tomman |
Posted on 19-05-20, 13:33
|
Dinosaur
Post: #334 of 1318 Since: 10-30-18 Last post: 16 days Last view: 10 hours |
Go offline then. Just like my personal site is doing in a week from now on, and noone will care. Neither I - the day I can afford it again, it will be back. Until then, no big deal. What is dead, it's dead. Ads are not invasive if done right. Sponsorship is not for everybody, I'll give you that one. Buttcoin mining is blocked out of PERFORMANCE reasons - it's MY computer, and you have no right to murder my battery life, turn it into a furnace, or simply deadlocking my browser because I dared visiting your minefield of a "website". Humans are selfish beings, so get over that and MAYBE we will see the light shining at the end of this murky tunnel. It's time to take back the control of our web browsers, and the Internet in general. Too bad the tech-illiterate dominate the world and simply don't give a damn on their rights to sane personal computing devices, which has led us to this dire situation. Just like politics in general, where a few with ill intentions rule over the herd of empty brains and conformists. Did I've mentioned how much I hate politics? And they ruined our computers too! (Yes, I take it very personal) Anyway, to not derail this thread (AGAIN!), I'll just say FUCK We DO need faster browsers, but we also need BETTER developers that don't think into $$$ as the first and only reason for everything! Turning JS into fast-downloading binary blobs is NOT the solution, people! Time to check out the Status Meetings reports from Seamonkey. Wonder if there is hope for a 2.49.5 release this year... or if at least I can suggest WG9S to compile langpacks for his Seamonkey builds (I miss Adrian Kalla's builds, another casualty of Mozilla being Mozilla) Licensed Pirate® since 2006, 100% Buttcoin™-free, enemy of All Things JavaScript™ |
wareya |
Posted on 19-05-20, 13:38
|
Post: #63 of 100 Since: 10-30-18 Last post: 1801 days Last view: 1366 days |
>Sponsorship isn't feasible for small/controversial/broad websites. Okay. >small You got me there, but you're not going to make much money off of ads on a small site anyway. Start a patreon or gumroad or something. Or run the site out of pocket. >controversial If you're too "controversial" to find sponsors, then you're "controversial" in a very bad way. >broad I don't see this being a problem. Surely you could sponsor specific publications if you're big enough, like big youtubers do. |
Duck Penis |
Posted on 19-05-20, 15:51
|
Stirrer of Shit
Post: #306 of 717 Since: 01-26-19 Last post: 1783 days Last view: 1781 days |
Posted by tomman Are you familiar with the concept of social Darwinism? (in other words: why there are hardly any such websites online) Ads are not invasive if done right. Nor are they profitable. Buttcoin mining is blocked out of PERFORMANCE reasons - it's MY computer, and you have no right to murder my battery life, turn it into a furnace, or simply deadlocking my browser because I dared visiting your minefield of a "website". Which is understandable. But for the browser to block it (while not blocking ads) is incomprehensible. I can't see slower websites being intrinsically inferior to ad-filled websites. And this is clearly a matter of taste, yet Mozilla forces theirs down my throat. The only logical explanation for this is that Google forced them to do it. Posted by tomman It's not their money, so that's not it. It's the demands from the web developers (who are too incompetent to write good anything, which is why they're webdevs in the first place) that someone else optimize their shitty code driving it. And to this Mozilla/Google gives in because it nets a short-term performance improvement in benchmarks, which in theory puts them ahead of the competition, which the webdevs can then sacrifice for a short-term development velocity improvement (alt: lower level of skill required = larger candidate pool = lower wages = more profit). And this cycle is overseen by Google and their subsidiary Mozilla, who both stand to gain from browsers getting more complex. Also remember that Mozilla is more concerned about politics than they are about browsers, and spend a large chunk of their incomes on random projects in Africa/India. In other words, they welcome anything which gives them more money for these, even if it lead to the death of the web in the long term, which is why they accept their current status. I agree with you that it's regrettable, but it's too late to do anything about it and worse things are coming anyway. The point of no return was passed several years ago. Unless you're aiming to become one of the few (in which case, good luck), there is not much of a point in complaining about things outside of your control. Posted by wareya There's a threshold effect. But yeah, for small places Patreon might be better. Provided they're not small and controversial, but then they can only take crypto donations/JS mining anyway. If you're too "controversial" to find sponsors, then you're "controversial" in a very bad way. Not really. If you're a political party with 5% of the vote, that's hardly fringe (or, well, it could be, but e.g. ALDE are both small and non-fringe), but you've more or less limited your advertiser pool to 0.05*(share of advertisers who want to make their political views an integral part of their brand) + 0.95*(share of advertisers who want to pretend they agree with you for some reason). For the really controversial places, not even stuff like Google ads or even HOT XXX WEBCAM MODELS/PDF DRIVER FREE DOWNLOAD will have you. And who decides if they're "controversial" in a very bad way? Zuckerberg the benevolent? I don't see this being a problem. Surely you could sponsor specific publications if you're big enough, like big youtubers do. How do you mean? If you're running a website about cars, then obviously car companies or whatever might be interested in sponsoring you. If you run a website about computer hardware, well you get the picture. But if you run something like imgur or twitter, then who'll advertise there? 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. |
tomman |
Posted on 19-05-20, 16:10
|
Dinosaur
Post: #335 of 1318 Since: 10-30-18 Last post: 16 days Last view: 10 hours |
You sound like one of those guys that would battle against the "right to repair" movement because "we lost this war, give up, embrace the suckyness". Just like my family and friends do, because each time I tell them why modern smartdevices are piles of junk designed to break, they SCREAM at me that I'm a social resent because I'm living in the middle of the worst era in a communist shithole, and the circumstances were more favorable, I would be already using a $1000 Samsung phone with sealed battery and unrootable Android. (BTW: I just has this discussion with two people over here - suffice to say, it didn't ended well). Let me get this point straight: that's how we got this world fucked up to this point. One thing is being pragmatic, but another is being a goddamned conformist, accepting the stats quo because "why bother" or "let me live my life". If we're going to accept the world of suck Big Commie Goverment and Silly Valley is trying to impose on us... well, might as well drink poison and end this suffering. I flatly REFUSE to accept this "new normal", damned be the consequences, because I know well the difference between "tech progress" and "consumeroid". Why in the hell I have to buy a new PC just to get more RAM or swap a CPU?! Why cellphone batteries have to be buried into a sandwich of glass and glue?! And why in the world a web browser has to provide a full blown OS, capable to run native code?! Instead... we get a new iPhone. And buttcoins. And a 1GB Facebook "app". Too bad we're pretty much alone on this battle. Still, I will stand up, with or without anyone's help. Licensed Pirate® since 2006, 100% Buttcoin™-free, enemy of All Things JavaScript™ |
Duck Penis |
Posted on 19-05-20, 19:58 (revision 1)
|
Stirrer of Shit
Post: #307 of 717 Since: 01-26-19 Last post: 1783 days Last view: 1781 days |
No point in screaming. It's like the weather. Sure, you can complain all you want, but it doesn't make it stop raining. The only reason things ever got this far was because the masses were willing to subsidize the development of, say, CPUs, because they needed them for business and play. Now that is done with smartphones instead, so that's where the development will be. The days of public opposition against DRM, rampant piracy, and 09 F9 11 02 9D 74 E3 5B (amazing that I could still type it out from memory) are long gone. Likewise, the masses aren't particularly bothered about their giant Facebook apps because it's easier to just spend $X on a new phone (realistically, they'd just ask their carrier to add a few dollars on to the subscription and send them a new one) with more storage, problem solved. So that's what'll happen, too. All you can hope for is that the future turns out closer to Psycho-Pass than Brave New World, although there's not really anything you can do to change it either way. An optimistic perspective could be that things will get far worse, so you should better enjoy things while they last. Although I suppose, for you, if things are shit now but maybe going to get better, it kind of cancels out. 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. |
Screwtape |
Posted on 19-05-31, 08:49
|
Full mod
Post: #262 of 443 Since: 10-30-18 Last post: 1121 days Last view: 192 days |
Posted by sureanem I mean, it doesn't strictly speaking matter whether that particular person was joking or not. There's six billion people in the world; I guarantee that there are people out there who genuinely believe the software industry's reverence for C is an instance of toxic masculinity, and I guarantee that there are people who genuinely revere C in a toxically masculine way. Both extremes are cartoonishly exaggerated (most of the time) and also perfectly correct (in a few tragic cases). My apologies if you didn't intend it this way, but to me it looked like you grabbed some evidence that was ambiguous, or at best ha ha only serious, to use as proof that some group you don't like are terrible people. And even if it was absolute, undeniable proof, why bring it up and sour the thread with needless vitriol? Believe it or not, buying your way out of resource limits is legitimately the easier option: it's work for your administrators, so it doesn't cost you developer effort, and the money spent is effective immediately, instead of after a six-month hiring-and-interviewing effort, and another six month training effort, and another two or three years on-the-job experience Yeah, which would make sense, but does Facebook really add features at such a high rate that it'd be infeasible? They already maintain one version in parallel for the third world (not making this up) because the real one got too bloated. Opportunity cost, man. Sure, they probably have enough engineers on staff that they could build a dozen Facebook Messenger clients in parallel, as long as they stopped doing everything else, but that's too high a cost for the expected benefit. Why hire even one engineer to work on rewriting an app when you can just get one of your existing admin staff to slot some extra meetings with standards bodies into their schedule? But after a few iterations of Moore's law, IE6 and XP have become paragons of bleeding-edge optimization. If Microsoft had just never tried to force Vista (which of course it would always have done since it was inevitable and had to happen), then computers would have gotten cheaper and bloat would have forcefully been kept down. If you're suggesting that sticking with IE6 and XP as hardware got faster would have left more computing resources for apps, why do you think apps wouldn't get even more bloated even faster when they didn't have to compete with the OS for resources? If you're suggesting that sticking with IE6 and XP would have meant hardware improvements would go towards cheaper, more power-efficient hardware instead of performance, then we wind up with cheap hand-helds that run XP as sluggishly and crashily as a 2001-era desktop PC would? Uh, thanks but no thanks. XP is just as much of a bloated, buggy mess as Windows 10, and Windows 3.1, and every other version: as much as they could get away with. If you claim that XP on modern hardware is elegant and optimised, you're really saying "I am OK with wasted resources as long as the total waste is less than X%", which is exactly what Microsoft believed when they were making XP in the first place, except that they probably used a bigger value for X than you did. But for the browser to block [cryptominers] (while not blocking ads) is incomprehensible. Advertising is a centuries-old industry, which just about every human interacts with regularly, as a customer or as a consumer. Meanwhile, cryptocurrencies are mostly known for Internet scams and ransomware. It's the same reason alcohol and cigarettes are freely available while other equally- or less-harmful substances are banned: nothing to do with the substances themselves, everything to do with tradition. It's the demands from the web developers (who are too incompetent to write good anything, which is why they're webdevs in the first place) that someone else optimize their shitty code driving it See that thing there, where you imply an entire group of people are incompentent, just based on their profession? That's kind of a dick move, you know? If you don't like the idea of C programmers being blanket-accused of toxic masculinity, maybe don't sink to the same level? All you can hope for is that the future turns out closer to Psycho-Pass than Brave New World, although there's not really anything you can do to change it either way. An optimistic perspective could be that things will get far worse, so you should better enjoy things while they last. The optimistic perspective is: you might not be able to change *every* circumstance, but there's always *something* you can change for the better, either for yourself or for other people. If you focus on the impossible things, everything you look at is doomed. If you focus on the improvable things, everything you look at is getting better, or could get better, or is already better, and that's a much healthier and much more *practical* attitude. The ending of the words is ALMSIVI. |
CaptainJistuce |
Posted on 19-05-31, 09:56
|
Custom title here
Post: #480 of 1164 Since: 10-30-18 Last post: 83 days Last view: 4 hours |
Posted by ScrewtapeThere's also the fact that the browser market, such as it is, has been incredibly competitive on battery life and performance lately, to the extent that one company was modifying their websites to trick another's browser into disabling performance and battery optimizations*. In such a market, allowing any asshat with two pennies to rub together to just hijack the users' processing power to dig for buttcoins is suicide. If a browser could turn off javascript without breaking half the web, they would do it faster than you can blink, because the performance and battery gains would be not-insignificant. *(That's actually why MS is giving up and making Edge just another Chrome skin, because Google was blue-shelling them repeatedly. They got sick of announcing a benchmark win, having Google sabotage them, and then getting called liars because Google's benchmarks showed Edge performing far worse than MS claimed. According to Edge developers.) --- In UTF-16, where available. --- |
Duck Penis |
Posted on 19-05-31, 13:48 (revision 1)
|
Stirrer of Shit
Post: #342 of 717 Since: 01-26-19 Last post: 1783 days Last view: 1781 days |
Posted by Screwtape Well, it probably isn't a widely-held opinion in that specific form. Anecdotally, I would however say it's a quite common view. Lots of Real Men Don't Eat Quiche-esque jokes about the matter, as well as the empirical evidence. If you see a female programmer, odds are they're a web developer and not doing embedded or something like that. If you discount MtF transsexuals from both groups, you would (anecdotally speaking) get a far greater reduction of the female programmers doing embedded than web. Anyway, my main point is that web developers are to programmers as security guards are to police: hardly anyone has as dream job to be a security guard, but if you don't make it into the police academy then that's what you'd often settle for. Yes, some people really do wish to become web developers, but they're few and far between. And thus, few web developers would be adept enough to be able to replace their use of JS with WebAssembly. Opportunity cost, man. Sure, they probably have enough engineers on staff that they could build a dozen Facebook Messenger clients in parallel, as long as they stopped doing everything else, but that's too high a cost for the expected benefit. I'm suggesting they hire additional engineers. Software development is as we all know something which parallelizes extremely poorly, but to hire another group who have absolutely no interactions with the first group would be more or less 100% parallel. The upside is that the app would be faster and better, which I imagine they would want. It wasn't a meeting with standard bodies, it was an arcane hack of the runtime, so it did take developer time. If you're suggesting that sticking with IE6 and XP as hardware got faster would have left more computing resources for apps, why do you think apps wouldn't get even more bloated even faster when they didn't have to compete with the OS for resources? That's a good point. If they'd have stuck with it for an additional 3-5 years then I suppose we'd have the best of both worlds, with enough el cheapo netbooks forcing applications to stay slim while they still could be run on desktop platforms. But anyway, pipe dream. XP is just as much of a bloated, buggy mess as Windows 10, and Windows 3.1, and every other version: as much as they could get away with. If you claim that XP on modern hardware is elegant and optimised, you're really saying "I am OK with wasted resources as long as the total waste is less than X%", which is exactly what Microsoft believed when they were making XP in the first place, except that they probably used a bigger value for X than you did. Yes, of course. But if computers get Yx faster, X = X0/Y. 2003 XP wasn't elegant and optimized, but 2019 XP is by virtue of comparison to any other alternative. It's not just Microsoft doing it. I used to own a laptop that ran Ubuntu 10.10 just fine back then. A few years ago, even Debian Stable w/ XFCE was slow on it. Advertising is a centuries-old industry, which just about every human interacts with regularly, as a customer or as a consumer. Meanwhile, cryptocurrencies are mostly known for Internet scams and ransomware. It's the same reason alcohol and cigarettes are freely available while other equally- or less-harmful substances are banned: nothing to do with the substances themselves, everything to do with tradition. I don't agree with your point about "less-harmful substances," but I digress. We already have a thread for that. Anyway, that doesn't exactly reflect well on Mozilla, now does it? They could have done nothing and let ad blocking take care of it. See that thing there, where you imply an entire group of people are incompentent, just based on their profession? That's kind of a dick move, you know? If you don't like the idea of C programmers being blanket-accused of toxic masculinity, maybe don't sink to the same level? I do believe it's true. Take for instance Reddit. They're a website company, 5th largest website in the US by Alexa, they reasonably should be able to afford good web developers. But when they make their new website, it's a slow mess that hogs resources. They could have just gotten unlucky, but this kind of stuff is "best practice", and Reddit are hardly unique in having slow websites. Go on any StackOverflow question, and you'll see the same: a question on how to accomplish something basic, top solutions 1-3 all use jQuery (or worse), and only in the bottom you find how to do it with vanilla JavaScript/CSS. This mindset is pervasive too: if you browse web development forums most discussion will be about new hot JavaScript frameworks rather than how to make good websites. So either they're incompetent, or they just have extremely shoddy best practices for no reason at all. At any rate, it appears to be more a matter of "a few names even in Sardis" than a vocal minority of bad apples, so I don't think it's incorrect to make the statement that web developers are lacking in aptitude. That C programmers get blanket-accused of toxic masculinity doesn't seem like a particularly outrageous or offensive conclusion to me, but to levy that criticism against the language feels like sour grapes and reflects poorly on the accuser. The opposite of that accusation which is often levied against web developers is unprintable, if rather commonly held. I'd much rather be considered toxically masculine. There's also that other common perception that people who a company don't want but are forced to hire get put doing web development. Anecdotally, people doing its opposite (e.g. low-level programming) seem (see patch notes - hardly an award-winner of ethnic diversity, let alone gender diversity) to be the demographic opposite of that group. I'm not saying this is actually the case, but as they say, no smoke without fire. The optimistic perspective is: you might not be able to change *every* circumstance, but there's always *something* you can change for the better, either for yourself or for other people. If you focus on the impossible things, everything you look at is doomed. If you focus on the improvable things, everything you look at is getting better, or could get better, or is already better, and that's a much healthier and much more *practical* attitude. Well, in what sense? I'm sure I could do something to improve some facet of something completely unrelated, but nothing much can be done about this particular development which is what we were discussing. To take the position that the things you don't like simply aren't happening seems somewhat idiosyncratic not to say insane. 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 19-06-01, 22:21 (revision 2)
|
Post: #48 of 77
Since: 10-31-18 Last post: 1209 days Last view: 1136 days |
>there are people who genuinely revere C in a toxically masculine way. I know a person (ax6) on Discord (pret server) who thinks "safe languages just create an illusion of safety, since CPUs are unsafe", "semantic type systems don't reflect what the actual hardware is doing", and "if you're making a mistake on C, it's because you're not a good enough programmer". He also claimed accidentally duplicating <code>if() goto fail; goto fail;</code> "happened because someone committed code without re-reading what they wrote" and "I mean, how many people looked at this and said "yep, looks OK"?"... Which seems reasonable enough, actually. I think "anything safer than C is bad" is a counterproductive and elitist (and toxic) attitude. I also think that blaming it on "masculinity", and claiming elitist behavior is somehow "masculine" is a discriminatory man-hating stereotype. |
wareya |
Posted on 19-06-02, 00:50
|
Post: #65 of 100 Since: 10-30-18 Last post: 1801 days Last view: 1366 days |
"toxically masculine" is actually just grammatical gymnastics for "regarding the toxic behavior that tends to occur when someone misunderstands values that are traditionally masculine". It doesn't have anything to do with masculinity itself. The situation you described is actually a perfect example of something that stems from this particular kind of toxicity. |
CaptainJistuce |
Posted on 19-06-02, 00:55
|
Custom title here
Post: #483 of 1164 Since: 10-30-18 Last post: 83 days Last view: 4 hours |
Posted by jimbo1qazThose all seem like things that should fall under "if you aren't using assembly, you aren't really programming." --- In UTF-16, where available. --- |
nyanpasu64 |
Posted on 19-06-02, 01:00
|
Post: #49 of 77
Since: 10-31-18 Last post: 1209 days Last view: 1136 days |
>Those all seem like things that should fall under "if you aren't using assembly, you aren't really programming." Good guess, pret is a Discord server where people try to decompile Pokemon games into "matching" C, where using the original (leaked or obtained via GPL) compilers produces matching binaries/ROMs with the same checksum. |
Kawaoneechan |
Posted on 19-06-02, 01:12 (revision 1)
|
A man of wealth and taste
Post: #243 of 599 Since: 10-29-18 Last post: 215 days Last view: 3 hours |
That's funny, a fair chunk of OpenPoké was decompiled. By hand. Not by me. ...also that is stupid. They are stupid. |
nyanpasu64 |
Posted on 19-06-02, 02:14
|
Post: #51 of 77
Since: 10-31-18 Last post: 1209 days Last view: 1136 days |
> ...also that is stupid. They are stupid. Are you saying pret is stupid? Because they decompiled the entire SM64 source code... quickly because it was compiled in debug mode lol |
Kawaoneechan |
Posted on 19-06-02, 05:36
|
FMOD user
Post: #245 of 599 Since: 10-29-18 Last post: 215 days Last view: 3 hours |
Even with the same compilers, bit-identical output from a decompile is a stupid desire. |
Duck Penis |
Posted on 19-06-02, 13:55
|
Stirrer of Shit
Post: #350 of 717 Since: 01-26-19 Last post: 1783 days Last view: 1781 days |
Posted by jimbo1qaz 2/3 of those are right. I mean, I don't see anything wrong with semantic type systems, but it is true that the hardware just sees a register. I think using C is like riding a bicycle without a helmet. Sure, on paper it might be more dangerous. But in practice, the other options just make you feel too safe and in practice increase the risk of accidents. Better to fear the machine and make sure you're doing things right. I think "anything safer than C is bad" is a counterproductive and elitist (and toxic) attitude. I also think that blaming it on "masculinity", and claiming elitist behavior is somehow "masculine" is a discriminatory man-hating stereotype. Posted by wareya But it's a reasonable stereotype. How many woman C programmers do you have? Or doing low-level stuff in general. I suppose I could count one (J. Rutkowsk., developer of Qubes OS), depending on your definition of woman (now that's for "Politics!"). Even then, <1% is not really a lot. Posted by jimbo1qaz Oh hey, that reminds me, the FSF ([email protected]) still haven't responded to my tip about GPL violation I sent them 2019-03-23. Posted by Kawa It means you've reproduced the source code, which is pretty cool. 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. |
Kawaoneechan |
Posted on 19-06-02, 14:14
|
Praise Celestia!
Post: #247 of 599 Since: 10-29-18 Last post: 215 days Last view: 3 hours |
Posted by sureanemI don't think you got my point there. You can reasonably expect the same or at the very least effectively the same behavior, but to expect bit-identical output from a higher language (it's all relative, C is pretty low) is quite frankly madness in my eyes. |
Duck Penis |
Posted on 19-06-02, 14:29
|
Stirrer of Shit
Post: #351 of 717 Since: 01-26-19 Last post: 1783 days Last view: 1781 days |
Posted by Kawa But how can you be sure you didn't miss some obscure edge cases? A modern compiler would all else equal probably get far more creative with UB. Think SM64's console-only PU freeze glitch, for instance. With a bit-identical compile, you can be sure you've reproduced it with 100% accuracy. 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. |
Kawaoneechan |
Posted on 19-06-02, 15:25
|
Off-Model
Post: #248 of 599 Since: 10-29-18 Last post: 215 days Last view: 3 hours |
That's the thing: you can't be sure you didn't miss some obscure edge cases. With a disassembly you can pretty much guarantee will result in a bit-identical end product, assembly and raw machine code being what they are. You rebuild the C code in any way, be it automated decompilation or intense manual labor, you're not gonna get a bit-identical compile, debug information or not. You can only guarantee it's bit-identical if you use the original source code and compiler. In fact, the very presence of that debug information would invalidate the result even then. |
Duck Penis |
Posted on 19-06-02, 15:35
|
Stirrer of Shit
Post: #352 of 717 Since: 01-26-19 Last post: 1783 days Last view: 1781 days |
Posted by Kawa If they have the compiler (via GPL), and ignore the build info (or spoof date etc, whatever goes into it), then they can use the compilation process to check that their manual/automated decompilation is correct. And if it is, they've recovered something which resembles the actual source code. Then they can apply various transforms to it to get something more probable to have been what the actual developers worked on, to find interesting stuff. The utility seems pretty clear-cut to me. 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. |