|
i don't think cross platform is as big a reason for the use of react/vue/angular (via the browser or electrum or ionic or cordova or whatever) as the dire state of native application development. when electrum offers 100% of the functionality you need for your desktop app and you can share code with your mobile app and your webapp why would you ever bother with garbage like uikit, wpf, qt or swing? put another way, if you want to see more high quality native apps direct your anger at the native app framework vendors not at javascript. javascript managed to deliver application frameworks people actually choose
|
# ? Jan 2, 2018 10:22 |
|
|
# ? Jun 13, 2024 06:30 |
|
the talent deficit posted:i don't think cross platform is as big a reason for the use of react/vue/angular (via the browser or electrum or ionic or cordova or whatever) as the dire state of native application development. when electrum offers 100% of the functionality you need for your desktop app and you can share code with your mobile app and your webapp why would you ever bother with garbage like uikit, wpf, qt or swing? the only reason people use js frameworks is that they work on the web react native is OK though the hot reloading is like babbys first small talk
|
# ? Jan 2, 2018 10:29 |
|
on the other hand, web app > native app since i don't have to install yet another loving dumb pointless app on my phone smartphones were a mistake
|
# ? Jan 2, 2018 10:35 |
|
CommunistPancake posted:discord is a web app that runs fine on web, desktop windows, desktop linux, android, and ios. so, i dunno, works for me i guess. Yet for 2 years and counting from release still does not have a spell check because electron shits itself when they tried to add it. I don't know if that says more about the front end devs being bad at coding or that web apps suck. But web apps suck.
|
# ? Jan 2, 2018 11:03 |
|
i've never wanted a spell check on a chat app and i see no reason to start now
|
# ? Jan 2, 2018 11:08 |
|
CommunistPancake posted:i've never wanted a spell check on a chat app and i see no reason to start now Good for you?
|
# ? Jan 2, 2018 12:14 |
|
thank you
|
# ? Jan 2, 2018 12:24 |
|
Da Mott Man posted:Good for you? just set up a list of spelling mistakes and corrections as an autohotkey script, op.
|
# ? Jan 2, 2018 12:35 |
|
I never thought I wanted a spell check on a chat app until I accidentally used one for a while and noticed how much poo poo I habitually misspell It's still a bad experience since the spell check hasn't figured out that sometimes I do foreign language shitposting but overall I think I look less stupid now
|
# ? Jan 2, 2018 12:50 |
|
Da Mott Man posted:Yet for 2 years and counting from release still does not have a spell check because electron shits itself when they tried to add it. I don't know if that says more about the front end devs being bad at coding or that web apps suck. But web apps suck. it says more that electron isn't great but is definitely better than any alternative out there
|
# ? Jan 2, 2018 13:26 |
|
eschaton posted:the difference is Lisp did all this stuff like 25-30 or more years ago, and it’s the rest of the industry that’s reinventing things in new terms a lisp user would look at a modern car and claim that they had ABS back in the day, sticking their feet through the bottom of their car
|
# ? Jan 2, 2018 14:20 |
|
the only thing a lisper won't claim to have invented first is a popular language
|
# ? Jan 2, 2018 14:24 |
|
great, a couple of smuggo mit graduates write a blog post and now their not invented here has become lore
|
# ? Jan 2, 2018 14:25 |
|
forth is much cooler for that though, they are (or at least were) a super-quirky community which tended to focus a lot on redefining the problems to suit the solutions
|
# ? Jan 2, 2018 14:41 |
|
forth never achieved enough dominance or collaboration to get much of a PR presence; at its height it was mostly EEs quietly applying it to their problems and never sharing their code with anyone. that it is so easy to build a forth means everyone makes their own incompatible needs-suiting implementation and the community gains no force multipliers. guilty here. in contrast lisp has many "enthusiasts" who write little code but treasure vague aspirations and legend. even today one need only mention lisp in passing for lispers to crawl out of the woodwork to explain how lisp is "good for ai", lisp has had every language feature forever, a metacircular evaluator is the most beautiful program in the world, &c, &c.
|
# ? Jan 2, 2018 14:55 |
|
deflecting from that since i mostly agree that lisp got a ton of things right way ahead of its time, and only really was made irrelevant due to not passing through the eye of the needle of actually affordable, but small/slow, hardware taking over in the 80s/90s, but as there are so few specifics held up here to argue about that i figured we could go more obscure with forth instead
|
# ? Jan 2, 2018 15:02 |
|
who started the "lisp is good for ai" meme anyway?
|
# ? Jan 2, 2018 15:33 |
|
Shinku ABOOKEN posted:who started the "lisp is good for ai" meme anyway? academics, perhaps?
|
# ? Jan 2, 2018 15:47 |
|
Shinku ABOOKEN posted:who started the "lisp is good for ai" meme anyway? mostly that it was a fairly natural fit for the symbolic ai of the 80s, symbols as a type and first-class values, and, by virtue of being a (flexible, as in impure) functional language, stuff like transforming trees of symbols, or defining search procedures over fairly complex data structures, is pretty straightforward compared to then-current competitors like pascal toss in there being commercial supported implementations (with actually functioning debuggers), macros to define e.g. pattern matching operators suitable for those tasks, and even clos (well, mit flavors or whatever variant existed in the system you chose) if you wanted to structure a larger system, and it certainly looks good in its historical context present-day ai is all optimization techniques, so no advantage there everything loosy-goosy of course, but these things always are, all the options are turing equivalent after all~ Cybernetic Vermin fucked around with this message at 16:09 on Jan 2, 2018 |
# ? Jan 2, 2018 16:05 |
|
Shinku ABOOKEN posted:who started the "lisp is good for ai" meme anyway? the mit ai lab. http://okmij.org/ftp/continuations/against-callcc.html
|
# ? Jan 2, 2018 16:59 |
|
Insomniac games decided some years back to turn all their internal authoring tools into web apps It was terrible for all the reasons you can imagine: https://www.gdcvault.com/play/1024465/Insomniac-s-Web-Tools-A They ended up trashing it all and rewrite everything on a platform that actually works (qt/c++)
|
# ? Jan 2, 2018 17:03 |
|
Shinku ABOOKEN posted:who started the "lisp is good for ai" meme anyway? I was going to suggest Doug Hofstadter.
|
# ? Jan 2, 2018 17:04 |
|
whenever i do my funny xplain side project i run into browser differences and unimplemented features probably every 15 minutes "how can i detect browser zoom? run a poll on window.devicePixelRatio, there's no event for it? no, only integral changes..." https://bugs.chromium.org/p/chromium/issues/detail?id=126287 "why can't i paste this data? oh, because chrome filters things out" https://bugs.chromium.org/p/chromium/issues/detail?id=487266 "there's no consistent way to modify a textarea, wtf?" https://bugzilla.mozilla.org/show_bug.cgi?id=1220696 i believe that if you're writing database skins everything is ok, but if you actually want a decent underlying platform to build a really loving good app, it's a hellish nightmare
|
# ? Jan 2, 2018 17:14 |
|
It's not only for security fixes that Chrome updates every two weeks. Will we see a standard WYSIWYG editor any time?
|
# ? Jan 2, 2018 17:56 |
|
Doom Mathematic posted:I was going to suggest Doug Hofstadter. Uggh-las Hofstadter, IMO.
|
# ? Jan 2, 2018 18:51 |
|
Cybernetic Vermin posted:deflecting from that since i mostly agree that lisp got a ton of things right way ahead of its time, and only really was made irrelevant due to not passing through the eye of the needle of actually affordable, but small/slow, hardware taking over in the 80s/90s, even this is part of the mysticism there existed multiple high quality lisp implementations on "actually affordable" hardware, and the market still rejected lisp as a platform
|
# ? Jan 2, 2018 19:10 |
|
(wrong-p 'skinner) #f (wrong-p 'kids) => #t
|
# ? Jan 2, 2018 19:44 |
|
Notorious b.s.d. posted:even this is part of the mysticism naah, i think mysticism is the claims that the stuff ran well on the affordable hardware. sure maybe on a properly expensive unix workstation, but the actually good bits had already ballooned in size at a time where stuff was as likely to trickle up from tiny home computers (in fact this is largely the direction that happened) as it was to trickle down from supposedly "advanced" systems 80s garbage collection pauses were surely a good way to get a quick pause to sip your coffee though
|
# ? Jan 2, 2018 19:48 |
|
lisp invented plangs
|
# ? Jan 2, 2018 21:52 |
|
tef posted:(wrong-p 'skinner) I and many others will buy this T-shirt if you make it
|
# ? Jan 3, 2018 00:44 |
|
Cybernetic Vermin posted:naah, i think mysticism is the claims that the stuff ran well on the affordable hardware. sure maybe on a properly expensive unix workstation, but the actually good bits had already ballooned in size at a time where stuff was as likely to trickle up from tiny home computers (in fact this is largely the direction that happened) as it was to trickle down from supposedly "advanced" systems in the 1980s, Unix WAS the affordable hardware. keep in mind a 386 system was about $25k in today’s money. people were not buying them to play video games apocryphally GC was so slow on the old LispMs that people would manually disable it, and let it run only when the machine was totally out. I think that’s mit cadr era tho, not the commercial units
|
# ? Jan 3, 2018 08:57 |
|
tef posted:(wrong-p 'skinner)
|
# ? Jan 3, 2018 09:07 |
|
Coral Common Lisp really wasn’t bad on a Mac Plus, and flies on a Mac II
|
# ? Jan 3, 2018 10:18 |
|
Notorious b.s.d. posted:in the 1980s, Unix WAS the affordable hardware. keep in mind a 386 system was about $25k in today’s money. people were not buying them to play video games I had a 386 and definitely played loads of video games on it
|
# ? Jan 3, 2018 15:50 |
|
Shaggar posted:I had a 386 and definitely played loads of video games on it Same, but not until say 1992.
|
# ? Jan 3, 2018 18:48 |
|
the 386 came out in 1985, followed by the 486 in 1987. we are accustomed to thinking of these as trash PCs to play video games on because we bought these chips five to ten years after they came out. even when a complete 386 system was $10k+ (in 1985 dollars) it was still considered to be a budget option!
|
# ? Jan 3, 2018 18:50 |
|
at any rate lispMs died because trash unix workstations and even macintoshes could run common lisp development tools and compilers faster than a lispM for 1/4th the price a reasonable lisp environment did not cost significantly more than a c++ environment, and the market still rejected lisp
|
# ? Jan 3, 2018 18:52 |
|
webass strikes again!!! https://twitter.com/swiftonsecurity/status/948698168895639553
|
# ? Jan 4, 2018 00:33 |
|
webassembly, like javascript, is close to the metal
|
# ? Jan 4, 2018 00:36 |
|
|
# ? Jun 13, 2024 06:30 |
|
From the Spectre paper it looks like javascript is sufficient. Maybe it's time to disable javascript until this is fixed. mystes fucked around with this message at 00:46 on Jan 4, 2018 |
# ? Jan 4, 2018 00:44 |