Screwtape |
Posted on 19-03-20, 08:51
|
Full mod
Post: #169 of 443 Since: 10-30-18 Last post: 1101 days Last view: 172 days |
Sure, but I think 2016 is when people really started to think there might be a systemic problem and not just some unfortunate coincidences. Anyway, this isn't the politics thread! Today Firefox 66 was released, which means Firefox Nightly got bumped to version 68. For the first time in as long as I can remember, I actually found some regressions and filed bugs for them: - browser.display.use_document_fonts=0 is not respected - Tree Style Tab requires an extra click to do anything Much to my (pleasant) surprise, both bugs got triaged, confirmed, and somebody found the exact commit that caused each regression within a couple of hours of my reporting them. I'm still a bit annoyed that the regressions happened, but I'm pleased to get a response so quickly, and now I'm keen to see how long it takes the regressions to be fixed. The ending of the words is ALMSIVI. |
CaptainJistuce |
Posted on 19-03-20, 09:48
|
Custom title here
Post: #341 of 1164 Since: 10-30-18 Last post: 63 days Last view: 11 hours |
Posted by Screwtape When people started NOTICING there was a problem is a different thing.
Every thread is the politics thread. Today Firefox 66 was released, which means Firefox Nightly got bumped to version 68. For the first time in as long as I can remember, I actually found some regressions and filed bugs for them: "WONTFIX Behavior as designed." I have faith in Team Firefox. --- In UTF-16, where available. --- |
Duck Penis |
Posted on 19-03-20, 23:32
|
Stirrer of Shit
Post: #104 of 717 Since: 01-26-19 Last post: 1763 days Last view: 1761 days |
Posted by Screwtape Fair enough. But it's up to them. For instance, they might prioritize getting unwanted elements to self-select out. This is my personal opinion, but I'd rather have a small project that deals with critical bugs and adds feature at a slow but steady rate, than a big project filled with incompetent and antisocial elements with a development velocity the speed (and quality) of cancer. More LOC doesn't mean more good. I have a theory that because software developers work so intimately with computers, we fall out of practice at identifying and processing emotions in ourselves and others (or maybe the kind of people who have problems with that are naturally attracted to software development). And so, like people who are hard-of-hearing yelling at each other indoors, we get super-emotional because we don't remember what volume-level regular people use, and we wish regular people would speak up and quit mumbling so we can observe their emotional state without having to guess wildly. And like people who are hard-of-hearing, even though we're behaving rationally and haven't done anything wrong, it's more pragmatic to learn to control our volume level and learn to lip-read than it is to try and make the rest of the world conform to our expectations. No, I think Occam's razor goes here. When you're among friends, of course you would use informal and even vulgar language, but at work of course you would not. There's just a disconnect - on one hand, you have software developers coming in from the wider Internet, where communication generally is quite informal and should be taken with a grain of salt. On the other, you have those coming in from the corporate environment, who expect everything to be handled completely formally. This is a bit of a simplification, but I think there occurs a cultural clash between these two groups. For instance, if someone at work would tell me to "go fucking kill yourself you dumb nigger," I would be shocked, but if someone were to use similar language in private, well then it's just obvious they neither believe I should commit suicide nor that I am African-American, they're just taking the piss. So when someone acts what you might perceive as 'aggressively' online, it's just because they have different expectations of the level of formality than you do. Possibly, those kinds of people you describe would be those same people who want to add code of conducts to everything. Not because of any deep political convictions, but just because if they follow The Rules™ then nothing can go wrong. There was a similar slashdot post on this subject, don't agree with it 100%, but it's still somewhat interesting: https://developers.slashdot.org/comments.pl?sid=10924699&pid=54904703
Well, which one is it? It seems to me as if you start with the conclusion (TPB is bad) and then work backwards to the facts. Why would the Central Intelligence Agency run a file-sharing website? How much does a sporadic DNS block really hurt the site? What obscure content isn't available? I've heard this last one time and time again, and it's always from a private tracker enthusiast. It bugs me, because I really hate private trackers, and also because as far as I can tell it's wrong. The issue with public trackers isn't bandwidth, it's retention. Most BitTorrent clients have UIs made to encourage people to stop seeding old torrents, while eMule and DC++ encouraged people to keep seeding their whole downloads folder. And it's true that for private trackers, the retention is generally better. But no amount of retention can compensate for the sheer amount of quantity that a public tracker would give you, and with proper indexers the search is much better for these purposes. Also, for non-obscure content (and I use that term liberally), public trackers have good speeds. No, private trackers are a harm to society. They incentivize people to rent seedboxes and provide large amounts of bandwidth, but the big issue is always storage. Just think - how many torrents have you tried to download where the issue was no seeders, and how many where the download was too slow? If private trackers wouldn't exist, the people who use them today would be using private trackers. Wild guess here, but those people are file sharing enthusiasts who would keep seeding anyway. The retention would be far better, since private trackers are just a massive duplication of effort. Of course, all of this could be solved if someone were to develop a BitTorrent client that was smart about files and handled cross-seeding and automatic source discovery (from a custom DHT) so that whole folders could be seeded without prior coordination like in the deterministic file hashing of DC++ or emule. But they won't, because things are in that crucial swamp of "good enough," where it's not atrocious enough to warrant immediate action and it's not good enough to motivate people into contributing out of enthusiasm for good technology.
Well, what would they do? They put the owners in prison, stole their domain, used diplomatic back-channels to bludgeon a free and independent country into submission under threat of trade sanctions, bribed another one into extraditing a resident despite not having such an extradition treaty (price tag: $60mil). What more is there to do? Steal more domains? Clearly, "do nothing and wait until almost nobody has a computer and thus is forced to use Netflix" is easier. I'm typing this right now on a computer with a Free operating system, Free drivers and mostly Free firmware, and I didn't have to ask anyone's permission, or pay $99 a year for a developer kit. What did the FSF have to do with this? As far as I can see, all they've done is to take credit for the actions of others. Linux, X11, Firefox, apt, none of these are GNU projects. As far as I can tell, the GNU project has made three substantial contributions to free software: * A C library * A compiler * A set of coreutils All of these are today technologically irrelevant: Musl is arguably superior to glibc, except for glibc having some bugs that programs designed around. LLVM is superior to GCC in benchmarks, I'm not qualified to comment on the matter any further. The GNU coreutils are not much better or worse than any other implementation of coreutils, but there's no reason to change them either. It wouldn't be too hard to create a Linux distribution that didn't use GNU, but there wouldn't be much reason to do it except for sheer spite. It turns out there already exists such a distribution, Alpine Linux. It still appears to use GCC, but that's no a part of the operating system. The only reason they still are relevant is due to inertia, like how the Linux kernel is too tightly bound to gcc now to ever compile under Clang, even though they would want to. And back in the day, you had BSD. As far as I can tell, that was free, just not Free Enough™. My point being, the Free Software Foundation/GNU has spent far more time attempting to take credit for other people's efforts, wasting money, and generally being obstructive than they have actually developing anything of value.
No, the move towards open-souce is just sleight of hand, to hide the real Microsoft stratagem:
As the innovation slows down on the desktop front, this gives Linux the opportunity to catch up to Microsoft on the desktop, especially with the backing of Valve. Microsoft of course would like to prevent this in order to gain a monopoly position in the market (which legally wouldn't be a monopoly due to Apple) for Windows 10, because there is no other way to make money - you can't sell the same product over and over again, you can't get end users to pay a monthly fee in significant numbers, and OEM money only goes so far. We've seen Embrace, and I think we'll see Extend real soon with the Windows Subsystem for Linux. (I'll put the rest of my reply in Politics!, since it doesn't have the slight connection to the thread topic that yours does.) There was a certain photograph about which you had a hallucination. You believed that you had actually held it in your hands. It was a photograph something like this. |
nyanpasu64 |
Posted on 19-03-21, 05:17 (revision 1)
|
Post: #28 of 77
Since: 10-31-18 Last post: 1189 days Last view: 1116 days |
I'd argue that rude maintainers are "unwanted elements" and "antisocial" (but not "incompetent"), which tend to create a hostile environment that drives off other people, regardless of skill level. |
Duck Penis |
Posted on 19-03-21, 16:21
|
Stirrer of Shit
Post: #107 of 717 Since: 01-26-19 Last post: 1763 days Last view: 1761 days |
I'm not entirely sure. It's true that a hostile environment drives off some people, but I'm not certain it's completely unrelated to skill level. Linus' famous rants didn't seem to hurt the project, for instance. There was a certain photograph about which you had a hallucination. You believed that you had actually held it in your hands. It was a photograph something like this. |
BearOso |
Posted on 19-03-21, 19:34 (revision 1)
|
Post: #59 of 175 Since: 10-30-18 Last post: 1450 days Last view: 1450 days |
Posted by sureanem You don't need any of that extra stuff if all you do is browse foot-dermatophagia mailing lists with emacs all day. A serious anecdote, though: Recently I was trying to update Snes9x GTK's translation support that had broken for some reason. I ended up just upgrading the autotools stuff for 1.58, but I was disappointed by the mess of files it produced. All these mandatory "GNU" files are strewn over the place. "Fill in the blank," templates would say, "your project is called GNU ____". Try to run make without some of these literally pointless GNU files and you get errors. So for the next version I set out to replace autotools. Warning: if you expouse this choice publically, you'll get a lot of backlash from seeming autotools-lovers. Ignore them, because they're rote-methoders that barely figured out "./configure; make; make install", and they're afraid it'll change. I started looking at CMake, which is quite popular, and Meson, which looked simple from when I was playing with Mesa git. I looked at Meson first, and a couple hours later I realized I was pretty much done. I wrote nearly all the build script in one go, and it ran and worked the first time. That's how much easier it was than autotools. No more ABOUT-NLS advertisement for GNU, no more config.* files or M4 macros or sed scripts littering the directory. I may eventually get into CMake because of its ubiquity, though. I'm not entirely sure. It's true that a hostile environment drives off some people, but I'm not certain it's completely unrelated to skill level. Linus' famous rants didn't seem to hurt the project, for instance. Linus's famous rants are not unjustified or rude, but angry. Some really bad code made it to him despite all the layers of maintainers, and the retort needs to travel all the way up the totem pole for it to have any effect. This sort of thing is common in any real workplace, but the Internet echo chamber amplifies and remembers it. If you're close enough to Linus that he'd accept your code and commit it directly to his tree, where it will affect millions of people, you need to take responsibility for your actions and be more careful. *edit* The only reason they still are relevant is due to inertia, like how the Linux kernel is too tightly bound to gcc now to ever compile under Clang, even though they would want to. It mostly compiles with clang, just a couple hiccups like asm-goto. Google Pixel devices all have the kernel built with clang. LLVM/Clang are quite good: excellent performance, clangd, clang-format, debugging tools, etc. They benefit from extra hindsight, having been started way more recently than the GNU toolchain. |
nyanpasu64 |
Posted on 19-03-22, 07:42
|
Post: #29 of 77
Since: 10-31-18 Last post: 1189 days Last view: 1116 days |
Interesting note: The pret Pokemon decompilation project/community primarily operates in a Discord server. (there's a meme about it "not being a community") One member (don't recall who) has obtained an ancient gcc fork named agbcc, used by Nintendo to compile pokemon games. Since gcc is GPL, after obtaining the agbcc binary, they could tell the company who developed it to release the source code: https://github.com/pret/agbcc I assume that proprietary compilers developed off llvm/clang are neither free to distribute, nor source-available, and anyone trying to distribute them could be C&D'd by companies. In a way, GPL licensing of gcc has actually meaningfully increased public freedom, and reduced the ability of corporations to stifle this particular community. (The actual decompiled source code repos are probably vulnerable to C&D though.) Maybe it's somewhat unfortunate that Google is trying to kill gcc for Android. (I've had a difficult time setting up Eclipse with GCC for TI Launchpad embedded processors. Also somebody offhand said that "nobody uses IDEs for embedded, everyone uses makefiles". Is that true?) |
BearOso |
Posted on 19-03-22, 15:35
|
Post: #61 of 175 Since: 10-30-18 Last post: 1450 days Last view: 1450 days |
Posted by jimbo1qaz obtaining :: stealing. If it was for private use, they have no obligation to provide the source code, especially if the third-party obtains it illegally. I'm sure a judge would take that perspective as well. Say I use a GPL recipe for brownies, only intending for me to eat them. You break into my house and steal those brownies. Do I have to tell you how they're made? Nope. Even if the GPL intends that, it's not going to hold up. Now, should Nintendo release the code in the spirit of good will and greater knowledge? Definitely. |
Kakashi |
Posted on 19-03-22, 15:42
|
Post: #105 of 210 Since: 10-29-18 Last post: 1876 days Last view: 1848 days |
God damn you, BearOso! I want dem brownies! |
tomman |
Posted on 19-03-22, 18:26
|
Dinosaur
Post: #218 of 1316 Since: 10-30-18 Last post: 32 sec. Last view: 32 sec. User is online |
Now that you guys mention GCC and game development, it turns out that Sega did used GCC as their reference compiler for the 32X (at least for the SH-2 part). That was the compiler they shipped on their SDKs, and if you take a look at some 32X binaries (at least on the CD games), you will find GCC symbols on those. Did they ever released the sources for whatever GCC version were they using back in 1994? Licensed Pirate® since 2006, 100% Buttcoin™-free, enemy of All Things JavaScript™ |
Duck Penis |
Posted on 19-03-22, 20:33 (revision 1)
|
Stirrer of Shit
Post: #112 of 717 Since: 01-26-19 Last post: 1763 days Last view: 1761 days |
Hey, this is a great thread. I was curious about what they used for the N64. So I went to TCRF and looked at strings. Found that "(null)" was present, which is how GCC represents null pointers passed to "printf("%s", NULL)". So I kept looking. There was this list of error messages, like when you try to divide by zero. I googled one of them that sounded kind of rare, "Virtual coherency on data". Turns out it's only used in other N64 games. A bit further down there was a reference to __osGetCurrFaultedThread. Out of curiosity, I google it and find "Manual: Nintendo Ultra-64 Programming Manual Plus Addendums". Apparently the devkit is publically available, didn't know that. Here comes the interesting part: in a pdf file that comes with the dev kit, it says this: Posted by ninpcman.pdf Disappointed, I think, "oh well, they must have rolled their own compiler" Anyway, I install it under Wine just to see what happens. The first thing that comes up when I open the installer? I checked, that's the real GCC. The COPYING file is still there, there is some gcc documentation, and there are some gcc strings inside the executables:
So what does that mean? They're still in business. Does it mean they have to give me the source code of the whole thing if I ask nicely, or would it only apply to their actual licensees? Can you even claim you own the copyright to GCC in your manual as long as you contradict yourself later on? There was a certain photograph about which you had a hallucination. You believed that you had actually held it in your hands. It was a photograph something like this. |
Duck Penis |
Posted on 19-03-22, 20:59 (revision 1)
|
Stirrer of Shit
Post: #113 of 717 Since: 01-26-19 Last post: 1763 days Last view: 1761 days |
Posted by tomman I'm no expert, but I think it's perfectly legal to make your custom copy of GCC for internal use without releasing it. Here's another interesting EULA, which indicates that I legally can "copy the Software for archival purposes," and use it "on any single computer with any compatible cartridge hardware". That should be the third different license so far for the same piece of software. This is in the same folder as their renamed GCC fork, by the way: EDIT: In my understanding, GPL makes NDAs on it null and void, so if you have a copy of a GPL'ed piece of software you by definition always have a license for it.
There was a certain photograph about which you had a hallucination. You believed that you had actually held it in your hands. It was a photograph something like this. |
BearOso |
Posted on 19-03-22, 21:01
|
Post: #62 of 175 Since: 10-30-18 Last post: 1450 days Last view: 1450 days |
Posted by sureanem Let's theorize. If you distribute the modifications in binary form to someone, you must provide the sources to them only. This implies an intent. Swapping your custom copy around a dev team should be OK. It's not going to be a problem giving coworkers the source code on request. If it leaks out or someone takes it, you're not distributing it to them, so you have no obligations. Regardless, the leaker or hacker, anonymous or not, is responsible now. If the person only stole the binaries and not the source, they can't fulfill the contract for anyone they would give it to and thus give up their own license to use and distribute the product. In the case of the Sega and N64 dev kits, they knowingly distributed their modifications in binary form, so they should definitely have provided source code. |
Duck Penis |
Posted on 19-03-22, 21:19
|
Stirrer of Shit
Post: #114 of 717 Since: 01-26-19 Last post: 1763 days Last view: 1761 days |
I don't understand your argument. With the binary, there comes a license. To take away the license file and continue distributing it is a violation of the GPL, unless you make a later written offer. But it doesn't take away the legal license, just the actual license file. Say my boss gives me gcc.exe + license. I take away the license and give it to my coworker Jim. He might not know it's covered by the GPL, but he still has a license to do whatever the GPL gives you the right to do, given by my boss or wherever he got it from. Say he then uploads it to the internet. Everyone obtaining the copy gets a license for the source, and even though they aren't informed of this right they still have it. It's not like playing tag, where the last link in the chain only has the responsibility. If I find a flash drive on the street with a copy of emacs on it I could contact the GNU project and ask for the source, I don't have to find whoever dropped it to get it. (Note - don't insert random flash drives you find on the street into computers) There was a certain photograph about which you had a hallucination. You believed that you had actually held it in your hands. It was a photograph something like this. |
BearOso |
Posted on 19-03-23, 01:56
|
Post: #63 of 175 Since: 10-30-18 Last post: 1450 days Last view: 1450 days |
Posted by sureanem No, but if you’re the only distributor of a binary, the onus is on you. Someone who only possesses and uses the binary and does not give it away (free or paid) is not obligated to provide source either. So even though Nintendo used a modified version of GCC, it stayed put. If someone dug up an old binary, Nintendo doesn’t have to give them source because there was no intent on their part to give out the binary. If that person gives it out intentionally, they need to provide the source. The GPL works really well in protecting the innocent party in this case. |
Screwtape |
Posted on 19-03-23, 02:58
|
Full mod
Post: #175 of 443 Since: 10-30-18 Last post: 1101 days Last view: 172 days |
> if you have a copy of a GPL'ed piece of software you by definition always have a license for it. Sure, you can *use* it, but you can't necessarily *distribute* it. The GPLv2 says to distribute a copy of the software, you must include the source-code, a written offer to provide the source-code yourself, or pass along the written offer you received from your provider. I'm guessing these dev-kit owners didn't receive the source-code *or* a written offer, it would have been impossible for them to legally distribute it. > It's not like playing tag, where the last link in the chain only has the responsibility. It really is. If person A published a program in source form, and international mega-corporation B distributed binaries to a billion users, it wouldn't be fair for mega-corporation B to leave all the responsibility of source-distribution on person A. In practice, it's not a big deal because person A generally pushes their source-distribution responsibilities onto international mega-corporation C (i.e. GitHub), so if B abuses C's resources they can fight it out among themselves and A doesn't care. But it should be *possible* for A to take care of things themselves, and if they do, they shouldn't have to shoulder other parties' responsibilities. > If I find a flash drive on the street with a copy of emacs on it I could contact the GNU project and ask for the source, Ignoring the question of whether finding (or losing) a flash drive counts as "distribution", you could contact the GNU project for a copy of Emacs even if you *didn't* manage to acquire a binary, because the GNU project gives out as many copies of Emacs as they can. In the case of a console dev kit, the console manufacturer only gives dev kits to people in their licensing program. If they include the source-code with each dev-kit they distribute, then they don't have to include a written offer to provide source code (because it's already provided), and so the dev-kit recipient can't pass on that written offer (because there isn't one), and so the dev-kit recipient is responsible for providing source to the people they give binaries to. The ending of the words is ALMSIVI. |
CaptainJistuce |
Posted on 19-03-23, 04:01
|
Custom title here
Post: #349 of 1164 Since: 10-30-18 Last post: 63 days Last view: 11 hours |
Posted by Screwtape So you're saying the GNU project is evil? --- In UTF-16, where available. --- |
Screwtape |
Posted on 19-03-23, 04:45
|
Full mod
Post: #176 of 443 Since: 10-30-18 Last post: 1101 days Last view: 172 days |
No, that's the Extensible Vi Layer. The ending of the words is ALMSIVI. |
Duck Penis |
Posted on 19-03-23, 12:31
|
Stirrer of Shit
Post: #118 of 717 Since: 01-26-19 Last post: 1763 days Last view: 1761 days |
Posted by BearOso They did distribute it though. This devkit was given out to lots of people, I think anyone who was a registered/legitimate developer. Whoever they gave the binary to had a right to get the source code from them. However, since they did not include a proper written offer, I believe it makes them GPL violators. There's this quote from the GNU project:
They included the license with the program, but no written offer. Well, I did the prudent thing and asked the GNU project - if it is, they'd want to know about it, so we'll see. There was a certain photograph about which you had a hallucination. You believed that you had actually held it in your hands. It was a photograph something like this. |
BearOso |
Posted on 19-03-23, 16:36
|
Post: #64 of 175 Since: 10-30-18 Last post: 1450 days Last view: 1450 days |
I'm under the impression the Pokemon version wasn't a dev kit and wasn't given out, only leaked. That doesn't make them GPL violators if they don't give out the source. The Sega and N64 dev kits obviously require the source. |