|
KARMA! posted:idiocy!! yeah no ok maybe that was a bit much im sorry internet forums poster KARMA!
|
#
?
Jul 4, 2014 17:20
|
|
|
|
| # ? Jun 11, 2024 17:26 |
|
|
Notorious b.s.d. posted:in other words, the ivy leagues are not particularly welcoming to non-trad students. no if ur gonna be there fulltime not much happens explaining why u are ~*nontraditional*~ in your app essay is actually a good way to get admitted
|
|
#
?
Jul 4, 2014 17:43
|
|
|
http://www.php.net/manual/en/datetimeimmutable.modify.php DateTimeImmutable::modify — Alters the timestamp
|
|
#
?
Jul 4, 2014 19:15
|
|
|
ya we already covered that in the PL thread and it doesn't do what you think it does and is correct.
|
|
#
?
Jul 4, 2014 21:55
|
|
|
PHP: it doesn't do what you think it does and is correct
|
|
#
?
Jul 4, 2014 21:56
|
|
|
wait this is the PL thread so idk where it was but rest assured it's somewhere
|
|
#
?
Jul 4, 2014 21:56
|
|
|
Tiny Bug Child posted:wait this is the PL thread so idk where it was but rest assured it's somewhere you were right, it is in the PL thread
|
|
#
?
Jul 4, 2014 22:07
|
|
|
tbc was right
|
|
#
?
Jul 4, 2014 22:07
|
|
|
https://medium.com/code-adventures/farewell-node-js-4ba9e7f3e52b lol
|
|
#
?
Jul 4, 2014 23:16
|
|
|
|
|
#
?
Jul 4, 2014 23:37
|
|
|
|
|
#
?
Jul 4, 2014 23:47
|
|
|
idg why you posted that guy's pic. is it because he's cuter than probably any yosposter?
|
|
#
?
Jul 4, 2014 23:57
|
|
|
Note this guy has a ton of (like over 500) node packages that people use. He doesn't maintain them and whines when nobody steps up to maintain them.
|
|
#
?
Jul 5, 2014 00:06
|
|
|
pretty sure I've taken 500 hundred dumps with more substance than his node packages
|
|
#
?
Jul 5, 2014 02:43
|
|
|
AWWNAW posted:pretty sure I've taken 500 hundred dumps with more substance than his node packages yeah probably https://www.npmjs.org/~tjholowaychuk
|
|
#
?
Jul 5, 2014 02:57
|
|
|
They're basically PHP functions ported to JavaScript wrapped as a package for Node.
|
|
#
?
Jul 5, 2014 03:08
|
|
|
akadajet posted:Note this guy has a ton of (like over 500) node packages that people use. He doesn't maintain them and whines when nobody steps up to maintain them. I wondered about that. I've quite taken to Jade, its a nice concise syntax to write HTML in, but the Scala interpretation of Jade is somehow better documented and has some nice extra language features like a little more control of whitespace collapsing and the like. Seems so weird for the library to spawn the idea just stop giving a poo poo about progress shortly after gaining traction.
|
|
#
?
Jul 5, 2014 07:56
|
|
|
Maluco Marinero posted:Seems so weird for the library to spawn the idea just stop giving a poo poo about progress shortly after gaining traction. welcome to open sores projects
|
|
#
?
Jul 5, 2014 08:15
|
|
|
Arcsech posted:welcome to open sores projects
|
|
#
?
Jul 5, 2014 10:47
|
|
|
yeah but the things poettering and hogsberg are replacing have decades-old design assumptions that make them a hindrance in the modern day systemd and wayland both own gnome 3 on the other hand though yeah that could def have been handled better. like maintaining 2.x in parallel until at least uh 3.8, and also at least pretending to give a poo poo about API stability in gtk3.
|
|
#
?
Jul 5, 2014 12:14
|
|
|
http://www.jwz.org/doc/cadt.html
|
|
#
?
Jul 5, 2014 12:36
|
|
|
Wayland and systemd aren't either of those things.
|
|
#
?
Jul 5, 2014 12:39
|
|
|
the example bugs are(were, who knows) in GNOME
|
|
#
?
Jul 5, 2014 13:06
|
|
|
Suspicious Dish posted:Wayland and systemd aren't either of those things. no one ever thinks they're part of the problem
|
|
#
?
Jul 5, 2014 13:51
|
|
|
quote:But that's what happens when there is no incentive for people to do the parts of programming that aren't fun. Fixing bugs isn't fun; going through the bug list isn't fun; but rewriting everything from scratch is fun (because "this time it will be done right", ha ha) and so that's what happens, over and over again. quote:Wayland is intended as a simpler replacement for X, easier to develop and maintain. GNOME and KDE are expected to be ported to it. quote:With Wayland we can move the X server and all its legacy technology to an optional code path. Getting to a point where the X server is a compatibility option instead of the core rendering system will take a while, but we'll never get there if don't plan for it. quote:Is Wayland network transparent / does it support remote rendering? Notorious b.s.d. fucked around with this message at 13:54 on Jul 5, 2014 |
|
#
?
Jul 5, 2014 13:52
|
|
|
Brain Candy posted:the example bugs are(were, who knows) in GNOME the last time i submitted a bug to gnome it was completely ignored for four years, then fixed by accident when they rewrote a component from scratch the system works!
|
|
#
?
Jul 5, 2014 13:56
|
|
|
Mr Dog posted:yeah but the things poettering and hogsberg are replacing have decades-old design assumptions that make them a hindrance in the modern day better rewrite from scratch! Mr Dog posted:gnome 3 on the other hand though yeah that could def have been handled better. like maintaining 2.x in parallel until at least uh 3.8, and also at least pretending to give a poo poo about API stability in gtk3. at this point the best thing would be to just wind it down entirely, because holy poo poo what is even the point gtk3 and gnome 3 have nothing like the uptake of their predecessor projects
|
|
#
?
Jul 5, 2014 13:57
|
|
|
We've talked about this before. systemd and Wayland both explicitly have compatibility with the legacy systems they're replacing as an explicit goal. There are things that we just can't do with yesterday's technology, and extending it is too difficult since the old model and old assumptions became false in the meantime since.
|
|
#
?
Jul 5, 2014 14:07
|
|
|
Notorious b.s.d. posted:oh this twenty year old codebase with millions of users is a pain in my rear end They're doing a surprisingly good job, but it isn' just Poeterring each systemd release has a very long authors list.
|
|
#
?
Jul 5, 2014 14:34
|
|
|
Suspicious Dish posted:We've talked about this before. systemd and Wayland both explicitly have compatibility with the legacy systems they're replacing as an explicit goal. of course systemd is compatible with the old stuff no one was ever stuck with init/syslogd/all the other crap it is engulfing in the first place that's how we got all these replacements the way things are going, systemd will be more like x, where we can't replace it but eventually someone will build a saner system on top that tries to mostly ignore it, and eventually people will throw up their hands in disgust and rewrite the platform entirely and put a huge emulation layer on top.
|
|
#
?
Jul 5, 2014 14:42
|
|
|
hail satand
|
|
#
?
Jul 5, 2014 15:18
|
|
|
suffix posted:of course systemd is compatible with the old stuff Sure. Nothing lasts forever, and as innovations in computers and tech continue, any model that's sane for today will eventually fall apart 30 years down the road. That doesn't mean we shouldn't even try to improve what we have today, and instead stick to stuff that's 30 years overdue now. systemd and Wayland are already large improvements over the status quo.
|
|
#
?
Jul 5, 2014 15:19
|
|
|
Suspicious Dish posted:That doesn't mean we shouldn't even try to improve what we have today, and instead stick to stuff that's 30 years overdue now. systemd and Wayland are already large improvements over the status quo. lol i upgraded to a linux distro with systemd, and it had me a bit worried because, you know, pulseaudio but the system started up fine, and everything seemed to work, except the systemctl interface was pretty bad and confusing, but well, linux so it had me in a bit of existential crisis, could it be that poettering had made something that wasn't a giant pile of poo poo? that wouldn't randomly crash and take down the whole system? but then journalctl started hanging and i noticed that journald was writing corrupted log files, and everything was right with the world but man, that was an intense 10 minutes
|
|
#
?
Jul 5, 2014 15:57
|
|
|
imagine, a log system in 2014 that corrupts he own log files "all right, let's get the specs down on this log system. what's the single most important thing it should do?" "it should use dbus!" "oookay, anything else?" "it should also be an init system" "and it should be a hard dependency for gnome 3" "ooh, i know, the config names should collide with those of the linux kernel" "and it shouldn't run on any other operating systems" "uh, guys, shouldn't we also have something about storing the user logs somewhere" "gently caress that i don't give a poo poo about your logs bitch, i'm lennart poettering!" suffix fucked around with this message at 17:31 on Jul 5, 2014 |
|
#
?
Jul 5, 2014 15:58
|
|
|
journalctl stores user logs just fine. I haven't seen log corruption at all, and my logs go back to February, when my new laptop came in. Did you file a bug at all, or can I look at your log files?
|
|
#
?
Jul 5, 2014 16:01
|
|
|
yes sure, let me just send log files from my system in an opaque format i can't inspect myself, so some scrub-level developer who thinks he's kernel people can tell me it's my sound card, or the kernel, or user error, or intended behaviour, or anything but his non-euclidean eldritch horror of an overcoupled mess of a system i just said, i remember pulseaudio but i think i deleted the log files that failed --verify anyway, so journalctl would stop hanging. i get a lot of "<hex address>: unused data (entry_offset==0)" but that's not the same i think
|
|
#
?
Jul 5, 2014 16:12
|
|
|
is there even a systemd-using distro that doesn't pipe all the journal logs to syslogd? just grep /var/log/messages like you normally do you big baby (or is it /var/log/syslog? linux is about choice you see, it's very important that i not be able to rely on things like where the gently caress the system log is located because if we standardised on that then that would be Not The Unix Way) computer systems look nothing like they did back in 1980 when X and sysvinit came about. like, there are more differences than similarities. regarding wayland, the iphone happened a few years ago and it kind of changed what normal users are willing to tolerate from a user interface (specifically, you cannot get away with tearing, stuttering and janky resizing any more like you could ten years ago). osx has been doing this right for over ten years now and x11 hasn't because of the loving Change Is Bad greybeard bridge. people who know more than you about x11 have all come to the conclusion that you can't just retrofit a vsynced composited rendering model onto x11. you can't. you just loving can't, the drawing model of your dumb as bricks dma-controller-with-bells-on 1980s-era graphics cards designed in the day when RAM was priced as if it were made of crushed unicorn horn that x11 targeted just is not compatible with a composited desktop. x11 was a good solution for that hardware, modern hardware bears no resembleance to it, x11 is not a good solution for modern hardware. literally every single other commercially successful os with a unixy kernel on it realised this decades ago but even today the bsd beard squad still have to be dragged into the 2000s kicking and screaming. and when sysvinit was designed, unix machines all had professional administrators, virtualisation was not a thing on unix machines, your security threat model was bored undergrads who wanted to play practical jokes, and nobody in their right mind would rip out half of the peripheral bus and mass storage and expect the network stack and "display server" to magically reconfigure itself. today we call that "unplugging the thunderbolt display from your laptop and going to starbucks", and sometimes even non-technical people do this and have the gall to expect it to work. x11 applications and sysvinit scripts continue to work though via the backward compatibility layers though fwiw.
|
|
#
?
Jul 5, 2014 21:09
|
|
|
suffix posted:yes sure, let me just send log files from my system in an opaque format i can't inspect myself, Yes, systemd is complex, because the problem space it's trying to solve is a large and difficult one, and bugs can happen. Please let us know when your journal files are corrupt, because it shouldn't be happening, and if it happens, we want to fix it. It could be that in some weird edge case where your /var/ is mounted through sshfs, the file locking doesn't work and journals stamp on each other. Linux administrators tend to put the system into billions of weird configurations, and while we try and support most of them, it's impossible to know about it. Don't get mad at Lennart for having the galls to try and improve the ecosystem to the point where we aren't consistently lagging 30 years behind. Keep in mind that he isn't the only one working on systemd. Lots of people, including me, contribute to systemd because we like the ideas behind it, its mostly-clean codebase, and the grand vision that it has. I can talk about very specific features that Wayland enables that X11 could never, or that systemd enables that sysvinit could never that directly lead to a better user experience for people, and aren't just an architectural purity thing.
|
|
#
?
Jul 5, 2014 21:20
|
|
|
go does look neat also lol
|
|
#
?
Jul 5, 2014 21:23
|
|
|
|
| # ? Jun 11, 2024 17:26 |
|
|
Soricidus posted:nah, there's plenty of open source stuff that's maintained pretty well. this culture of constantly rewriting everything from scratch seems to be a feature of the javascript world. well, and linux too since the advent of lennart poettering I guess quoting 1969 hamming again quote:Indeed, one of my major complaints about the computer field is that whereas Newton could say, "If I have seen a little farther than others, it is because I have stood on the shoulders of giants," I am forced to say, "Today we stand on each other's feet." Perhaps the central problem we face in all of computer science is how we are to get to the situation where we build on top of the work of others rather than redoing so much of it in a trivially different way. Science is supposed to be cumulative, not almost endless duplication of the same kind of things.
|
|
#
?
Jul 5, 2014 21:32
|
|
