Notices by Glitch (glitch@pl.glitch.pm)
-
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 05-Apr-2024 17:35:08 JST Glitch @Containment @lanodan @hj western artists seem to perpetually suffer from the issue of congregating on crappy Venture Capital social media (which their content would be a shite fit for to begin with since those sites recompress the hell out of it) and then complaining when VC starts to do the rugpull. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 05-Apr-2024 17:26:20 JST Glitch @Containment @lanodan @hj I dunno, it's more usable than it's english equivalent (deviantart).
Pixiv is okay insofar as primary sources for art go, but it's awful as an archive. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 22-Mar-2024 00:53:42 JST Glitch @lanodan Yeah. In most other cases of hostile relicensing by companies (a lot of GPL products... It's rare to see this with a BSD license), they usually do an extensive amount of rounding up all relevant contributors first to try and get the CLA signed/rewrite the patches of those that are unreachable before pulling this trick.
It's normally a process that takes on the order of 2 to 3 years, not one git commit. Even if they'd prioritize SSPL for all new contribs, they'd need approval to remove the old license from the code. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 22-Mar-2024 00:47:28 JST Glitch @lanodan Wouldn't that be infringement, as far as I can tell they're stripping out the old BSD text entirely.
Which as far as I understand it, they can't do unless they have all contributors with significant patches in the current codebase agreeing to it. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Tuesday, 19-Mar-2024 06:39:41 JST Glitch @Moon @phoenix danke.
Very cool, might see if I can build something with it. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Monday, 04-Mar-2024 22:05:55 JST Glitch @wolf480pl is this the rust version of "regex DDoS attack in nodejs library"? -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 16-Feb-2024 06:04:40 JST Glitch @Johann150 @lanodan @arcanicanis I think a lot of it comes down to the fact that it's really difficult to "predesign" a standard before you have a fully working product too.
Even putting aside the politics that always rear their ugly head when it comes to defining standards, I can speak from experience that one thing no standard (that isn't a frankenstein backport of implementations) survives is contact with reality. You're always going to find some obscure edgecase or issue you weren't aware of during standard design.
A lot of the really old web standards have clear roots back when it was only universities (so much smaller and easier to communicate these things), and even there you can find issues with how there were exciting new issues when a general audience got involved. (See: the complete disaster of handling email spam, the failure of domain WHOIS records until literal laws had to step in to fix things.) -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Sunday, 21-Jan-2024 06:53:35 JST Glitch @a1ba I love how top right is basically just a list of tumblr sexymen. ;-; -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Thursday, 11-Jan-2024 01:51:22 JST Glitch @lanodan @mia windows' DNS resolution is integrated with the network configuration panel if memory serves me right (which means it dumps things in the registry)
And yeah, it's specifically because the DHCP client mangles resolv.conf that some distros sub in something like Dnsmasq - it's easier to just override it with 127.0.0.1 in all applications and then redirect it to what's needed than trying to untangle which part of the stack set the DNS *this time*. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Thursday, 11-Jan-2024 01:45:35 JST Glitch @lanodan @mia iirc on some distros this is the default moreso because DNS resolution on Linux isn't determined by what network you're connecting to (like it is on Windows/Mac), but is instead just handled by a simple resolv.conf file.
If you're not using systemd (which messes with resolv.conf via NetworkManager) its usually easier to shoe in dnsmasq and rely on it instead for something more advanced than "set an IP and forget". -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Monday, 18-Dec-2023 05:04:34 JST Glitch @lanodan @cwebber from the top of my head, undo/redo works with moving and renaming files. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Sunday, 05-Nov-2023 06:22:53 JST Glitch @lanodan @Moon @moth_ball for where it counts, XP source code was leaked a couple years ago I think?
the problem really is just that it's fucking radioactive unless Microsoft opens up the license. (which they might not be legally able to do, at least if you want a "functional" source code release rather than a source dump, since XP shipped with drivers on the install disc.) -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Tuesday, 10-Oct-2023 06:41:19 JST Glitch @lanodan @solidsanek TBH I suspect most of the tower defense genre moved over to mobile games after Plants vs. Zombies was a big mobile runaway success. A lot of random flash games that seemingly vanished ended up on phones. (Infinite runners, dress-up simulators, that kinda jazz.)
Maybe there's some gems in there but it's gonna be riddled with the mobile MTX bullshit.
(Don't let this discourage you from gamedev tho or anything, just speculating what happened to it.) -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Tuesday, 10-Oct-2023 06:37:35 JST Glitch @lanodan @solidsanek yeah, Rogue Legacy is imo the worst one since it intentionally invokes the name of the original Rogue and I think it's also the one that formalized the "modern" read of the genre.
Like, if we're talking about "credible" indie games that went slightly beyond the usual nerd audience, you had Spelunky in 2008 (which does still do the hardcore permadeath thing), Isaac in 2011 and then Rogue Legacy in 2012. From that point onwards, the definition of roguelike kinda lost all meaning.
I guess if I were a bit liberal in my interpretations, I'd count Noita as a roguelike? It hits almost all the marks of the "berlin interpretation", but that's basically the only one in recent memory that hits the marks while being popular (and Spelunky 2 I suppose).
Vampire Survivors gets called a roguelike and that game is the definition of an arcade title. People just started conflating the definition of "arcade" with "roguelike" over time I suppose. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Tuesday, 10-Oct-2023 06:18:31 JST Glitch @lanodan @solidsanek tbh that definition was killed off in the public eye thanks to the double whammy of Isaac and Rogue Legacy.
Nowadays the only unifying factor seems to be "permadeath" (in big scare quotes because the amount of roguelikes that do true permadeath is also zero these days, they almost all do some form of metaprogression). -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 11-Aug-2023 00:59:07 JST Glitch @lanodan @feld @phnt if memory serves me right the only real shortcomings right now are things like "no gists/snippets" (got stalled in bikeshedding) and things that I suspect are somewhat beyond giteas scope (ie. Pages).
The rest more or less works as expected although it's UI really apes a lot from GitHub (not necessarily a bad thing.) -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Friday, 11-Aug-2023 00:55:05 JST Glitch @lanodan @feld @phnt No idea about scaling but gitea has always worked very well for me at least.
Gitlab always feels like 25% of a forge if you don't have their enterprise plan. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Tuesday, 01-Aug-2023 20:32:19 JST Glitch @lanodan @emacsen @xocolatl ORMs are great as long as you're just doing basic CRUD operations. Thankfully, that's 99% of the operations you'll ever need to do.
For the 1% of really complicated queries, that's when SQL will always be better because of how purpose build it is to return tables/do complex updates and deletes. You can do them in ORMs but it starts turning into a contortionist exercise at that point because even the best ORMs aren't made with those in mind. (Also for stuff that needs really optimized queries - ORMs may not choose the most optimal retrieval methods for a complex query.)
And *even then* it's all being bolted on top of some pretty awful design choices that should be changed cuz we've learned more about how to properly design a language since the 90s. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Monday, 17-Jul-2023 23:08:55 JST Glitch @Suiseiseki when the layperson talks about the three major desktop OSes it's gonna be windows, macOS and Linux (or just the major server OS). That's what the shitfit was about. Something heavily relying on GNU tools at the time got popular, it wasn't named GNU so the GNU project insisted it had to be prefixed to the name so they would seem more relevant.
No OS since the early-to-mid 2000s has been just the coreutils anyway. Most practical desktop OSes have been mostly their DEs like KDE and Gnome. Calling it KDE/Linux is just as appropriate as GNU/Linux, yet the KDE people didn't mobilize an army of reply guys to tell you why it's GNU/Linux. -
Embed this notice
Glitch (glitch@pl.glitch.pm)'s status on Monday, 17-Jul-2023 06:15:22 JST Glitch @lanodan once you get past a certain sample size the amount of ways people refer to one thing will always include *many* incorrect references, especially when it comes to things like separators, dividers and similar such (it's the main reason I dropped leetspeak from my handle many years ago - there were like 5 ways to refer to me and it irritated the shit outta me).
I think one reason why GNU/thing is so prevalent however is in part because of how insistent the FSFs shitfit was over Linux being named that way, which led to a lot of people doing it to all their projects because their takeaway is that that is how GNU would prefer to be credited.