set without actually giving distinct titles to each set?

I have read the manual, and I still don't know what's the propose of "System:inbox", "System:archive". Since the last update almost all my files are showing in the Inbox page, why?

>>9921 It's just a way to keep track of new files. Like an email inbox. Archive them if you don't care, the only difference is the little letter icon on the thumbnail goes away and the files aren't in the inbox anymore.

>>9902 There are two ways to do it. You could export and import all of you tags via services->review services->tags->local tags->advanced service-wide update or set up the tag server, that would function the same way as the ptr one.

My PTR sync is taking an exceptionally long time to get through some updates, and using nearly no CPU while it says its processing (current one its processing has been 1.5 hours with hydrus locked, average cpu use 0-5%). Is this the expected behavior?

>>9931 Thanks, I think I'm gettting the hang of this nicely I'm trying to get the downloader to work with an *.booru.org site. I've duplicated and modified everything that references furry.booru.org to *.booru.org in Network->downloader definitions->(manage gallery url generators|manage url classes|panage parsers) and linked it with manage url classes links. Now whenever I try to download something, it finds the right number of tagged images to download, but it says ignored and "no data found in document!" in the notes. What am I doing wrong?

>>9936 it might help if you run a full analyze maintenance operation on your db putting your db on an ssd makes a massive difference too, when i upgraded ptr update times went from hours to minutes

Hi, brainlet here So basically i want to know if i can do what another anon asked in another Q&A thread >>3209 (number 2 question) in which the dev told it wasnt implemented yet (it was 2 years ago..) I would like to know if its possible now, and if yes then how could I use the client when the harddrives are not plugged in. Like i tried it the other day but it said that it couldnt open because the files were missing and my options were to select the correct location or cancel withour opening the client. thx in advance

>>9945 There's an option to make multiple databases, you could have one for the images on your computer, another for the images on your external drive. Don't think the two can interact though https://hydrusnetwork.github.io/hydrus/help/database_migration.html

Is there any feature like danbooru "pools" where you can group images together that belong in a sequence? I think there's a feature like that in hydrus but I'm not sure how to access it.

>>9954 you can use page: and chapter:, and then change to sort by chapter and page. You can also create your own sorting method in the options

>>9936 >>9941 Thanks for replying, more recent usage after running all verification checks has shown continued freezing (when doing services → process now, and when leaving the Hydrus client idle to pick its own processing timing). The client never unlocks after it reaches this state, for example idle overnight → left it for 8 hours while I worked on the PC the next day and hydrus never became responsive). I got around this mystery issue by setting it to process for 8 hours as part of the shutdown procedure, which I had noticed never freezes; since allowing it to do that, it doesn't freeze when idling anymore. I suppose whatever the issue was has disappeared by finishing the PTR sync. I'll pick up an SSD some time and migrate the files to it as well.

When downloading with hydrus from yande.re with is it possible to make it download the JPG version (if avaiable) instead of defaulting to the original PNG? I'm not sure if there's an option or a change i can make myself to make it happen (and possibly toggle it back when i want the PNG) Also i'm not sure why but when downloading from danbooru and yande.re, the rating tag (safe, questionable, explicit) is not taken, while sankaku gets it normally However, in sankaku, some pictures (honestly i've found just one for now, but there might be others in the future without me noticing) don't seem to get the tags even with the option to get all tags checked (specifically it was post 716789, it only got medium:tags and nothing else) The same image gotten from danbooru managed to get all tags however

>>9954 I just make the pool:tag myself honestly, altho it gets kinda shitty when an image belongs to two pools and i tag it as such, because when collecting by pool, that image is going to make a collection by itself instead of getting collected with the correct pools

My display res. is 1366x768, so some settings windows (e.g. 'manage subscriptions') are a bit too big and I have to use the scroll bar to get to the buttons on the bottom. Are there any settings that tweak it?

what systems folder systems did you all use before you moved into Hydrus?

>>9590 >>9591 >>9715 >>9778 Same broad issue for me. I'd put a paid bounty on a new import option: Import by selected order (i.e. original timestamp). Could be a simple "on/off" extra option so people who don't care don't have to use the extra sorting step. (Drag files to Hydrus → Optional sort on Import window → Import normally or Import by optional sort rules → etc).

What ever happened to the rating system? I was just about to rate my stuff when I noticed the starts were gone for a long time. Is it still there as a namespaced tag or is it gone completely? >>9990 Windows explorer

How does one download from pixiv? Because whenever I try to add an artist id it keeps saying the error "Did not find any files on first sync".

>>3288 (471) >Yes, I'd like to have this for all exports, including mouse drag-and-drop. Some sort of default phrase, like: >[artist] - [series] - [unnamespaced] >that'll try to autopopulate, falling back to the current ugly sha256_hash.jpg. Was this ever implemented? If not, are there any plans to?

>>10022 services->manage services->add->local like/dislike rating service/local numerical rating service

not sure if this is the right thread to ask but….is there much of a pron video renaming database yet? thx

What's a good tag for this? Breasts resting/being laid on something. I see it commonly in 3d but I don't seem to have a tag for it in my db from boorus. Is there a booru for 3dpd? I know there's 3dbooru and idol.sankaku.complex but those seem to be focused around idols/cosplay sets. Or are western 3dpd whores allowed too?

How am I supposed to catch up with PTR? Processed tags are really behind compared to downloaded ones.

I'm new to hydrus, so apologies in advance. I've been kinda absent-mindedly making commits to the PTR (because why not), but I'm pretty sure they're almost all 'filename:' tags that get appended by various downloaders. How do I stop making a mess of the PTR and fix my mistakes? I'm figuring I can blacklist filename tags at import time, but I'm not sure which is the toppest-level setting I could change to make that a universal change.

>>10065 https://cuddlebear92.github.io/Quicksync/index.html

>>10078 Thanks anon.

>>9985 You're doing things like pool:thiscomic and pool:thatseries, right?

How do I rename a tag other than just replacing it with a new one?

>>10094 you can't

Is it feasible to move my Hydrus install to my synology NAS and just have it constantly running until I want to check on it? I'd like to just have it in a docker container or something and have it download my subscriptions without me needing the client open.

Is there any way to search through tags for my images that i got from PTR so I can see how worth it was to run this in the background?

Are there any other tag repo's out there? like other than the standard hydrus one? I have roughly 20 years worth of images saved from the internet but other than the occasional self tagging not a lot ends up auto tagged (other than stuff that's on the usual boorus/generic memes) Considering setting up my own repo eventually given the amount of files and tags/organization I have setup over the past few years

>>10100 Well, there's this but I don't know if it'll help you: https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts

Any hints on making a booru.org downloader? I tried to make one for a 0.1.11 booru on the latest release and it returns the message about not finding any results and suggesting that my tag search is incorrect, but I know for certain that it's not and it's just not finding anything for some reason.

How do I get hydrus to download dA pictures that are nsfw? It just ignores the files and has "veto if nsfw and not logged in" in the notes. I'm not seeing a way to log in, either.

>>10107 I think the 0.1.1 downloader is a little messed up, because I got it to work on one *.booru.org site, but not a different *.booru.org site and I can't figure out the difference.

(Couldn't post a new topic for this, 8ch fuckery I guess, but it probably belongs here anyway.) I bet you thought this was a thread made to suck on Hydrus Dev's cock. HAHA, WRONG! There's so many things about the Hydrus network that irritate the crap out of me, and it's clear that, as they're deliberate design considerations, they're not going to ever change. Therefore, I want to make my own. Problem is, I haven't really developed shit before, so I just want to know what I'm in for. Maybe I'm mistaken, but I'm under the impression that Hydrus Dev was only starting out when he began creating his as well, so maybe he could offer a few pointers, being already so far into his own project. Anyway, here's some notes I've been making on how I might do things differently. I imagine Hydrus Dev has already considered a lot of these ideas before, so being more world-weary, and having lost that old shine in his eyes, maybe he could tell me why they're fucking stupid and wouldn't work at all.

Pyrus Database - Store tags and other filesystem-related metadata in two separate databases for full reversibility. Since I consider reversibility so important, I think it deserves this level of security, to be stored in a more stable database of its own which gets written to less frequently and is therefore less prone to breakage. This secondary database will contain all forms of metadata the source filesystems could possibly support; not just the basics like filenames or creation dates, but also comments music tags, or even multi-purpose "description" fields. Users could also create custom keys such as "Likes" for archived YouTube or Twitter posts. Naturally, this level of data retention may seem redundant to many users, so it wouldn't be mandatory. - Also store the full source paths of programs, or, optionally, just the final directory name. When certain files are "marked", they may all be stored (or at least symlinked) in a single internal directory with all of their original filenames. This would allow certain flash files, and even games or programs to function normally within the database. Naturally, these directories should be excluded when you start the server up. I'm thinking of a hierarchy like this.
db/.
db/everything else
db/symlinks
db/symlinks/(Games/(Visual Novels/))euphoria
db/symlinks/(Programs/)Hydrus Network
* "symlinks" subdirectories up to user discretion, even if they don't seem necessary to me when programs must work with relative paths for this level of abstraction to work anyway.
- No silly client-side case conversions. Come on, I mean we have the full fucking range of the UTF-8 character set here, not 7-bit ASCII. If users really feel like it, then maybe include an OPTION to use this kind of filter. With this and the previous point, some measure of reverse compatability with the Hydrus PTR might be possible, but that sounds like too much of a hassle. - Since community spirit is important for a federated user participation-based booru project like this, maybe even throw in a third database for user ratings and comments.

Pyrus Client - Develop a GUI and TUI in tandem, or if it has to be a GUI, at least don't use a terrible library that forces you to look at the hideous default GTK3 theme. - Don't do anything absolutely fucking stupid like develop a garbage "Media Viewer" implement and integrate it so deeply into the operation of the program that other features must get held back for it. - Allow users to create custom execs for each filetype, or even for entirely different database variables. If a user doesn't want to define another set of execs for their client, then they can simply write "rifle"/"xdg-open"/"whatever" into each field. Since we've abandoned the Media Viewer, which is used to untangle the messy jungle of a filesystem behind the curtains, we can go even more abstract and create variables for several client-specific factors to be passed as arguments for the exec. %s might mean all selected files %a might mean all files (this is dumb and dangerous) %v might mean all files in the active view and so on. For example, and that is if that's the way the program works, an exec binding might look like this: Bind Control-E: exec 'Image Viewer.exe' – %v Which, though it'd make a ridiculously long command, I don't think would have to be particularly RAM or CPU intensive. After all, this is the kind of thing ordinary file managers like Ranger do. You could get a whole lot more complicated in the long-term by introducing a whole scripting language which can bring up windows in the client itself, or use if statements and the like, but maybe it'd be best to adhere to the KISS philosopy. So about how stupid are my ideas?

where are the proxy settings? google didn't help.

>>9327 (OP) So, I planned to organize my art collection. All pictures are saved from gelbooru and all are named as "gelbooru ID number, spacebar, characters and artist tags with spacebars between them". Now, the questions. 1. Can I somehow ask Hydrus automatically download from gelbooru a group of pictures based on IDs of existing ones, and selectively add tags (with namespaces) from the said booru? Or add existing pictures directly to it, but still do the tagging based on their IDs only. 2. Can I link every picture to its respective gelbooru link based, again, on its ID, so I can later synchronize tags of selected namespaces with the booru's ones?

>>10140 If everything is on your computer like that, there's an option to 'add tags based off filename' when you import. That'll bring you to a screen where you can make some RegEx to parse tags out of them. Test a small amount before you import everything. This is an example of what I ended up doing with a bunch of files from Pahael, that seems like it has similar naming conventions. If you've never used RegEx before, I didn't either before the day I did this, there's links in the program that'll bring you to 'how to RegEx' and a really nice practice site, plus you can see what I did in the corner there. It ain't perfect, but it'll get you 90% of the way to having everything tagged

>>10141 Yes, I saw 'add tags based off filename', but the problem is I don't have original tags in my files because I changed them manually. Only IDs are left untoched, that's why my request is that strange.

>>10022 Sorry, I wasn't asking about the file manager, what I want to know which methods you used. Like which file names and folder hierarchies did you use to sort everything without tags.

>>10041 Thanks. >>10053 I imagine resting_breasts or something, consdering how boorus tags work. >>10143 GIF/PIC/VID folders, each with 2D/3D/Nip/CG subfolders and random pics thrown in. I didn't organize it much before, most of the pics were in to be sorted folders.

Also, is there any way to make it show the rating on the pic, like it originally did with the stars?

Is there a way to set it to constantly restart itself so I can get all the maintenance done?

>>10162 No need, just set the maintenance time to a couple of hours and start it once

>>10163 Thanks

silly question: i have a headless server connected to a large RAID array, which i would like to run hydrus on (among other things). is there an easy way to run some kind of headless hydrus server on this machine, which can download from pixiv/DA/etc in the background, and then connect to it from my desktop without having to use VNC? or is running hydrus locally and then connecting to the server with sshfs/nfs/etc a better bet?

Does Hydrus have a way to organize parent and/or sibling relations between posts like in pic related? I guess you can copy a hash of one of the files and add it as a tag to every related file, but that's kind of clunky.

>>10166 That's a feature yet to be added. Closest thing you can do is stuff like >>10083

>>10166 >>10167 Hydrus does have this feature, if you think of it as duplicates rather than parent/child. Pools I think solve a different problem and that sort of thing is best solved with tags anyways, something like "pool:unique name here" works pretty well for grouping things that aren't duplicates but are related.

Is there any way to force the old download system? I'm trying to redownload files by their MD5 and it kind of just crashes on the new system.

>>10217 Also, I'm doing this for the tags, so something like the URL downloader won't work.

when I open a small file, Hydrus zooms in and horrible artifacts show up. Is there a way to deactivate the auto-zoom? I fail to find that option. Also is there a way to change the interpolation algorithm? It would be useful to things like pixel art that work better with nearest neighbour.

>>10219 yes it's in the options

Can I get a subscription group to not publish to a single page, but rather multiple pages under a page of pages so it doesn't mess up my top bar of pages?

>>9327 (OP) Hy, very new to Hydrus, working on getting used to it and loving is so far. Thank you everyone who has contributed to this… I'm not sure on one thing though… Is it possible to filter out tags in a search? A way to say "return media that does NOT have this tag" I've also looked at blacklisting, but I'm not too sure on that yet…

>>10239 Yes, like most boorus you can filter tags by preceding them with a hyphen

>>10240 That's fantastic, thanks for the fast response.

why are my pixiv downloaders fugged? I'm using v325 and the default pixiv url classes and parsers but not getting much luck. am i missing a step? here's a note: Page Parser pixiv single file page parser - new layout: Content Parser creator tag: Unable to parse that JSON: No JSON object could be decoded… (Copy note to see full error) Traceback (most recent call last): File "include\ClientImportFileSeeds.py", line 1052, in WorkOnURL all_parse_results = parser.Parse( parsing_context, data ) File "include\ClientParsing.py", line 2098, in Parse raise e ParseException: Page Parser pixiv single file page parser - new layout: Content Parser creator tag: Unable to parse that JSON: No JSON object could be decoded

How do you download from pixiv and have it tag each image with the respective pixiv illustration ID? I tried so hard to figure this out on my own but I swear nothing works. I don't understand what to do with the text things on "Hydrus-Presets-and-Scripts", so I just dragged the images into "import downloaders" but even so "all tags" still won't include illustration ID. This is the one thing I need most but I'm not seeing it.

>>10246 The nips changed things yet again, but should be fixed tomorrow (and hopefully more permanently) thanks to effort from a user, and pixiv tag search is back as well. Let me know if you run into more trouble with it.

>>10268 Thanks for the response but I appear to still be hitting some snags; I haven't tried every url class and parser but i'm getting some more errors about JSON parsing or something or other; any ideas?

Is there any way to auto-page number comics? Like, select all the pages, then hit a button and it will pass through and tag them with page numbers in the order they're in?

What happened to the deviantart login option? I noticed it was missing when I couldn't download anything from that site.

I'm trying to run hydrus from source, it seems to work fine until I place my existing hydrus install DB folder, then it crashes on launch: https://hastebin.com/ubasipoyey.sql If I use a clean db folder it works perfectly fine. Where did I mess up?

>>10299 This is happening to me too, it gives me an error saying "ImportError: No module named ordered_dict" when I try to launch from-source-Hydrus + my main DB, though checking all the scripts it seemed to indeed be present in them according to IDE.

Currently i am using xnview mp as my image viewer and i love its configurable interface. I consider it the gold standard as far as folder based viewers go. I ve tried using hydrus in the past but found it lacking in many aspects so i never used it beyond doing some testing. My biggest problem is the UI but i have some other questions/feature requests as well: 1. Is there a public todo list/roadmap? Or a more extensive FAQ than the one in help? 2. Why are 2 thumbnails generated for every image? Seems like a waste of space. 3. Image viewer as a tab option? 4. Image viewer control options? (hold down LMB to 100% and pan, release to fit) 5. Browser drag to select? 6. Browser preview on right side with same controls/options as image viewer? 7. Custom selection tags sorting that sticks - creator,series,character,meta,rest

>>9327 (OP) is there a way to add tags based on directories higher than the third? I only see one, two, and three under the "add tags based on filename" option when importing.

>>10319 Regex.

>>10322 I managed to get hydrus to accept my regex without any errors in the add tags based on filename window, but I still don't see any tags listed. What else do I have to do beside just past in the regex? here's the line in question (?=\/[^])*([\w*\s*\d]*)

>>10327 >>10322 never mind, I got it. [\+]*[^_\/.]+

is there a way to see stats about your database like your most common tags with parents and siblings considered?

I've recently started using this software but I want to quickly start tagging my existing images. hydrusadmin has a number of booru tag archives available on mediafire, but these only go back to around 2014. Are these any versions of these tag archives that are more up to date?

Is there a simple method to give a selection ordered tags in the selected order, i.e. selecting ten files and defining one button press to give them tags page:1 through page:10? My current workaround is exporting the selection, importing it again and adding the page tags through the import dialogue, which seems a bit cumbersome and has a habit of breaking the order when files have different mimes.

>>10331 What's wrong with PTR?

>>9888 >>9931 I'm trying to access this option, but this is all I see when I follow these paths. Is there an option I've not got switched on or something?

>>10365 I actually figured it out. Look under help if you're a retard like me, where you can toggle advanced mode.

Is there a sensible way to download/subscribe to xkcd (or other webcomics)? Off the top of my noob head, an RSS/Atom parser might be a reasonable way to work subscription, but trawling the archive could be harder to standardize/generalize across different comics on different sites.

Is there a way to search my subscriptions? It's hard remember what I've already added.

>>10370 Is network > downloaders > manage subscriptions helpful?

Hey fairly new to using Hydrus and I can't seem to get the pixiv downloader to work. I've set up a login and tried using a few different artists id's in a number of different ways just to make sure but it never starts the search. I feel like I've tried all the basic stuff but I might've missed something

I want to use the series category but I don't want it to display the header. Is there a way to hide it?

Is there any problem with downgrading your DB? I want to use the old downloader to get tags for my pics by using MD5s.

Any way I can grab a whole gallery, scraps and all from furaffinity? I tried the alternative with furry.booru.org but it wasnt doing it for what I was searching for. Furfinity wasnt listed in the simple downloader gallery option.

>>10374 Unfortunately, this will likely not work. You can sometimes go back a version or two if I haven't made any big db or object changes, but I have been doing a lot recently and I can't guarantee anything. If you can convert those md5s into gelb gallery urls using some small script, can you paste them into the 'url downloader'? That'll act a bit more like the old downloading, putting everything into one queue. Otherwise, if things crash, try breaking it up into smaller jobs. The new system is probably spiking hundreds of new threads with all those new queues in your image in >>10217 . Still, I think the url downloader is your best bet.

>>10379 afaik, It needs the login manager. I should have something ready in a few weeks.

>>10373 file->options->tag presentation

>>10372 Please try again with help->debug modes->report mode->network report mode on and let me know what you get in the bug thread or release thread.

>>10367 RSS often does not have like a 'page=3' parameter–it only ever shows the latest 20 or whatever posts, so you wouldn't be able to search far back. I recommend writing a parser for the archive. This would be 'simple' as far as hydrus parsers go, but it would be work if you aren't familiar with the new download system yet. If you just want to scrape the files and don't care about page numbers or anything, you could generate all the BASE_URL/1234.html urls for all the potential comic pages and then fudge your way through a Simple Downloader formula to just grab the image url.

>>10331 Sorry, no. The PTR is your best bet. Or getting a friend who has synced with the PTR to export a Hydrus Tag Archive for you, which you could sync with to 'local tags' or whatever you want to do.

>>10330 Not yet. I'd like to have a tag cloud and other top lists to summarise this.

>>10389 >try breaking it up into smaller jobs I tried. After a lot of crashed and about 2-3 days it finally managed to download all of the ones I put there. I think it was going slower due to the number of requests having destructive interference. Still, that was only 100 out of the ~75k I wanted to download. >till, I think the url downloader is your best bet. I'll try. I didn't even know it worked with search pages and actually managed to get the tags, GJ dev.

>>10389 >>10405 Also, the URL downloader is working great. It's not stalling for hours like the gallery ones.

>>10280 are bumps permitted? I'm still having this issue on v327; I can post long garbage failed post pre-parsing conversions if it would help and am open to suggestions

>>10407 Hey, a v326(?) client should have got some new url classes and parsers. Some of the old html legacy stuff as in your screenshot is still in the client, but it should no longer be active due to the url class being replaced. It should be redirecting to an api url that is full json. You should have: pixiv file page -> pixiv file page api Like: https://www.pixiv.net/member_illust.php?mode=medium&illust_id=69073387 -to- https://www.pixiv.net/touch/ajax/illust/details?illust_id=69073387 and pixiv file page api parser To parse that json. The 'new layout' parser shouldn't be doing anything anymore. Are you getting errors with these new objects, or does the old stuff still seem to be kicking in? If you have some custom url classes or something here, the v326 update may not have applied neatly. Trying a download with help->debug->report modes->network report mode may be helpful to diagnose this. It'll spammily say all the urls it is going through and which parsers it is trying. Please feel free to pastebin or email me the full error texts if this is not an easy fix. I will try and keep an eye on this Q&A thread, but I don't follow it too closely or I end up just writing long posts about bullshit and not getting the work I wanted done. Feel free to jump over to bugs or release thread, which I check more often.

>>10288 >>10299 Sorry for the problem here. >>10402 (10359)

brand new user here, is there a way to make a query that runs through a Gelbooru user's favorites while also grabbing tags? I found that using Simple Downloader to DL favorites works well, but I can't get tags this way, and Gallery downloading doesn't seem to recognize the Favorites URL

>>10410 Gelbooru doesn't have this feature yet, so you can't. It looks like this is on their to be implemented list, though. Alternatively, you can use Danbooru with fav:username query. https://gelbooru.com/index.php?page=tracker&s=view&id=515

>>10412 thanks, but I think I found a workaround by going to network -> downloaders -> manage default tag import options, and editing Gelbooru's options to always allow tags. now, each URL I insert into Simple Downloader will grab 50 images with tags now if only the Gelbooru admin would actually work on the site instead of pushing patreon subscriptions

Can I ask for help abt regex / getting tags from filename?

I have no idea what I'm doing so I let Hydrus make repositories and I downloaded and manually installed all the updates (I only realized later there was a full install option, wish I knew but too late) and I clicked process updates. How long is this going to take? The window is now completly locked up and the process took 4gb ram and keep writing stuff on the HDD and it's still happening today. The bar isn't updating at all. I have 70k files. I just hope that by the end of it I'll finally have tags.

>>10464 So I ended up killing the process. Loading up Hydrus shown no changes at all. Dammit, what the hell was the point? How do I get automatic tags on my files? What am I doing wrong?

>>10461 Sure, what's the filename like and what do you want to get from it?

>>10464 >How long is this going to take? A long while. >The window is now completly locked up and the process took 4gb ram and keep writing stuff on the HDD and it's still happening today. It's a memory intensive process that locks up the program, you need to just let it run while you don't use hydrus. >>10466 It takes a long time to download, after that it needs to process the files before it starts to actually tags pics. It started processing, so the downloading part is complete, but you need to let it do it's stuff. Hydrus automatically processes stuff at two points: one is in idle mode and the other when you close it and allow it to do stuff. You'll have to allow it to process the tags either by closing it and allowing it to do whatever (takes longer but, since the window doesn't just ignores inputs, it never crashes) or just by leaving it running on background, though it might crash if you try to use it while the program is doing something.

>>10469 Thanks, I get it now!

>>10464 >>10471 If you want to catch up with PTR quickly, you can use a full installation with tags processed too. That's what I did but I'm not sure how you would migrate your files one client to another one, I just did a clean install with no files. https://cuddlebear92.github.io/Quicksync/index.html

I have some tags for characters that are like asuka+soryu so I set them as siblings to their more prominent/correct counterparts, in this case character:asuka soryu. I just discovered that when I do a search with -character:* I get all the asuka+soryu tags despite them being siblings to and displaying as character:asuka soryu. Is there a way to search that includes siblings so that I can do searches like -character:* and not catch any thing that shouldn't belong?

>>10478 See >>10385 (5115) , that's a bug.

Is it possible to resize file list in "add tags based on filename"? It takes up most of the screen and I have scroll up and down to access advanced stuff or file list headers.

Planning to import 800 plus manga/doujinshi all of similar genre. I would like a method to guarantee each are assigned at least their title for grouping. Bulk rename utility was mentioned above for extracting metadata to the file name though I didn't see that option. Trying to avoid importing folders individually to titles.

Planning to import 800 plus manga/doujinshi all of similar genre. I would like a method to guarantee each are assigned at least their title for grouping. Bulk rename utility was mentioned above for extracting metadata to the file name though I didn't see that option. Trying to avoid importing folders individually to titles.>>10533 >individually just to add titles

So I'm downloading a gallery (from Sankaku) and while I'm watching it, I'm noticing a number of files are completely devoid of tags. I know the tags exist since I can view them through the URL it still grabs. What do I do to force it to fetch the tags? Just rerun the gallery download?

I'm running hydrus from an external hard drive due to the lack of space on my laptop, it worked perfectly fine for a while but now hydrus just keeps not responding when doing database maintenance, specifically vacuuming external mappings. Has anyone had this problem before or know how to fix this. I was also wondering if it would be possible to have hydrus itself on my laptop and the database on the external hard drive.

How do I force the client to process the tags on exit? That's the only time when I see any progress, doing it through this "process now" button just make the client crash

Is it possible to delete the repository updates? I was trying to clean everything out, and noticed that the files in it outnumber my actual files now.

>>10567 Set maintenance on exit to last a couple of hours and let it run

>>10579 Oh, thanks The last few mappings were painfully slow, this way is much faster, I didn't know you could extend the time

How much would hydrus kill my SSD if I move it there?

>>9327 (OP) Is there a way to set "system:" predicated tags for export folders? I was able to do it somehow in the past but I can't find a way to do it now. I'm a little behind, using version 326.

>>10494 It is still on old panel system, so I am afraid not for now. >>10533 >>10535 Use the filename tagging dialog, either through manual import or an import folder. Don't select to delete the files after import, in case it goes wrong. If the titles are in folder name, then select like 'last folder name' checkbox on the simple panel with 'title' as the namespace, but if it is more complicated, you'll need to use regexes on the advanced panel. Work on one doujin first before you try wrestling with hundreds in one go. >>10543 Sure, requeue the downloads, either by running the search again or by going into the file list and exporting/importing to a new queue or try againing, and change the tag import options to say 'get tags even if file already in db'. If the tags still don't come up, please send me some links in the bug or release thread and I'll see if there is a parsing issue. >>10545 I recommend you turn off vacuuming under options->maintenance and processing. External drives are pretty slow in some ways, and it'll probably take hours to run a multi-gig vacuum. The benefits aren't worth that time, so just turn it off. >>10583 You should be good. I run it on an SSD and know many users who do as well and haven't had any reports of trouble. A very busy PTR-syncing client will be doing, say, 5GB a day writing max. So lets say a max of 2TB a year. Modern SSDs seem to top out at about 1-2PB of writes, so even with a very busy client working every day, you are talking centuries of operation. Hydrus runs 10-100 times faster for many operations on an SSD. My (nice) laptop processes about 50k rows/s. I highly recommend it. Here is background reading: https://hydrusnetwork.github.io/hydrus/help/database_migration.html >>10590 They should be listed in the edit export folder panel, in the standard autocomplete dropdown. Do you see something else? Can you take a screen, to compare with my pic related?

>>10494 >>10627 I just double-checked this today as I was in the area, and my statement here is fake news. It is actually on the new panel system–can you please take a screenshot and post it here/email it to me so I can see what you are dealing with and maybe shuffle some things around?

>>10627 Thanks, I'll put at least the db and thumbs on the ssd then, 2TB a year is nothing, I assumed hydrus was writing much much more than that especially during syncing

How do I remove this second line in "location"? I successfully migrated the full database on the SDD, now I want to move just the actual files (and not the database or the thumbs) into my storage HDD, and I realized too late that the "client file" include everything.

>>10681 Oh nevermind, all clear!

>>10469 So I finally got the repo all downloaded and processed. But I'm not seeing any changes on my untagged files. Any steps I'm missing?

>>10692 Are you sure it has completely downloaded and processed everything? Check services->review services->PTR to see how far it got. Typically, due to a bandwidth rule to slow things down, it takes a couple weeks to sync completely. It does a little bit more every time it has a chance to do some idle work. Give it time and you'll be caught up and seeing plenty of tags for any popular file. If you are completely synced, let me know and I can give you some searches to test things out.

>>10694 No idea what's happening. It looks like it's completely done. I had 30k untagged files for a very long time (had Hydrus for a year and never used the PTR stuff until now) and I'm certain the number hasn't changed. Some pics that I'm absolutely sure should be on popular boorus still aren't being tagged. Maybe I need to copy all mappings to local tags?

Since I am entirely confused, I'll just ask like the baka I am. New user. I dumped a 200 file folder in the program. I went through them for sorting (more of a test, that particular folder was sorted already.), deleted a few, kept the rest. This just moved them from system:inbox to system:archive, but as I didn't have them tagged, I just moved them back to the inbox. How this is pretty much my first question: Should I only get things out of the inbox after tagging them or just treat system:archive as "tag this!"? The other thing is actually tagging because I can't figure this shit out. I downloaded the public tag repository but how do I add tags to a file? How do I even see the tags because manage tags window is just empty. And how do I create my own tags?

>>10706 Yeah, that looks all correct. Please hit database->regenerate->autocomplete cache and then restart the client. Has that fixed the issue? If not, open up a new search page, and set the buttons below the tag entry to 'all known files' and 'public tag repository'. Select 'system:hash' and paste this: 93805c3a61f2fe74bd1956872dac1f6bfd7d16093f7faed1b01bcb59f333c740 (its an sha256 hash) Let that search run–it should produce a 'hydrus' thumbnail, meaning a non-local file that has some tags, including 'creator:ilya kushinov'. Does it? Then remove the system predicate so you have an empty search. Type in 'huge' into the tag entry and wait a second–do you get a whole ton of tag results for 'huge x'? Try entering one of them–do you then get many thousands of the hydrus thumbs?

>>10715 I recommend you use the inbox as a metaphor for 'shit I haven't looked at that may be good or not' and the archive as 'stuff I know I like'. You can add tags as you filter your inbox, but you may find instead giving the files you want to tag later a 'tag later' local tag or local rating so you can find them again in batches is easier. You can set up a simple keyboard shortcut (like 't') to give a 'tag later' tag which can speed this stuff up. To add tags, open manage tags (F3) on a selected thumb and then check which 'tab' you are looking at. You'll probably have two–one for local tags and one for public tag repo. It probably defaults to 'local tags', which you can change in the options. Then type in tags like you would in a search and they will 'pend' in the list above. Hit ok, make sure it all looks good, and then, when you are ready, hit the new 'pending (x)' menu on the main gui to upload your tags to the PTR. You can also save this up and only upload in batches, rather than always be uploading five tags at a time. If you don't see any tags at all, not for any files, please check my post just above, >>10718, and let me know what you see.

>>10719 Ok… I think the repo I downloaded manually might not work? The one I got from "I don't know what I'm doing" has another name and this one has tags when I search. Also for some reason a few images are now tagged.

Hm, alright I think I understand how at least the local tags work now. I'll probably try with my own autism before I use the public repos. One thing though. In the repo that works tags have different colors. All the ones I make are the same two colors. blue for words and a greyish blue for the ones with a : in them. Any way to change that?

>>10722 Ok, I posted bit early. I made a character: tag and it was green. Sorry for being a newfag

>>10718 Thank you, looks like it's working now including your test There's still quite a lot of untagged files but still much less than before. Thanks m8, and it's so fast on a ssd too

>>10722 You can set the color of any namespace in the settings menu

>>10731 right, found it. Another thing came up: Multiple images belonging together. Example being a short comic spanning two files. Do I have to tag those under "shitty comic" and "page1" or is there an option to bundle images like this?

>>10736 This is a long subject, but tl;dr is: There is no good solution for this yet. Doing page:n tags is a bit cumbersome, and the ui isn't built for easy page/chapter navigation yet. I have plans, but I generally don't recommend you import paged comics, or little 2-4 page mini-comics, just yet. Unless you find a way in hydrus that really works for you, keep them in a separate folder/cbz and use a different program to read them for now.

>>10627 Thanks! Exporting using 'system:' predicates works fine, I was just being an idiot. I VNC'd into my machine using a pretty low res monitor, which vertically compressed the 'edit export folder' dialogue, hiding the tag suggestions. Many panels offer scrolling, so I haven't noticed the issue until that dialogue.

Any way to get permanent thumbnails? The previous image program I used maintained a permanent thumb DB. I'll be perfectly fine with just setting thumbnail cache = 3gb / timeout = 10y (their max values), but just curious if perma-thumbnails is a thing. (I'm very often changing to artist/character reference tabs while slowly tagging the last decade of images and the default thumb settings meant those were being forced to reload their thumbs quite often).

Recently, I get this error when trying to start Hydrus:
Traceback (most recent call last):
File "/opt/hydrus/client.pyw", line 20, in <module>
from include import ClientController
File "/opt/hydrus/include/ClientController.py", line 14, in <module>
import ClientCaches
File "/opt/hydrus/include/ClientCaches.py", line 5, in <module>
import ClientRendering
File "/opt/hydrus/include/ClientRendering.py", line 2, in <module>
import ClientImageHandling
File "/opt/hydrus/include/ClientImageHandling.py", line 3, in <module>
import cv2
ImportError: No module named cv2
How should I go about trying to fix it?

>>10755 Is this on Arch? Looks like they somewhat messed up the bindings with the new version: >>10732 (10656) >>10737 (10656)

>>10754 Can you describe more what you mean here? What you are looking for may already be what happens. Atm, hydrus stores a 'master' thumb file at 200x200 and a resized file at 120x80 or whatever you have set in the options. It loads that resized one whenever needed and stores it as a bmp in memory. That cache timer only handles the memory timeout. (and the memory cache is obviously cleared every time you shut the client down) Depending on various factors, your client may not have its resized thumbs generated, so if they are loading slow the first time, it could be because the client is noticing they are missing and going back to the master to generate a new resized file. This will eventually happen once for all files if you change the thumb size in the options. If you find thumbs load slow every single time, let me know. This can also be true for users who store their thumbs on an HDD instead of an SSD. There is some advanced info about splitting up your db across a system with both kinds of drive here: https://hydrusnetwork.github.io/hydrus/help/database_migration.html Which can greatly speed a client up.

>>10764 Hmm, my only real point of reference was the other program I have used for ages, which seems able to load them notably faster I guess, so I came to the wrong conclusion. My bad. The speed of Hydrus thumbs isn't a problem for me either, I just got my head in the wrong spot and then was thinking like I should try and stop it from grinding the drive to general new thumbnails or something. It could also be that this hard drive is just simply way worse than the hard drive the other program is using. Hydrus is currently sitting on its own HD, with the original image collection I'm importing from on a different drive. Timing of loading for reference, with both programs loading approx same # of image thumbs on the screen: Hydrus: Thumb population ~1-3 seconds, can usually hear hard drive during entire population (leans to 1sec end much more than 3, but sometimes "quite" slow) Firestone: Thumb population ~0.2-0.5 seconds, no noise I'll move Hydrus to an SSD sometime.

I have a large set of images I downloaded from a 4chan thread which are not tagged in the PTR. I know that the greater majority of these images are posted and tagged on e621. Is there some way I can automatically search for these files and pull their associated tags? Related question, I have high res patreon versions of images which have been publicly released in lower resolutions. Is there an easy way to parse the tags from the lower res versions, which are tagged on boorus, on to the high res ones in my archive? I've been doing it by hand but it's a pain in the ass.

My client hangs on the maintenance window every time I shut it down. No progress indicators to indicate that it's doing anything, nothing in the client log or debug client console. I'm on windows 10.

>>10772 That's interesting. Maybe they do something clever and cache 'directories' worth of thumbs together or something. Or maybe pre-load thumbs depending on where you are navigating. They probably have faster decoding and rendering tech too, as well. If you are interested, you can see the master and resized thumbs under install_dir/db/client_files, and then look for the 'tXX' and 'rXX' folders. The client pretty much just pulls from there, converts to bmp, and then fades/slaps them on screen. If you run db or thumbs from an HDD, making sure you are defragged (and maybe with some pseudo-file-ordering, so files in the same directory are physically near each other on the partition) can really boost performance. If it hasn't run in a while and you only fairly recently added most of these files, their thumbs could be all over the place on your drive, which is adding the ~8ms latency for each one as your drive head bounces around.

>>10773 As far as I know, a lot of the e621 tags are on the PTR, so I suspect your images have been slightly altered by Cloudflare or similar and hence aren't exactly the same files as on e621. You might like to test this first by figuring out a fairly common artist in your collection and doing an e621 search, with tags, for that artist. If the downloader pulls new files (likely with higher res or ~+5% file sizes), then what you have are probably not exactly the same files and you are probably out of luck. If you do have the exact same files and they just happen to not be tagged on the PTR, the easy solution is to run the same e621 searches for the most common artists and fill up the blanks, and the advanced and more precise solution is to get into File Lookup Scripts a la: https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts/blob/master/Scripts%20-%20Tags/md5%20-%20e621.png Which is a pretty prototype system I made a while ago and lets you look up tags for individual files via md5 searches under the manage tags dialog. Turn it on under options->tag suggestions and import new scripts under the network menu (although you may have e621 already in your client). To merge tags across different versions of the same file, check out the duplicates system. It is a first version, so the workflow is a little awkward and rough, but it ultimately works. Check out the help here: https://hydrusnetwork.github.io/hydrus/help/duplicates.html >>10778 Can you go to options->maintenance and processing and turn off shutdown processing and try again? Does it work ok then? If so, something in your db maintenance is eating up so much CPU that the gui is hanging. You might like to try, when you have some time, running help->debug->debug modes->force idle mode and letting the client sit for a bit so it can sort out whatever huge job it needs to on its own time. I'd be interested in knowing more here–what job is running so hot and whether it clears out ok!

What's the correct process for removing a PTR tag sibling from local machine results? I see that the client says if you petition it on the PTR it will be removed locally whether its accepted by a reviewer or not, is there a non-intrusive way to unassociate that PTR sibling and return the originals? It doesn't seem like adding the inverse on local nor using the censor function accomplishes that goal.

>>10783 I'll try defragging later on and see if that helps, thanks. The entire thing has been imported over the last couple months so far (~60k images now). The speed isn't really an issue overall either, I just confused myself with what was likely an unfair comparison (i.e. what is perhaps a thumb DB optimized by relying on roughly sequential files in a folder as you mentioned).

>>10787 There isn't a good way yet, although a comprehensive "I want these kinds of siblings, not these kinds, and I prefer this one over this one" control for sibs and parents remains a long-term goal. I may be able to chip away at this in small parts rather than wait for it to come up in the big job poll. If you just want to remove something real quick now, petition it with a reason like "Do not approve, I just want it gone from local", and I'll deny it with no harm done. Copy and paste the same reason and it'll all come to me as one batch job I can clear out in one click.

Alright, So big oof I did , and I need help I managed to delete all of my images, and recover them back. I am now trying to import them all back, but none of them are importing properly (at all) What do? Clean install? Is there some workaround I can do?

>>10793 I think hydrus remembers the hashes of deleted things so if you deleted them in hydrus then that might be it. I would also like to know how to import previously deleted files.

>>10796 >>10793 To reimport files that have previously been deleted, check the 'file import options' of the import context. A button for FIO is on the 'import files' dialog that pops up when you drop some files on the client. Inside FIO is a checkbox for 'exclude previously deleted files'. Uncheck it and previously deleted files will be treated as new. Let me know if this fixes your issue, >>10793 , but if the 'not importing properly' issue is an actual error, rather than 'x files previously deleted', let me know and we can get deeper into it. How did you delete all your files–from inside the client, or by accidentally clearing out the client_files folder in the db? Is there something I could do in the hydrus UI that would have stopped you making this mistake?

>>10797 I think my issue is more that the checkboxes don't show up for me. I'm using kubuntu 18.04. Playing around with qt and gtk themes didn't help. >>10797

>>10809 I should probably provide an image.

Can hydrus hook up to yiff.party the patreon scraper? it looks like the site might to belly up because bui is a stupid fucker who let normalfag camwhores bloat the servers and blow up the cost of running them

>>10797 Sry for the late reply, school shit. Anyway, I "Deleted the files" by moving all the files to an external drive using migrate db, then i tred to made a backup in that same directory. All that did was copy what was left on my c drive (not the images) to the external drive, and removed all the images from the client_files folder.. So now I recovered the images and if i try to import them, they arn't being placed properly / arnt importing at all. But funnily enough, if i drop the correct image to the correct place, the thumbnail will still be gone, but the image will come up if i double click it. Uh, I dont think you need to change anything in the UI, the backup thing said preferably an empty destination. Maybe somehow explicitly say that files will be deleted if they dont match the current db? idk. NeWay, any advice on what to do next? Clean install? (lul)

>>10627 I probably called the dialog wrong, this is the actual dialog Half of the options below are obscured until I scroll down, makes batch importing comics a bit uncomfortable because I have to scroll down to "#" field every time

Does the Source time of each file get saved when importing a gallery? Is there a way to sort files using this source time?

>>10832 Yeah you problem is definitely different from mine as I'm not getting those check boxes by those regex forms

When searching through your files, is there an inclusive OR operator for tags? Like if I want to show all files that have either of the tags I'm searching for, not just both

>>10820 >>10809 That is pretty strange! I have seen something similar on Linux where hidden/no border is how the particular window manager shows 'disabled' controls (where most window managers will show it greyed out in some way). But I don't think those checkboxes are disabled at any point in their generation (so it shouldn't be, say, that they were once disabled and didn't update visually on being re-enabled). Attached is what I see in Win10. Is the 'MB' dropdown next to the active max gif filesize active? There should be a checkbox after it, so what happens if you click in the '32' spinctrl and then hit tab twice? Do you get a dotted 'selection' border around where the checkbox should be? Can you alter its state with a spacebar hit in that case? What happens if you manually resize that window, do the controls redraw in a neater way, or maybe flicker as you resize to something more correct? Do you see checkboxes correct anywhere else in the program? file->options has a ton of them in different layout scenarios. If any show up right, please take a small screenshot and we'll see if there is a common theme to the hidden ones. Your shot here is also missing up/down arrows on the spinctrls and I assume a nice border on the MB dropdown choice control.

>>10828 There should be some support by default, but it was always a bit shaky due to the site being unusual in a couple of ways and obviously very error-prone. Try drag and dropping an artist page onto your client, one formatted like this: https://yiff.party/200110 It goes in the watcher for now, but I think it is probably more appropriate as a gallery (which would then be subscription-able). I understand there were some issues figuring out what to parse and how to get nice tags. This page has the latest attempt: https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts/blob/master/Download%20System/Parsers/png https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts/blob/master/Download%20System/Parsers/png/yiff.party_api_parser_2018.09.20.png https://hydrusnetwork.github.io/hydrus/help/adding_new_downloaders.html

>>10831 I am sorry, I have some difficulty understanding your English. Did the backup in the client delete your files in the external location? I do not recommend using the client's backup if you have a complicated db you made using the migrate database dialog. I will make sure this system provides a warning if you try to run it with a split db in future. In any case, I assume the backup choice location was a mistake? If so, we'll ignore it for this issue. I think the problem here is that the client thinks it has those files, so when you try to reimport them, the problem is not that they are 'deleted', but 'already in db', so it doesn't bother to copy the source import file to the client_files structure, since it assumes the file is already there. If you are running an older version of the client, please update to a new one, as this should be able to detect on file import that a file is missing and hence add it properly. If you are on a new version of the client, please show me what errors or status you see in the import window, like my image.

>>10832 Thanks. I will see if I can fix up that layout.

>>10835 Not at the moment. It is only used for subscription and watcher check time calculations. Your best bet is to go by import time, although this obviously has its limitations. You should be able to sort that list by source time, and then if you multi-select and right-click->show these in a new page, I think they'll come out in that order. This isn't a great or permanent solution, but it helps for subscriptions sometimes. I've always wanted good 'this file was created/posted on 2015-06-20' metadata, but there isn't really any yet. I guess it would be something like a 'date:2015-06-20' tag, but I am not sure. The import time that's currently used is always there, uncontroversial, and completely under hydrus's control, so I can't easily extend it to support other timestamp domains where it is less precise there may be conflicting opinions.

>>10840 Not yet, but I expect to work on that next year.

>>10853 Awesome, thanks for the quick response

>>10850 FileMissingException The thumbnail for file 000ad63cd6ef2353e368b85019fe9ed694fbbad6687f0421472f496c34061e6c was missing. It could not be regenerated from the original file for the above reason. This event could indicate hard drive corruption. Please check everything is ok. Traceback (most recent call last): File "include\ClientCaches.py", line 1889, in _GetResizedHydrusBitmapFromHardDrive path = self._controller.client_files_manager.GetResizedThumbnailPath( hash, mime ) File "include\ClientCaches.py", line 1078, in GetResizedThumbnailPath self._GenerateResizedThumbnail( hash, mime ) File "include\ClientCaches.py", line 340, in _GenerateResizedThumbnail self._GenerateFullSizeThumbnail( hash, mime ) File "include\ClientCaches.py", line 290, in _GenerateFullSizeThumbnail raise HydrusExceptions.FileMissingException( 'The thumbnail for file ' + hash.encode( 'hex' ) + ' was missing. It could not be regenerated from the original file for the above reason. This event could indicate hard drive corruption. Please check everything is ok.' ) FileMissingException: The thumbnail for file 000ad63cd6ef2353e368b85019fe9ed694fbbad6687f0421472f496c34061e6c was missing. It could not be regenerated from the original file for the above reason. This event could indicate hard drive corruption. Please check everything is ok. This was the error I was getting, but updating to the new version seems to be letting me import files properly. Thanks! Will update after I import more files.

>>10848 I think the check-boxes are just invisible. clicking the approximate location of the checkbox next to the no limit was able to enable/disable the spinctrl and was able to successfully import a previously deleted file using the same method. Check-boxes are invisible everywhere including file->options, couldn't find a single one.

Oh crap! Sorry, I posted the posts in the wrong thread! Here we go again… Anons, how can I bulk-download booru-imageboard image files in bulk? I get the error of pic related. I tried to install "bionus" "Imgbrd grabber", that's the thing that didn't want to work. I already followed instructions to reinstall the .dll for this, so I am at a loss what to do now. Pls halp ;_; Is something like this possible with hydrus itself!? I never tried… Originally: >>10868 (10654) >>10869 (10654) >>10870 (10654)

>>10871 Here is the image again… I really am a mess today, excuse me.

>>10872 …And of course you lads don't speak Kartoffel, so here is the translation: >The execution of the Code cannot be continued, >because MSCVP140.dll couldn't be found. Through re-installation >of the program the problem kan possibly be solved. (I already tried re-installing.)

>>10871 Yes, you should be able to bulk download files several websites, as long as you use gallery download system. Unless I've misunderstood you, this should help: https://hydrusnetwork.github.io/hydrus/help/getting_started_downloading.html

>>10871 Hey, I am the developer of hydrus. I'm afraid I can't help much with imgbrd-grabber as that is a different program to the one I develop. I do not use it myself. Hydrus also lets you download from boorus, but it works in a very different way. If you would like to get started, please check out my help and getting started guide here: https://hydrusnetwork.github.io/hydrus/help/index.html That said, that MSCVP140.dll error looks like an error with a missing Microsoft Visual C++ package. Try installing the redistributable package here: https://www.microsoft.com/en-us/download/details.aspx?id=53840

>>10863 I am sorry, I am not sure what is going on here. My best bet is wx is not playing well with your window manager in some way. I don't do anything too controversial with how I create and place checkboxes, and if they are all missing and other funky stuff like spinctrl arrows are hidden, I suspect some .so file somewhere is being called the wrong way by wx itself (or the gtk interface, or some other low-level shit I don't know much about). I am not an expert at Linux window managers. I know some of the discord users had somewhat similar display issues to this, so you might want to ask around here: https://discord.gg/3H8UTpb Or otherwise see if you can tell your Linux to launch hydrus with a different window manager/style? Is that possible?

I have a few questions with tumblr nuking porn and all 1. Where can I put in my tumblr account? It doesn't seem to be an option in the manage logins section for me 2. I guess this one's a bit of wishful thinking but is there any way to get hydrus to download from the links in the posts themselves? i.e a lot of artists put links to uploadr, imgur, MEGA, etc for higher quality images / videos / animations 3. I'm having some slight issues importing a certain tumblr, the first one I did seemed to go fine, it imported all the images / gifs from the posts themselves but this second one only shows like 12 images even though it says it downloaded 80+ items

>>10890 Also wondering this. Trying to save a few blogs before they're killed off. One of them is one of those stupid (semi-private) ones that only open in the sidebar in the dashboard so searching through the posts is a pain in the ass. Any chance of the downloader working on those before the 17th and the blogs all get destroyed?

Is it possible to download a board on pinterest? I don't know if I am using the program wrong but it doesn't seem to give any returns when I enter the URL to a board.

I'm using the gallery downloader (f9 > left arrow > right arrow) to download from pixiv, but for probably 19 hours of the 20 I've had the tab open, it's been frozen on the same number. At first I thought the client was DLing too fast, so it got itself ip banned from pixiv or something, but I slept on it and it's been stuck for so long. I attempted to restart the client, but it says "1 page says: 241 queries are still importing. Are you sure you want to close it?" Here's just a screenshot of what I'm looking at I guess. I've never done this before, so I have no idea of how it's supposed to behave, or if it broke indefinitely in my instance of it.

>>10898 Sometime between my submitting this and now the client itself froze, and the number shot up from 643 (what it was frozen at) to like 741. I say "like 741" because it actually froze again at the time I started typing this post, and I thought it said 741. But it unfroze sometime after I started typing. It's currently at 921 and climbing in real time, as it does stuff. So perhaps pixiv ip bans last for almost a full day if the downloader goes too quickly? I believe this PC had internet access without downtime since this started, so not sure what the holdup was. If the following context is worth anything, I remember adding another pixiv "numerical artist id" a few hours after it froze, and within an hour prior to making the above post I added a few more (I think three).

>>10898 >>10900 Sorry for posting three times in a row, but to be honest, before I showered just now I saw a popup at the bottom right of Hydrus just telling me it's ripping a subscription of mine, from pixiv. Well I just finished showering and the progress bar for that subscription popup hasn't moved, nor has my gallery download tab (though the number is more inflated than before. Is Hydrus just willy-nilly getting itself ip banned from Pixiv? I used to use a program called "PixivUtil2" to rip from pixiv, and I'd download many thousands of images using it a day. You can set your own delay between ripping images, but it defaults to 5 second delay I think, which I haven't changed. My may concern with it was I saw my internet go offline during it once, and it failed to rip images, but it didn't produce a .html file or anything telling me it failed at all, and what it failed (but it will do that if the image itself is offline, not your internet)… but I used the program anyway… If Hydrus is just getting me ip banned from Pixiv and thus limiting its download speed to clusters of hundreds of images every ~18 hours that is fucked. That would mean I should still rip from Pixiv using PixilUtil2 for insurance in case the gross downtime Hydrus habitually induces might cause me to miss images that artists delete before Hydrus picks them up. I thought I had a better conclusion to say that that but if so I forgot it…

>>10902 I'll just post four times in a row to say this sucks… the subscription popup is still on the same amount of progress as it was four hours ago, as is the gallery download tab. It just isn't fast enough at saving images if Hydrus' method is to run itself into a wall by downloading hundreds too quickly, getting ip banned, then doing it again as soon as it's unbanned. At least that's what I assume is happening, since the same images remain "downloading" for well over a dozen hours. In the end I will have to rip using PixivUtil2 anyway as a backup since Hydrus can't rip it fast enough (at this rate it literally only ripped 981/48,204 in 27 hours, but PixivUtil2 can rip thousands a day without hitting any wall [granted it does it one at a time with a 5 second delay between each image rip but still]). Also isn't it worrying to just run something that gets yourself habitually ip banned by pixiv. It's not even ripping the images at a rate I consider worth it. I think I forgot to mention that I can't close the subscription popup by right clicking (the usual way to popups that appear there), so all around this is no good.

>>10904 Hey m8, I am putting the release together now so don't have time to give you a good reply, but I think it is overwhelmed with the number of queries you have running there. Try pausing all those downloaders (ctrl+a the list and then click pause files), restart the client, and then unpause a dozen or so at a time.

>>10907 Hi, thank you for the reply. It's ok if you don't have time to give me a "good" reply by your standards- any help is appreciated, since otherwise I feel my client has an 18 hour downtime between progress. I didn't notice at first that doing what you said did anything, but I eventually noticed the "f" tab started showing a box, all the way down. When I did what you said, I was able to close the client. It's currently still closing but yeah, I will temper how I treat the downloader so I'm not just flooding it and hoping it works. Thanks again. To be honest with PixivUtil2 I only had confidence in running one instance of the program at a time (to rip only one artist's gallery at a time), so if I can do a dozen at once that'll really do it quickly for what I know.

>>10907 My client is still shutting down 14 hours later (I went to bed and just woke up) and I'm wondering if it's ok to force terminate it in task manager. The taskbar icon for it turned to the white generic file icon with the fold at the top right (don't remember if that's normal), and the tiny window itself still says: "shutting down gui" "waiting for daemons to exit" "SychroniseSubscriptions" I thought since the only thing it was doing was ripping from pixiv that terminating the program in task manager would be fine but I was hoping it'd resolve itself but maybe it's not. It's not "frozen" according to windows since I can still click without it saying "not responding", but it's stuck.

Is it possible to download entire Twitter feeds, or is it only possible to download individual posts?

>>10916 Nevermind, I found it, was looking at the simple downloader when I should've been looking at the Network tab

>>10893 >>10890 bump, nearing the dawn of the final day

>>10915 I ended up just forcing the client closed after at least 36 hours of it shutting down, but just alt + f4ing the window didn't close it (but it made the window disappear as well as its taskbar thing)- I tried opening the client again but it said it was already open. Tried again twice, and it waited 10 seconds before the retry, but still no, so I forced it closed in task manager. It didn't corrupt my db or anything, and in limiting the downloads to 12 at a time it ripped like 2k more (981 to 2,895) before stopping, but it did stop. I suspect I'm ip banned from pixiv again and it'll restart when it's unbanned in another 18 hours… It's still been doing nothing for a few hours now, this time with only 5 queries "downloading file" (the rest not doing anything because I paused them, or the other 7 of the dozen I unpaused finished). One thing I don't get though is a handful of the queries (all pixiv artist lookups) have a 3-character string under the "items" tab, eg. "163 - 1Ig". I just assumed that means the pixiv account has pixiv ugoiras Hydrus can't rip, hence the "ignored" files but I don't know.

Is there a way to mass import tags from a booru for local files? I'm thinking it would be something like tag import options but that I can use on local imports, or even on a search page.

>>10849 thanks found the sadpanda option under simple downloader. worked pretty well, not perfect but way better than doing it manually.

>>10925 maybe try a different parser. there's two for tumblr https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts/blob/master/Download%20System/Parsers/png/tumblr_api_post_page_parser_2018.09.20.png https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts/blob/master/Download%20System/Parsers/png/tumblr_api_post_page_parser_-_with_post_tags_2018.09.20.png

>>10931 Looking at the documentation it seems they already were there in the list so no need to download them again but I'm still very unclear as to how I actually use those, I mean this page here: >https://github.com/CuddleBear92/Hydrus-Presets-and-Scripts/wiki/2.-How-to-Use-the-Parsing-Scripts Says that "For now the Parsing scripts only work for single files." which is a bit of a bummer considering I'd like to export entire tumblrs, although this info may be outdated? For instance it also says in the instructions that "1. Turn on the "file lookup scripts" in file>options>tags>file lookup scripts" but if I go to that section there's no such option at all, it's actually on tag suggestions but it only shows gelbooru and iqbd like in pic related

Does the tag censorship function actually hides images with a particular tag? It sees all it does is hide the tag itself.

>>10935 Go to new page -> download -> gallery and then select tumblr from the drop down. I used it and downloaded all the art on mutliple tumblrs, the only thing it didnt seem to grab was the tumblr hash tags but i dont really care about that. If you want to export entire blogs then use something like tumblrthree https://derpibooru.org/uppers/guide-to-tumblr-grabbing-tools

>>10935 Oh yeah and if you just want to grab your own blog then tumblr has that function already built in.

Do I have my proxy set up right? I'm trying to route the traffic through tor, but there are two boxes to fill in the proxies. There's one for http, and one for https. So far I just put socks5://127.0.0.1:9150 in both boxes. I haven't tested it out myself yet though because danbooru and pixiv are just giving me 403 errors, not letting hydrus download, but that's another issue I can try to figure for myself.

how can I change the bandwidth limit ?

>>10392 pixv isnt working for me. when i punch in a number it just spits out error! in status then says it will expire in an hour. i found the network report mode but it doesnt look like it has any useful information.

>>10964 I can confirm this exact error message. Only difference for me is the pixiv ID.

Is there a way to make Hydrus always max the vertical dimensions of the media preview window? It used to behave that way for me but then it stopped, on the same version number I think. In the same vein is there a way to restore the horizontal width of the entire left part of the client (media preview window included in there) once you've changed it, and can you set the dimensions for the media preview window+that entire left part of the client as the new default for all following tabs you open? Another thing I'd like is to automatically max the media preview window to the image you've selected. Obviously for browsing that isn't usable since images can have any dimension, but I tend to keep the client open on a tab with an image selected, but upon restart I have to manually eyeball scaling the media preview window to the proportions of my screen size and all that. Would be nice to just hit a button to automate it. I don't think any of this functionality exists yet, so I'm just leaving feedback I guess.

>>10971 Actually the proportions of the left side of Hydrus in my first tab stayed on restart, and the only other tab I had open emulated those proportions. Any following tab I open emulates the first tab proportions as well. So this is quite alright. I remember losing it on restart once though, but maybe I had multiple tabs open, with sizes discrepant from the first tab? Anyway I won't fuck with the first tab proportions of the left side, since I manually maxed the vertical res and made the right edge pixel perfect to the image it's displaying. So this is good enough for me. It's a hentai page, which is like 16:9 ratio vertically and is fine to standardize across any following tabs.

>>10915 >>10908 >>10928 Hey, I apologise again. I've had trouble fitting time in to keep up with threads this week. The client is generally ok with force killing with the task manager, and if you ever have to wait more than and hour, I recommend it. The new downloader I wrote is not great at handling many many simultaneous download queues. I do not recommend more than twenty or so active at one time. Try double-clicking one of those download queries, and its details will appear in the 'highlighted' panel below. This will give you more details, and likely a 'waiting on bandwidth' notification. You may not be IP banned from pixiv, but instead hit hydrus's bandwidth limits, which are fairly conservative. Hydrus is a relatively 'slow' downloader, intending to sync with galleries over days and be in it (and all the subsequent file processing workflows) for the long haul. Previously, the hundreds of simultaneous bandwidth checks were likely what was locking up your client. It sucks, but the whole thing works better than the old system, which was one download per page! You just have to manage the speed yourself now. I am sorry for your trouble here. If you haven't checked it out, I have some help here: https://hydrusnetwork.github.io/hydrus/help/getting_started_downloading.html >>10971 >>10975 This code is a bit of a mess, I'm afraid. Try the pages->management and preview panels menu to control it a bit better. By default, it tries to save the current width/height of the sashes of the current page on exit, but it sometimes spergs out and starts growing one pixel every shutdown or something. There's very little dynamic sizing in the client at the moment.

>>10929 Not really. This turns out to be a difficult technical problem (one I tried to fix by writing tag repositories). There are plans in the works to automate the advanced 'file lookup scripts', and there's a very advanced Hydrus Tag Archive solution which turned out to be a bit awkward to use for normal syncing. Your best bet for now is to sync with the public tag repo.

>>10957 Yeah, afaik, put the same string in both boxes. 403 is 'forbidden', btw. Maybe you got IP blocked on your normal connection, or maybe you have state-level blocking, hence the need for proxy?

>>10960 network->data->review bandwidth usage Be very careful with it. Don't delete all the rules, since hydrus will just fuck itself and murder your CPU and you'll probably catch an IP ban from the servers you are talking to. I recommend you only alter the existing rules a little bit, and revisit if it is still a problem.

>>10964 >>10968 I think pixiv changed it so only logged in users can access that API. If you haven't got one, make a throwaway account and put its details in network->downloaders->manage logins. If you are logged in and it still gives you the "error":true result, let me know. That "message" data is some moon runes I think, so we'll see if we can decode it.