Where does all this bloat come from?
I can undestand if you work with/consume high quality multimedia - they objectively require big amounts of memory and, preferrably, fast processors.
But what if it is text? Web pages? UIs? Why they are painfully slow? Goddamn, minimalism is popular today, you get to work with vectors - how VECTORS consume so much memory and CPU? Why a page with SOME. BIG. WORDS. on the front loads like it's dial-up all over again?
What the fuck with people, /g/?
>>60403616
Browsers are shit, we just kind of deal with it by slapping in more RAM.
>>60403616
> But what if it is text?
That actually is fast. Change your editor if it isn't.
> Web pages?
Often aren't that simple these days.
> UIs
Generally not that simple either. In particular multithreaded UIs are black magic. Most don't do them. Though of course on a GPU internally, some things may be running in a distributed fashion.
>>60403616
Literal megabytes of JS bloat which does stupid page effects that nobody asked for.
Botnets
>>60403616
By the way, the biggest problem starts at people using the wrong algorithms long before we get to optimizing muh memory allocation / deallocation and so on.
(You) might be doing some code and it might catch on because it does the job.
But (You) probably don't understand shit about when to use a certain data structure or algorithm at this point. Doesn't stop your code from catching on.
You may now be Adobe and since the money is coming in it's all fine, you just slap on hacks in the worst places and enjoy the money rolling in. "Innovative" marketing like selling software as a service will do way more for you than making it actually better.
>>60403616
Backwards compatibility.
>>60403724
For on-disk size. But in general you can't blame it much for memory and CPU usage.
>>60403739
Nope. We use old code and when it becomes inconvenient we wrap it in something else. Stuff gets translated from one language to another(in modern CPUs even the machine code gets translated to micricode first). All of this piles up.
>>60403764
Microcode or firmware is a method to keep machine details from software that runs on more machines than just that one. It has very little of a RAM / performance impact overall.
And if you wrap "inconvenient" things in wrappers that perform poorly and become "inconvenient" themselves, then that's not a necessity of backward compatibility - it's just you doing a shitty job. Maybe you were instructed to do so, but it's a shitty job regardless.
>>60403847
Virtually all software is "doing a shitty job" then. Along with every programming language.
>>60403847
>And if you wrap "inconvenient" things in wrappers that perform poorly and become "inconvenient" themselves, then that's not a necessity of backward compatibility - it's just you doing a shitty job
Literally modern web.
>>60403616
Nu males. The answer is nu males.
These are people who refuse to learn from the mistakes of the past, or learn anything at all really. If it runs on their iMac on San Francisco fiber, it's good enough for the web as a whole.
>>60403889
Can you show some examples that affect a lot of frequently used programs?
>>60403944
But there isn't a whole lot of wrapped legacy code contained in most websites?
>>60404094
>weebs
The real reason is that most programmers are more concerned with their income than the quality of their code. If not too many faults short term can be found them it's good enough. The problem is that it works a lot of the time...to the point that many who work as programmers really should be called stackoverlow copiers, because they'll actually copy code without knowing what it does... In production software.
I had all the evidence for realizing this awhile ago, and came to the conclusion too. However I haven't accepted that until recently due to how disappointing it is. No wonder people have 'imposter syndrome'so often. They absolutely should, they are imposters.
>>60404094
Kill you're self
>>60404038
X11.
>>60404070
They are bad code by HIS definition.
>>60405137
> X11.
If there is a wrapper or bypass somewhere in there (I think I heard it had some cruft, but IDK where), you kinda failed at pointing out where and how significant its performance / memory impact is.
Even after what, 30 years (not a situation that has kinda always been getting worse with layers on layers of horrible shit), it doesn't seem like everyone is running for Wayland immediately.
>>60405137
> They are bad code by HIS definition.
Whether there is bad code on the web or not wasn't what the quote was about.
>>60403616
a bunch of background javascript crap doing shit nobody even understands, but it undoubtedly ends with everything you do on the website ending up in google's archives
>>60403616
Back when computers were shit there was more incentive to do fancy shit to minimize memory & cpu usage. These days however there is enough memory and processing power available that theres not really any reason to minimize it.
There's also the issue of needing to scale to larger datasets and also of needing to keep shit secure, both things which are hampered by doing fancy shit to optimize for older shitty platforms.
>>60403689
Basically this. Disable JS by default and be surprised by how much better everything is.
>>60406607
>Disable JS
>become unabke to read page contents
Javascript.
Javascript tracking shit, tons of bloated js libraries loaded from third party servers and in general using js wrong.