|
God of Mischief posted:Forking the language is a fine solution if the issue is important enough or internal to his company. It is not a fine solution for a minor issue that he wants fixed because it affects other major tools (git/github). Then those tools will use his fork, or it isn't as big an issue as he thinks it is.
|
#
?
May 4, 2014 22:03
|
|
|
|
| # ? May 31, 2024 23:00 |
|
|
code:
|
|
#
?
May 5, 2014 00:12
|
|
|
tef posted:matz on diversity "let em eat rails girls" I do not think matz is saying what you are implying he is saying. It seems he is addressing the "fairness" in how the talks were chosen. That he did not want rubyconf to alter the way they were "blind" picking their speakers. And I think he was trying to say that if too few women were picked via rubyconf, at least railsgirls would address that shortage. Now whether or not you think the blind picking was fair, as in was there encouragement of people who weren't white and male is a fair argument to pick on. But I don't think matz was trying to argue women should just be relegated to railsgirls to give talks. See this tweet in followup: https://twitter.com/yukihiro_matz/status/380481952941211648 matz has a pretty good grasp of English, but I don't think 1) that he has a very nuanced command of English and 2) because of that does not choose his words to make sure they can't be misconstrued.
|
|
#
?
May 5, 2014 01:28
|
|
|
fritz posted:The first time I can remember hearing about ruby was in this context: http://geekfeminism.wikia.com/wiki/CouchDB_talk Ruby had been around for 16 years when this happened, so I think this says more about you and virtually nothing about Ruby.
|
|
#
?
May 5, 2014 19:15
|
|
|
Ruby was little known before 2004.
|
|
#
?
May 5, 2014 19:18
|
|
|
Sinestro posted:
Well I would also have "flojera" if I faced the same scenario. The comment was probably added by some poor mantainer who just puked seeing the code. If this is not the case, the guy deserves to be thrown out the closest window.  One of the first pieces of code:
|
|
#
?
May 5, 2014 20:53
|
|
|
Strong Sauce posted:I do not think matz is saying what you are implying he is saying. Why would I read any of this when I can feel outraged & superior instead??
|
|
#
?
May 5, 2014 23:17
|
|
|
Smugdog Millionaire posted:Why would I read any of this when I can feel outraged & superior instead?? Truly, you have embraced the Ruby philosophy.
|
|
#
?
May 6, 2014 15:44
|
|
|
Novo posted:Ruby had been around for 16 years when this happened, so I think this says more about you and virtually nothing about Ruby. I was in Old Industry until a few years ago so I wasn't necessarily hip on all the fashionable things going on out in webland. The ruby programmers I've met since then haven't exactly changed my opinion a whole lot.
|
|
#
?
May 6, 2014 17:25
|
|
|
Smugdog Millionaire posted:Why would I read any of this when I can feel outraged & superior instead?? Although we can interpret the first bit in a number of air quotes, the followup is less damning, but is as easily interpreted one way or the other. it was provided as an example of matz not being so nice.
|
|
#
?
May 7, 2014 00:44
|
|
|
..btt posted:Truly, you have embraced the Ruby philosophy. Is this, we should be judged as if we are matz but act as if we are dhh?
|
|
#
?
May 7, 2014 00:46
|
|
|
Judge not, that ye be not dhh
|
|
#
?
May 7, 2014 22:48
|
|
|
Coworker is taking a data structures class at DeVry and was tasked with taking a deck of cards and cutting it at an arbitrary point. This is like day one of class, so the only structure they're working with is an array of size 52. I'm guessing they just want to see that they can index/loop over arrays correctly, not go out of bounds, etc. Their instructor gave them a sample implementation that looked fairly terrifying to me: they had a separate function that could be used to rotate each card forward one spot in the array (with the front card moving to the back) so we're already looking at a Θ(n) operation. Then they looped that operation based on the cut point, so I think we're looking at O(n2) — just to cut a drat deck of cards!
|
|
#
?
May 8, 2014 05:46
|
|
|
Maybe it's being used as a precursor to teaching them bubble sort. They will probably revisit the algorithm later to explain why it is bad. Although, I doubt it, since it's a data structures course, and not an algorithms course.
|
|
#
?
May 8, 2014 05:52
|
|
|
Based on the stuff I've seen him working on for his other classes: I think you're giving DeVry curriculum too much credit. Also I feel like any data structures course worth its salt needs to spend time discussing algorithm complexity.
|
|
#
?
May 8, 2014 05:53
|
|
|
Star War Sex Parrot posted:Coworker is taking a data structures class at DeVry and was tasked with taking a deck of cards and cutting it at an arbitrary point. This is like day one of class, so the only structure they're working with is an array of size 52. I'm guessing they just want to see that they can index/loop over arrays correctly, not go out of bounds, etc. This sounds very much like how I learned calc in high school. "Mathematicians are lazy, we always have a shortcut. We're gonna do this once the hard way, though, so you know why we have the shortcuts." It seems like a pretty good jumping off point for introducing structures that do it more effectively.
|
|
#
?
May 8, 2014 12:01
|
|
|
It might not even be other data structures. The trick to doing that problem in constant space is to realise you can make a rotation from three reflections.
|
|
#
?
May 8, 2014 12:11
|
|
|
coffeetable posted:It might not even be other data structures. It's a data structures class, not algorithms.
|
|
#
?
May 8, 2014 22:18
|
|
|
Maybe it's to contrast between an array and a linked list. It would be a reasonable way to do it with a linked list.
|
|
#
?
May 8, 2014 23:24
|
|
|
NinjaDebugger posted:It's a data structures class, not algorithms. It's not like data structures is a separate subject from algorithms.
|
|
#
?
May 9, 2014 00:11
|
|
|
Sebbe posted:"Who the f*heck is Tadayoshi Funaba and why can he reject sensible patches unilaterally?" That reminds me of the fun I had with Behdad Esfahbod of Gnome when they changed VTE to send random amounts of keypresses on the mouse scroll wheel, based on trying to guess what the program using it would do with uparrow. https://bugzilla.gnome.org/show_bug.cgi?id=518405 Now, imagine what a random number of uparrow/downarrow keys do to incredibly uncommon things to be running in a terminal window, such as bash. 6 years then wrongly closed as a duplicate of an unrelated bug. I still maintain the out-of-tree patch that squashes the broken behavior because every loving terminal emulator aside from stock xterm uses libvte. Edit: I feel like I may have linked this one here before, so I'll share an old horror story involving code. Years ago I worked at a company that bought a biometrics company based on their "unique algorithm." Prior to purchase I was asked to validate that their code in fact worked - it did, although my builds didn't work as well as theirs did for some reason. Since I was new, my worries were glossed over by the incoming "head of biometrics" (A third party, not part of the company being bought) who had a weasel-explanation for it involving magic buildservers that would get brought over. Fast forward a year, we've been fighting to get this stuff to work (but the binary DLLs do work) and I finally get my export-controlled copy of the NIST biometric matching algorithms... and boy do they look familiar. The code I got from the now-purchased company was exactly the same - except with all the copyright stripped off and replaced. With that revelation came another - the "magic build server" that never showed up was, shockingly, imaginary. The working binary-libraries were actually a competitor's code with the licensing cracked. And the kicker, as everything went into liquidation? The head of biometrics wasn't actually a third party, he had setup the the deception in the first place. He even got out with the good monitors before the building was locked down. Good times. Harik fucked around with this message at 01:25 on May 9, 2014 |
|
#
?
May 9, 2014 00:59
|
|
|
The answer is that terminals are non-standard and all terrible. We're emulating a piece of hardware made over 40 years ago that doesn't have a mouse, let alone one with a scrollwheel. It's non-standard what PgUp / PgDown will do, and same with Up / Down. I think vte is doing as good a job as it can, here.
|
|
#
?
May 9, 2014 01:25
|
|
|
Suspicious Dish posted:The answer is that terminals are non-standard and all terrible. We're emulating a piece of hardware made over 40 years ago that doesn't have a mouse, let alone one with a scrollwheel. Yeah, that's the argument, but it's not very good - look through the use cases and there's none in which multiple arrow-keys are better than a single pageup/pagedown key, and in many they're significantly wrong. If a program implements scrolling at all, both will work, and the only "advantage" is that the terminal dictates how far it scrolls. In programs that don't do scrolling, but do have commandlines (I.E. shells) arrow keys are absolutely the wrong thing to use, and pageup/pagedown are correctly ignored. On a text-based list (email/usenet/whatever) scrolling and navigation are different concepts, and using the scrollwheel to move the selection is completely foreign to every other user-interface. I think the only time it has the appropriate behavior is in a pager like less. That's not a strong argument. You're right in that terminals are awful - I deal with them constantly. I just went with what felt like the least-worst solution, and Behdad went with one that had the benefit of being pedantically correct in the single use-case that could achieve partial-page based scroll. I'm also a fan of "Buy new hardware". Middle-mouse-paste is already an abomination, way to make it 100x more awful guys. Harik fucked around with this message at 01:34 on May 9, 2014 |
|
#
?
May 9, 2014 01:31
|
|
|
Middle mouse paste is something every window manager should aspire to enable.
|
|
#
?
May 9, 2014 01:42
|
|
|
JawnV6 posted:Middle mouse paste is something every window manager should aspire to enable. Window managers have nothing to do with it.
|
|
#
?
May 9, 2014 01:48
|
|
|
Ok.
|
|
#
?
May 9, 2014 01:56
|
|
|
My terminal just scrolls its contents on buttons 4 and 5. No matter what is running in them.
|
|
#
?
May 9, 2014 01:58
|
|
|
Zombywuf posted:My terminal just scrolls its contents on buttons 4 and 5. No matter what is running in them. Try screen. The code in question is active when "alternate screen" mode is enabled.
|
|
#
?
May 9, 2014 02:02
|
|
|
Harik posted:Try screen. The code in question is active when "alternate screen" mode is enabled. Well I haven't got around to installing the latest version of screen that actually supports rxvt-unicode-256color, but when I gently caress with $TERM to make screen work it still scrolls the scrollback, which has interesting results in screen.
|
|
#
?
May 9, 2014 02:17
|
|
|
Zombywuf posted:Well I haven't got around to installing the latest version of screen that actually supports rxvt-unicode-256color, but when I gently caress with $TERM to make screen work it still scrolls the scrollback, which has interesting results in screen. Yeah it hasn't kicked into alternate mode until it needs to. I think jumping into screen history and scrolling around there a bit will do it, or run something like vim inside screen? I used to know exactly where in my workflow everything went wrong but I've long since forotten. I got bit by it 6 years ago, made a patch, then grumble and build-from-source every time my distro updates VTE ever since. Oh, rxvt? Looks like they rolled their own, it doesn't depend on gnome vte. I may switch but all terminals are awful and better the devil you know...
|
|
#
?
May 9, 2014 02:25
|
|
|
Harik posted:Oh, rxvt? Looks like they rolled their own, it doesn't depend on gnome vte. I may switch but all terminals are awful and better the devil you know... rxvt has support for searching the scrollback :-) Go on, you know you want to. I'm pretty sure it pre-dates Gnome Terminal by a long way.
|
|
#
?
May 9, 2014 03:13
|
|
|
I'm always sort of amused by the fact that OS X has better terminal emulators than Linux.
|
|
#
?
May 9, 2014 04:23
|
|
|
Plorkyeran posted:I'm always sort of amused by the fact that OS X has better terminal emulators than Linux. Speaking of coding horrors: if you restart your computer with Terminal.app open, it'll open back up after the reboot with the windows in the same place, same size, with the contents of the terminal window as you left it... Except the contents are an illusion and you're dumped in a fresh terminal. Doesn't restore your pwd, history, anything.
|
|
#
?
May 9, 2014 04:42
|
|
|
Coding horror: using Terminal.app
|
|
#
?
May 9, 2014 04:48
|
|
|
It does preserve the directory you were in, but the other stuff kinda sucks.
|
|
#
?
May 9, 2014 04:48
|
|
|
Plorkyeran posted:I'm always sort of amused by the fact that OS X has better terminal emulators than Linux. Because everyone came from years of good terminal emulators on unix to the godawful early Terminal.app, so they wrote their own. Lots of them! A few good ones stuck around. I mostly kid, but I did spend nearly a year on OSX and just couldn't get used to it, ended up going back to Linux. Anyway, I tried rxvt-unicode but it can't handle (╯°□°）╯︵ ┻━┻ properly, and as an IRC native that just wouldn't do. Also the font-support is weird, if I specified multiple fonts it put them in huge boxes: code:
|
|
#
?
May 9, 2014 04:49
|
|
|
Terminal emulators are always going to suck, no matter which ones they are. The VT103 state machine is a terrible mess, and apps abuse the hell out of it in weird ways. The majority of apps only work because some terminal emulator has X weird feature. There's no standard for proper Unicode support in terminals. The "grid of characters cells" approach to terminal layout is just incompatible with Unicode, actually.
|
|
#
?
May 9, 2014 05:01
|
|
|
The grid of character cells isn't the problem, its the pervasive (wrong) assumption that 1 byte = 1 code point = 1 glyph.
|
|
#
?
May 9, 2014 05:07
|
|
|
http://cpprocks.com/what-if-c-looked-more-like-python-or-coffeescript/ As a person who likes to write C/C++, Python, and Coffeescript, I am deeply offended.
|
|
#
?
May 9, 2014 05:31
|
|
|
|
| # ? May 31, 2024 23:00 |
|
|
I'm pretty offended when people say "minimal syntax" but mean "an ambiguous and ad hoc grammar that doesn't use curly braces or semicolons to appease my OCD".
|
|
#
?
May 9, 2014 05:34
|
|