>>12050 There isn't a single alt-chan that has the hardware to handle this amount of users and not have latency issues. Even 4chan itself had latency issues at times. The people that can't handle latency issues are too used to reddit and twitter.

It's because the script this site runs on is overengineered to "fix" latency. In theory it's supposed to be faster than 4chan on less hardware, but that's probably only true for the back-end.

>>12050 Better a slow site than a dead site. But yeah, the site has been a buggy for a few years now, whereas I've never had issues with smug or zzz. CAPTCHAs have always been an issue.

there's a few tells that this is not set up in a best way possible. even though https is there, http/1.1 is used and i guess that's the reason why ws connection is made on same host but another port. sure, connecting to it is deferred to when someone scrolls to the bottom of the thread, but that's still paying 2 sockets per visitor. ws is dead at the moment too, yet the script is banging its head against it. it would be interesting to see what kind of cache size is there, maybe it's just being trashed all the time.

>le tuned for shitty hardware website xD just buy more ram 8chan man

>>12050 Is that why the website feels like shit to use? Images take ages to load, replies take ages to get sent, everything is slow as fuck.

>>12072 It's as simple as buying a dedicated server and putting everything behind Cloudflare.

>>12162 There's certain facts of life about this site that make those 2 things not as simple as you'd think.

>>12164 I've done so before successfully with a fairly big chan. The only problem is finding a hoster that isn't ass for controversial content, but that just needs a few hours of research and some trial & error when you're new to ib hosting.

if theres suddenly tens of thousands lurkers the bandwidth for that to be very smoothly served becomes prohibitively expensive quickly. you can get 1gb links pretty cheap now but anything above that starts having huge monthly costs.

>>12164 >>12162 >Cloudflare Yep, Cloudflare is compromised the same as google is, they are extensions of the (((united states))) government and its (((intelligence community))) Cloudflare can block or make content unreachable from controversial sites Cloudflare can profile and target individual users to make them stop reaching a website if either of the site or the individuals are too controversial Cloudflare can also ban people over the whole internet by blacklisting the IP so that any captcha under their control will not let you pass Under cloudflare and other similar websites/services, every IP existing online has a reputation score, when the ip reputation gets low you start to see content loading slowly or some sites not being reachable anymore https://www.ipvoid.com/ip-blacklist-check/ There are multiple sites which create their own ip reputation lists and they all cross reference between each other Barracuda_Reputation_B JustSpam_org (cross-references from barracuda + spamhaus) zen.spamhaus.org (cross-references from cloudflare) cloudflare (cross-references from ??many?? + barracuda + justspam) >barracuda + spamhaus --> JustSpam >cloudflare --> spamhaus >barracuda + justspam --> cloudflare The complete chain of ip lists dependency is circular, spamhaus references from cloudflare, but cloudflare references from justspam which also references spamhaus, so spamhaus is both the source and the output of cloudflare: >barracuda + spamhaus --> justspam + barracuda --> cloudflare --> spamhaus --> (cloudflare) Meaning if you get a low ip reputation on barracuda then all the other ip reputation lists update to reflect that score, meaning you get a low ip reputation everywhere, and that negative score is circular so whenever it gets raised somewhere it could get readjusted to negative from the circular influence And once they decided to lower your ip score for whatever reason they ask you to give them personal identifying details if you appeal them to remove you from their blacklists, it's a scam to dox people Any site hosting free discussion should never have neither of google or cloudflare as third party or as content management

>>12050 you are only counting 8ch spinoffs and soyjak+kohl. Missing information here, like a fourth of the refugees are posting on kissu and other sites

>>12333 kissu is being monitored https://eylenburg.github.io/altchans.htm

>>12336 According to those stats all 8chan boards together currently have something like 6% of the posting activity all 4chan had right before the hack 4chan: 515,722 posts each day 8chan: 32,712 posts each day

>Soyjak PPH 1415 >8chan PPH 2208 Total sharteen death.

Only 50% of posters loss after 4chan came back could have been more, 8chan sure gained more people than it lost after this 4chan hack event I remember this site before had as top board /hisparefugio/ , at something like 140 posters per day, every other board was dead

>>13030 how much of that is 2 chair sitting, d'you reckon?

>>13030 Bunch of battered wives.

>elephant in the room i thought this site was written in js???

>>13030 >/vg/ holy fucking shit that place got destroyed

>>13205 To be fair, a good chunk of the /v/ and /vg/ stats are shit up by the faggot that was spamming thousands of posts a day to kill the usability of the /xivg/ threads on both boards.

>>13208 And that spam reached its max output few hours before 4chan reenabled posting

So….the site owner isn’t going to bother with the final solution anymore? Just going back to sleep?

>>13232 He posted an update 12 hours back in the meta thread I think, basically its already mostly happened, with some parts still happening. So maybe an extra day or two is my interpretation.

>>13232 The loading speeds have already got better and the uploading speed has also improved

>>13264 I thought the site was faster because all 4chin refugees left

>>13030 If you wanted a real measurement you should be charting things from when the hack first started, not a week in.

Website has been updated https://eylenburg.github.io/altchans.htm >After 4chan's return, its PPD recovered but did not fully rebound to pre-hack levels a week after. For example, the various video game boards together dropped from 210,075 PPD pre-hack to 165,150 PPD, /pol/ fell from 89,475 PPD to 71,475 PPD, and /biz/ even shrank to less than half of its pre-hack activity. This suggests some users either remained active on altchans or reduced their overall imageboard activity. Many altchans retained somewhat elevated activity compared to pre-hack levels, with Soyjak.st's /soy/ + /qa/ at 13,475 PPD (still above the pre-hack 10,225 PPD) and 8chan.moe's video game boards at 23,450 PPD (down from 45,900 PPD but significantly above the previous 325 PPD). Holotower.org's /hlgg/ board notably increased to 24,100 PPD, surpassing its downtime peak, indicating strong retention for Vtuber discussions (whereas 4chan's /vt/ PPD did not fully recover). > >While 4chan remains the dominant imageboard, the hack exposed its user base's willingness to explore alternatives - but for the majority of users it also showed a lack of coordination and insufficient knowledge of other imageboards I wonder what would have happened if 4chan had been offline for a month or longer - or forever.

>>13642 >1 week after 4chan's return for /a/ + /animu/ >1,200 PPD Doesn't feel like it.

i told you for years you could cut load speeds to a fraction if you made webp the format for thumbnails and banners but you soyim keep saying "le jewgle made webp so its bad!" (somehow the irony is not lost on you that the internet was made by the US DoD) so you deserve what you get with your gigantic png thumbnails. don't say i didn't warn you