The web wasn't conceived as an application delivery platform, and the browser wasn't conceived as an operating system.
I agree wholeheartedly with this. That's why I suggested we need a new content and application platform that isn't a browser. I don't think it would be a popular concept though since people are so tied to their browsers even if they don't know what that is even. As for languages, I say we do it all in C since it is tried and true, mature as fuck and easy enough to learn. Most of the popular languages these days are only popular because of the same reasons that JS is- it is something to brag about knowing even if it is flawed and immature. But the language wars are for another day...
It might be neat to have something where the site/application is just a blob, and that gets downloaded and executed in a VM (so malicious things don't happen, theoretically).
So you visit Poal, say, and you get a blob that then gets executed to present you with the interface, and has normal methods of networked interaction (TCP/UDP directly) instead of goofy ass async JS HTTP requests.
Adobe was actually working on an ActionScript VM that would essentially deliver app code in a blob to the browser. It would have been a VM version of Flash Player on steroids. Combined with the media rich Flash Player display system, this could have enabled exactly the kind of thing we desperately need today. Of course the hatred of Flash and Adobe along with Steve Jobs lies about Flash on the iPhone killed off the work done towards that goal. We could have had something amazing, but people just couldn't control their hate and greed. Now we are stuck with bloated JS frameworks and even more bloated browsers for who knows how long. Thanks, haters and jews.
What the UTTER fuck are you even agreeing with? X11 was a FULLY NETWORKED CLIENT / SERVER protocol builtin the 80s to PRECISELY DO EXACTLY THAT.
They have been imagining the separation of interface from the software since the 50s and had working technology at least in the 80s, although I am not aware of any similar implementations to X11 in the 70s. Might Xerox Parc built some stuff already?
Http as a protocol was one guy in the 80s sitting around going, I am employed by this nicely funded government scientific agency, I have a network of computers here, how do I share my files with everyone? Oh, I know, I will just do what everyone had been theorizing in scientific and math circles since the 50s and just implement a simple file sharing protocol based around the musing of lots of people in the 20 years before including Ted Nelsons ideas that wrote about this in the 60s all of which was perhaps influenced by people that dreamed about this sort of stuff in the 40s and earlier.
I mean, I know what you mean about simplicity but the web was not designed for anything. It was an accidental creation of limited minds working with limited technologies all sitting on top of theoretical and practical work at had existed for a long time before we got access to it.
I was going to mention X11/X Windows, but I didn't want to stir up shit with that because it isn't fully what we would need now. We need a hybrid of X11 that can utilize both client and server power so that we can delegate the work to whichever end is best for the task. I don't want to take a step back to the concept of dumb terminals, X or otherwise, and thin clients are not the answer either. We need something new that is based on similar concepts but takes full advantage of the computing power we have today. We also need it to be something that doesn't enable further snooping, spying and data collection but still gives us options on where the code runs. Fuck the cloud and fuck the idea that we rent/subscribe to everything, which thin clients will lead to. I want something better and more privacy/security minded.
What the utter fuck are you talking about? How is synchronous vs asynchronous in any way relevant to the conversation?
The only reason anything web based was synchronous was incidental to bandwidth being 14.4k baud modems and server processing not able to keep up with the power of the average cel phone we carry in our pockets today. Actually, I haven't checked, but I bet the avarage raspberry pi kit you can buy today had more processing power than anything that could have been called a server in 1990/2 era or whatever.
To the extent that ANYTHING was designed with synchronous transactions in mind was incidental that any coder working on any of this stuff in the early 90s was a young adult in the 80s, most of whom had no access to any nice machinery because it was just that rare, and couldn't even BEGIN to imagine any of the ways the technology would be used.
If anything, we should be bitching and complaining about why full async support did not appear until 1999 when Lisp machines, unix, multics and everything from Lisp through APL through Prolog supported all the concepts necessary to implement a fully async networked world even back then. It all technically existed and was working in places like Xerox Parc, various university research departments and even started to be deployed in corps that could afford a fully networked async world.
Fuck, Lotus Notes was built in the 80s for the CIA (or NSA? can't remember) and had FULL SYNCHRONOUS AND ASYNCHRONOUS replication capability from the start, all the encription, user management, database (okay hierarchical databases but whatever) already built. All that would have needed to happen would be to open source the technology, reduce it to a set of protocols and use the Notes/Domino framework as a reference implementation that others could use to implement their own ecosystems and we would have had fully synchronous behaviour from the very beginning.
The web sucks because humans ... actually I don't know why it sucks. I'm not even sure that it sucks.
(post is archived)