HomeSearchPrevious dayNext day

00:10:32<fireonlive>he’s grown up to use a bouncer of sorts 🥲
00:11:35Arcorann (Arcorann) joins
00:35:38hitgrr8 quits [Client Quit]
00:42:55<fireonlive>https://www.fastly.com/blog/debunking-cloudflares-recent-performance-tests
00:43:00<fireonlive>fight fight fight!!!
00:43:18<fireonlive>'Lies, damned lies, and (Cloudflare) statistics: debunking Cloudflare’s recent performance tests' via https://news.ycombinator.com/item?id=36731490
00:43:40<nulldata>(December 6, 2021)
00:43:49<fireonlive>kill em
00:43:50<fireonlive>me
00:45:30<fireonlive>ok archiveteam is moving to discord
00:45:34<fireonlive>sorry JAA
00:45:51<fireonlive>i need my message delete back
00:48:14<nulldata>"Their tests compare JavaScript running on Cloudflare Workers, a mature, generally available product, with JavaScript running on Compute@Edge. Although the Compute@Edge platform is now available for all in production, support for JavaScript on Compute@Edge is a beta product. We clearly identify in our documentation that beta products are not ready
00:48:14<nulldata>for production use. A fairer test on this point would have compared Rust on Compute@Edge with JavaScript on Cloudflare Workers, which are at more comparable stages of the product lifecycle."
00:48:27<nulldata>Not a very strong argument on that point lol
00:51:17<@OrIdow6>fireonlive: Just say /s
00:51:22<@JAA>fireonlive: You just need to git gud.
00:51:34<fireonlive>i like OrIdow6's suggestion better
00:51:37<fireonlive>:p
00:51:54<fireonlive>JAA: git: 'gud' is not a git command. See 'git --help'.
00:51:55<fireonlive>?
00:52:18<@JAA>IIRC, there is in fact an extension that adds that command.
00:52:48<fireonlive>xD
00:52:53<nulldata>"Our comparable product isn't great and is in beta so they should have used a different product that might make us look better"
00:52:58<@JAA>Oh, there's a game: https://github.com/benthayer/git-gud
00:53:19<fireonlive>ahhh!
00:53:31<fireonlive>you mean i shoudln't just rm -rf gitrepo; git clone?
00:53:40<@JAA>But I think this is the one I was thinking of: https://github.com/fsufitch/git-gud
00:53:57<fireonlive>awesome
00:54:24<nulldata>Will gitgud save the repo IA just DMCA'd? :P
00:54:32<@JAA>The neat thing about Git is that you can easily extend it by just installing `git-foo` in your PATH, and it'll work as `git foo ...`.
00:55:09nicolas17 quits [Ping timeout: 265 seconds]
00:56:49<fireonlive>huh interesting
00:59:09nicolas17 joins
01:10:10sonick (sonick) joins
01:13:13<fireonlive>lol gitlab is down
01:13:37<fireonlive>"503 An internal server error occured." (oops a typo); git clone shows 502
01:13:52<fireonlive>ah now status page is orange :D https://status.gitlab.com/
01:14:34<nukke>This is their second major outage this week
01:16:54etnguyen03 quits [Ping timeout: 265 seconds]
01:17:19<@JAA>Again?
01:17:49<@JAA>They fucked up their Terraform deploy workflow the other day.
01:18:46<@JAA>Seems to work fine here.
01:20:08<pabs>nicolas17: blast from the past! I did reverse engineering on CHM back in the day, the chmspec I wrote is still online :)
01:20:57<pabs>http://chmspec.nongnu.org/latest/
01:22:37<pabs>re browser engines, the Ladybird thing seems like it will eventually be nice. completely community oriented
01:22:46<fireonlive>looks like it's serving for me now
01:23:35<pabs>nicolas17: any idea what MS' current offline-help things are? just open HTML in a browser?
01:24:36<nicolas17>pabs: they have offline-help stuff for Office and Visual Studio and maybe Windows, but they aren't available for third-party developers to use
01:24:54<nicolas17>or they are, if you're making a Visual Studio extension or something
01:25:06<nicolas17>they have nothing to replace hlp/chm
01:25:47<nicolas17>March 2001: At the WritersUA (formerly WinWriters) conference, Microsoft announces plans for a new help platform, Help 2, which is also HTML based.
01:25:47<pabs>I see, wonder what app devs do. probably custom stuff or use existing FOSS stuff
01:25:55<nicolas17>January 2003: Microsoft decides not to release Microsoft Help 2 as a general Help platform.
01:26:22<@JAA>Oh, the 2023 Heavy Metal Knitting World Championship is cancelled. :-(
01:26:36<nicolas17>https://en.wikipedia.org/wiki/Microsoft_Help_2
01:26:38<nicolas17>https://en.wikipedia.org/wiki/Microsoft_Help_Viewer
01:27:59<nicolas17>apparently macOS is in a similar situation, the "publicly available for third party devs" help system is at least in a better shape than Windows CHM, but it doesn't seem to be actively maintained/updated, and some modern apps from Apple are using a new private thing
01:28:59<pabs>FOSS seems to be settling on raw HTML/CSS, with JavaScript for search
01:29:19<pabs>I think GNOME might have a custom thing though
01:29:22etnguyen03 (etnguyen03) joins
01:29:39<nicolas17>GNOME has yelp, KDE has khelpcenter, I never got either to work well with debian doc-base
01:30:48<nicolas17>I hate that I have /usr/share/doc/python3-doc/html/library/ installed but it's still more convenient to use https://docs.python.org/3/library/
01:32:43<nicolas17>the latter appears on internet search results and is linked from pydoc
01:32:45<@JAA>man :>
01:33:14<@JAA>And GNU has info, which I hate.
01:38:48<nicolas17>IIRC some manpages in Debian used to say "in general, GNU loathes manpages and writes info pages instead" but I think that wording is gone now :)
01:43:58chessnoob280 quits [Ping timeout: 265 seconds]
01:46:33<nicolas17>ah found it
01:46:35<nicolas17>" The GNU folks, in general, abhor man pages, and create info documents instead. Unfortunately, the info document describing tar is licensed under the GFDL with invariant cover texts, which makes it impossible to include any text from that document in this man page. Most of the text in this document was automatically extracted from the usage text in the source. It may not completely describe all features of the program."
01:47:32<fireonlive>https://www.youtube.com/watch?v=mV34qVjw6Ws NSFW
02:00:50<fireonlive>[Identified] We have identified the problem and mitigated the worst of it. Our infrastructure team is continuing investigations. Details in https://gitlab.com/gitlab-com/gl-infra/production/-/issues/16042
02:01:01<fireonlive>"2023-07-14: Intermittent apdex dips for web and patroni services due to row-level lock contention"
02:04:11<@JAA>They accidentally nuked many PATs and aren't sure if they can revert it. Big oof.
02:05:04<nicolas17>...did they?
02:05:27<@JAA>> The incorrect SQL query also caused many PATs to be revoked, incorrectly. We do not yet know if we will be able to reverse this. The revoked PATs likely has downstream effects (failed API requests) for users.
02:05:46<nicolas17>I didn't see that x.x
02:08:05<@JAA>Huh, that comment has been removed again.
02:08:35<@JAA>@kslaats commented this at 01:58:
02:08:39<@JAA>> The root cause is a recently deployed change that runs an incorrect SQL query when revoking PATs. This led to a DB performance issue and poor response times, which we have mitigated by adding a DB index.
02:08:43<@JAA>> We are planning to deploy a hotpatch to "revert" (stop the new query from being called) the change. The incorrect SQL query also caused many PATs to be revoked, incorrectly. We do not yet know if we will be able to reverse this. The revoked PATs likely has downstream effects (failed API requests) for users.
02:13:08<nicolas17>well it just got worse, I'm actually getting request failures now
02:13:21<@JAA>Same
02:15:13<nicolas17>503 server unavailable
02:16:13<nicolas17>and it's back
02:16:17<nicolas17>maybe?
02:18:42<@JAA>Comment from 12 minutes ago:
02:18:43<@JAA>> Running a new pipeline now that GitLab.com is stable.
02:18:44<@JAA>:-)
02:19:27<nicolas17>it broke the deploy? lol
02:20:53<@JAA>That's from https://gitlab.com/gitlab-org/gitlab/-/merge_requests/126600 which disables the potential PAT nuking.
02:21:06<@JAA>At least based on the description
02:21:18<nicolas17>yeah I saw now
02:21:58<nicolas17>>MR to fix incident failed to merge due to same incident
02:22:39<@JAA>I have to say I enjoy seeing all of this in the open though. It's a great way to see how this kind of company works internally, at least on the technical side of things.
02:24:52<@JAA>By the way, that comment I quoted was visible for less than 4 minutes. I just happened to load the page at exactly the right time. It was deleted less than a minute after fireonlive shared the link here.
02:25:21<nicolas17>now I see no activity from kslaats in the last day
02:25:28<@JAA>Yeah, same.
02:26:05<nicolas17>someone might have shouted at them in internal chat >.>
02:26:48<@JAA>RIP
02:28:07<nicolas17>new comment in the issue
02:28:25<nicolas17>still no info about "incorrect revokes"
02:29:48<@JAA>I bet that'll come later when the immediate outages are resolved.
02:29:52<nicolas17>JAA: btw it's not just technical stuff they do "in the open" :) https://gitlab.com/gitlab-com/gl-infra/reliability/-/issues/14530
02:30:18<fireonlive>:o oh wow
02:31:03<@JAA>Ah, nice
02:32:18<fireonlive>the deletion is something lol
02:34:00<nicolas17>JAA: another example https://about.gitlab.com/handbook/hiring/greenhouse/
02:34:20<nicolas17>if they don't have a *good reason* to keep something confidential, then it's public
02:36:32<fireonlive>that’s pretty cool
02:45:50etnguyen03 quits [Ping timeout: 252 seconds]
02:47:56lunik173 joins
02:54:59Naruyoko5 joins
02:57:00Naruyoko quits [Ping timeout: 258 seconds]
03:00:12emily quits [Client Quit]
03:00:50pseudorizer (pseudorizer) joins
03:00:55etnguyen03 (etnguyen03) joins
03:03:09nicolas17 quits [Client Quit]
03:15:50<fireonlive>https://gitlab.com/gitlab-com/gl-infra/production/-/issues/16042
03:15:55<fireonlive>it's been bumped to S1 now
04:12:23<fireonlive>ChanServChan o3o
04:21:32etnguyen03 quits [Ping timeout: 265 seconds]
04:26:05etnguyen03 (etnguyen03) joins
04:28:08razul quits [Ping timeout: 252 seconds]
04:58:09etnguyen03 quits [Client Quit]
05:40:27decky joins
05:44:47jasons quits [Client Quit]
05:44:47jasons (jasonswohl) joins
05:56:19BlueMaxima joins
07:21:41hitgrr8 joins
07:22:05qwertyasdfuiopghjkl quits [Client Quit]
07:26:37qwertyasdfuiopghjkl (qwertyasdfuiopghjkl) joins
07:30:33chessnoob280 joins
08:34:55razul joins
09:17:03BlueMaxima quits [Read error: Connection reset by peer]
09:40:03chessnoob280 quits [Ping timeout: 265 seconds]
09:41:22Dango360 quits [Read error: Connection reset by peer]
09:43:29Dango360 (Dango360) joins
10:21:21<flashfire42>https://www.facebook.com/images/assets_DO_NOT_HARDCODE/company_brand_privacy_center_policy/Privacy-SafetyGuide-Hero-WWW.png lol
10:51:53Mateon2 joins
10:52:22qwertyasdfuiopghjkl quits [Remote host closed the connection]
10:52:22razul quits [Client Quit]
10:52:22Mateon1 quits [Remote host closed the connection]
10:52:22Mateon2 is now known as Mateon1
10:52:38razul joins
12:09:01chessnoob280 joins
13:19:00chessnoob280 quits [Ping timeout: 265 seconds]
13:24:40lunik173 quits [Client Quit]
13:25:05lunik173 joins
14:12:57etnguyen03 (etnguyen03) joins
14:25:09chessnoob280 joins
14:38:41<@JAA>Oh, they have a delayed replica, that's neat.
14:39:04<@JAA>> For purposes of attempting to reverse the effects of the unwanted bulk UPDATE statement, we extracted a copy of the personal_access_tokens table from the main db's delayed replica.
14:39:08<@JAA>> This is the data state as of 2023-07-14 18:12 UTC.
14:39:44nulldata quits [Ping timeout: 252 seconds]
14:44:41Arcorann quits [Ping timeout: 252 seconds]
14:46:01<@JAA>Here's a writeup of how they do that: https://about.gitlab.com/blog/2019/02/13/delayed-replication-for-disaster-recovery-with-postgresql/
14:46:39<masterX244>smooth method for a emergency ctrl+z. always good to have safety nets
14:48:35nulldata (nulldata) joins
14:58:08chessnoob280 quits [Remote host closed the connection]
15:33:38etnguyen03 quits [Ping timeout: 252 seconds]
16:22:18etnguyen03 (etnguyen03) joins
16:45:04<fireonlive>oh that’s interesting
16:52:06sonick quits [Client Quit]
16:57:57fishingforsoup_ joins
16:58:02gfhh quits [Ping timeout: 258 seconds]
17:01:30<fireonlive>speaking of governments (shreddit) https://www.ctvnews.ca/canada/mp-files-official-grievance-after-taylor-swift-announces-new-tour-dates-with-no-canadian-cities-1.6450714
17:01:52fishingforsoup__ quits [Ping timeout: 258 seconds]
17:03:50etnguyen03 quits [Ping timeout: 252 seconds]
17:05:32AmAnd0A quits [Read error: Connection reset by peer]
17:05:49AmAnd0A joins
17:09:39<fireonlive>https://transfer.archivete.am/inline/BnGdp/IMG_3756.jpeg when worlds collide
17:10:09<fireonlive>or when a huge company also uses your name
17:19:44jasons quits [Client Quit]
17:20:16jasons (jasonswohl) joins
17:21:12etnguyen03 (etnguyen03) joins
17:22:27<nulldata>https://bird.makeup/users/elonmusk/statuses/1680082007873953794
17:22:33<nulldata>Huh wonder why
17:24:53<fireonlive>lol
17:25:11<fireonlive>50% drop in ad rev is impressive
17:31:16razul quits [Client Quit]
17:32:41razul joins
17:43:26etnguyen03 quits [Ping timeout: 252 seconds]
17:50:55etnguyen03 (etnguyen03) joins
17:52:28chessnoob280 joins
18:12:14<fireonlive>10:41:37 AM <+rss> Synit – A Reactive Operating System: https://synit.org/ https://news.ycombinator.com/item?id=36735620
18:12:42<fireonlive>hmm
18:13:50etnguyen03 quits [Ping timeout: 265 seconds]
18:47:31<fireonlive>TIL there’s an alternative fandom frontend: https://breezewiki.com/
18:49:07<fireonlive>and imgur https://imgin.voidnet.tech/
19:01:10<thuban>unfortunately imgin and rimgo break on many if not most imgur items lately
19:01:20<thuban>(i haven't investigated this)
19:01:22nicolas17 joins
19:04:31<fireonlive>ah boo
19:04:46<fireonlive>unrelated to you joining, nicolas17 :p
19:22:26cm quits [Ping timeout: 252 seconds]
19:26:08cm joins
19:35:51sec^nd quits [Ping timeout: 245 seconds]
19:40:58sec^nd (second) joins
19:44:17razul quits [Client Quit]
19:45:35razul joins
20:19:36<nicolas17>fireonlive: HN moment
20:19:49<nicolas17>https://pbs.twimg.com/media/F1FVhesaEAACnus.jpg
20:20:08<nicolas17>https://media.discordapp.net/attachments/980413884961808415/1129804522655580303/Screenshot_20230715-185825_Chrome.png
20:21:26<fireonlive>xD jeeeeeesus
20:21:53<fireonlive>real name in bio too :D
20:22:01<fireonlive>but the whole concept.. just sheesh
20:24:21<@JAA>> what kind of lawyer to check with
20:24:50<@JAA>Depending on your definition of 'lawyer', I'd recommend a notary for the death certificate. :-)
20:27:45<fireonlive>:3
20:33:03<@JAA>Here's the thread, by the way: https://news.ycombinator.com/item?id=36731320
20:33:23etnguyen03 (etnguyen03) joins
20:44:39Jstar269 joins
20:47:58cdub quits [Client Quit]
20:53:49Jstar269 quits [Remote host closed the connection]
20:59:42Minkafighter52 quits [Client Quit]
20:59:57Minkafighter52 joins
21:01:41ymgve_ quits [Quit: Leaving]
21:05:25IDK_ quits [Client Quit]
21:05:25Minkafighter52 quits [Client Quit]
21:05:30IDK_ joins
21:05:35Minkafighter52 joins
21:11:08IDK_ quits [Client Quit]
21:11:08Minkafighter52 quits [Client Quit]
21:11:12IDK_ joins
21:11:22Minkafighter52 joins
21:20:48<fireonlive>AI has certainly taken new directions: https://www.youtube.com/watch?v=QRl1HAtUItU
21:20:49<fireonlive>nsfw
21:21:08hitgrr8 quits [Client Quit]
21:26:41BigBrain quits [Ping timeout: 245 seconds]
21:27:35BigBrain (bigbrain) joins
22:05:54Jake quits [Quit: Leaving for a bit!]
22:06:11Jake (Jake) joins
22:25:29BigBrain quits [Remote host closed the connection]
22:25:56BigBrain (bigbrain) joins
22:28:51luckcolors quits [Ping timeout: 258 seconds]
22:33:23luckcolors (luckcolors) joins
23:01:26tzt quits [Ping timeout: 258 seconds]
23:02:59tzt (tzt) joins
23:05:12<@JAA>> The root cause is attributed to CTE not being applied to the update_all function called in the query, used during the revocation. This in turn caused it to attempt to revoke a much greater (~4 million) number of tokens. Being a transactional operation it timed out trying to complete (& in turn was a no-op, i.e no tokens were revoked)
23:05:17<@JAA>Re GitLab last night
23:06:33<@JAA>And then row lock contention and general sadness.
23:07:08<fireonlive>ooooof
23:08:07<nicolas17>so nothing was actually revoked, because it timed out because it could do that :P
23:08:20<@JAA>Yup
23:08:22<nicolas17>I'm disappointed at the lack of info on status.gitlab.com tho
23:08:23<@JAA>Yay, timeouts!
23:08:41<fireonlive>i'm such a badass i never use begin transaction;
23:08:49<fireonlive>😎
23:09:09<fireonlive>i mean at least they didn't rm -rf db1 again
23:09:11<fireonlive>:3
23:09:21<@JAA>That's from https://gitlab.com/gitlab-com/gl-infra/production/-/issues/16047 by the way.
23:09:24<fireonlive>i'd love if github was this open though... and a lot more things
23:09:39<@JAA>Yeah
23:09:43<fireonlive>like power companies, internet copmanies, ...
23:09:49<fireonlive>at the very very least public utilities....
23:09:53<fireonlive>>:(
23:09:55<nicolas17>yeah, they're very open and public but you have to *actually go into the issue* to see the details, I think they could have been a lil more verbose on the status page
23:10:03<fireonlive>ah yeah
23:10:38BlueMaxima joins
23:10:49<nicolas17>on an unrelated note, I'm making a proxy that gets webhooks from *github* and decides whether to forward them to the real destination or not, because github itself doesn't have flexible enough filtering for events
23:11:08<nicolas17>how dumb would it be to put the proxy code on a repo on gitlab? :D
23:11:35<fireonlive>(can't filter @ destination?)
23:11:51<fireonlive>i meeeeean; it'll mostly be available on gitlab :p
23:12:50<nicolas17>the destination is https://discord.com/api/webhooks/something, so no
23:13:07<@JAA>Both ends should support that. lol
23:13:46<nicolas17>github lets me say "send a webhook when event X happens, not when event Y happens"
23:13:59<nicolas17>it doesn't let me say "send a webhook when there is a push to branch X, but not when it's to branch Y"
23:14:11<@JAA>Ah, yeah
23:16:04<nicolas17>every time I push, there is a notification saying I pushed to repo1 branch main, a notification saying a bot pushed to repo2 branch main "update submodule for repo1", and a notification saying github-actions pushed to repo1 branch gh-pages
23:16:26<fireonlive>ahh
23:21:39ymgve joins
23:22:38<nicolas17>https://github.com/smarkets/marge-bot
23:22:43<nicolas17>this is a gitlab bot hosted on github, lol
23:46:43Arcorann (Arcorann) joins

HomeSearchPrevious dayNext day