Is the madness ever going to end?

<i>They [Electron and React native desktop apps] constantly crash and have no value over a native desktop application.</i><p>No they don&#x27;t. I use VSCode, Slack and Spotify <i>literally</i> all day, all at the same time, with at least 3 Chrome windows open with between 3 and 20 tabs in each, and I honestly can&#x27;t remember the last time any of it crashed. I use a few other Electron apps on and off and they don&#x27;t crash either. Electron can be a bit of a memory hog, but that&#x27;s more Chromium&#x27;s eager caching than Electron. I also have Safari, Firefox and Edge running most of the time and they don&#x27;t crash either. In fact, the only app I use that crashes regularly is Android Studio.<p>Admittedly I only really use OSX so perhaps it&#x27;s different on Windows or Linux, but the &quot;Electron crashes a lot&quot; trope is almost certainly due to either using poorly written Electron apps (going native won&#x27;t help you if the dev sucks) or running something else that&#x27;s killing them (don&#x27;t run that thing?). Electron is fine.

Is it these people&#x27;s full time job just to rant about tech, or do they actually produce anything of value?<p>Use PHP if you want. Use Electron if you want. Use an established http server, or write your own. Host on AWS or a Raspberry Pi in your closet. Write plain HTML in notepad. No one gives a shit.<p>The only people complaining are the ones not actually coding.<p>&gt; Also, if you found any of the content on this website useful consider supporting me on Patreon ;)<p>Or maybe this was the reason for this strawman-laden rant all along.

Why is this post flagged? Despite it&#x27;s tone it state a valid and perfectly shareable consideration about IT evolution. Ok, it&#x27;s a rant but...<p>Apart from that: I goes beyond that post. We keep pushing new things that&#x27;s confirm Greenspun&#x27;s tenth rule, or they try to (re)create normally ignoring past creation, things Xerox &amp;c implemented decades ago.<p>The most advanced modern notebook UI are actually limited implementation of classic text editors like the old Tioga, just far heavyweight and a bit more pixel perfect&#x2F;colorful, and FAR LESS featureful.<p>Modern web (aka web 2.0)? An user-depriving Xerox Star Office System and so on.<p>The common denominator is re-implement such past ideas in a way to keep users out of control&#x2F;out of the &quot;power of IT&quot; + plugged-in surveillance. A rant might be not that insightful but definitively pertinent...

This is a whiney rant that boils down to “what’s wrong with kids nowadays? Things were so much better back in my day.”

Previous: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=29898030" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=29898030</a>

Abstractions are vital aspects of computer science.<p>Raw machine code worked! Why are <i>idiots</i> wasting time with this Fortran abstraction nonsense! &#x2F;s<p>Javascript itself is still just fine - npm et al always struck me as a bad idea I agree, but I have <i>zero</i> issue with using what are basically linkers and compilers on JavaScript in the same way we&#x27;d do the same thing with a c linker and compiler, so I am not sure where the hate is coming from. If you are railing against splitting your code up into easily manageable separate files and reusable components then you are just wrong, sorry (no one wants an entire application in a single spaghetti file).

He answers is own question at the end:<p><i>&gt; A company may save both time and money in the beginning because, yes you can get something running quickly, but they always pay the price in the long run when they are faced with all the issues their bad decision has caused.</i><p>So, you&#x27;ll get cleaner and more robust software when the economic incentives do not richly reward moving fast and cheap at the beginning of product life cycles.<p>(This comports with my experience. Mostly, over the last 15 years, during the explosion of the online economy and the associated great economic dislocations, you could avoid the madness by avoiding web programming.)

Once again, author gives a couple vague points against single-page applications while conveniently failing to address all of the bullshit that came before with PHP and friends. These rants are always the same two or three substance-less paragraphs, concluding before offering any sort of concrete examples or reasoning.<p>Author expresses:<p>&#x27;In the past IT people, whether we&#x27;re talking about programmers or something else, were very clever people.&#x27;<p>Yes, an age before programmers started wasting time on these vapid blog rants

&gt;They constantly crash and have no value over a native desktop application what so ever<p>Well maybe, but you can blame the the top companies and greybeards for not doing cross-platform better. Abstraction is how we don&#x27;t have to deal with deciding which OS is the best for the average user.<p>But anyway, there was still a wide spectrum of quality for desktop applications way back then like there is today.

100% Agree!<p>It is reasonable to try all those modern frameworks, tons of npm packages, webpack, sass, preprocessors, typescript.<p>And then throw this shit as far away as possible, because right now you have a beautiful javascript support in Chrome, if you need it, and you can render everything you need with React when you compile, i.e. prerender, your website.<p>Complexity these days is way beyond what we are ready to support. While React seems to be cool at start, in the end it, as any other framework, is a pain in the ass, 10 times bigger than any other solution

Another episode of &quot;old man screams at the cloud&quot;?<p>All points in this rant boil down to gatekeeping.<p>Sure, Electron, React-Native, PHP, and whatnot might not have maximum performance. But they allow people to build things that couldn&#x27;t do it in the past. Now we get apps from different demographics than the author belongs to, which is a great win in my book.

I&#x27;ve not read the article, but in general the answer to the question &quot;Is this madness ever going to end?&quot; is &quot;No.&quot; and usually because, if it isn&#x27;t outright clickbait, it&#x27;s only madness from one parties perspective, and indeed, more often than not, the madness is actually the new normal.

Although the author sounds a bit &quot;boomer-ish&quot; and &quot;get off my lawn&quot; in his blog post, I do agree with him in essence.<p>Especially his quip about websites being just JavaScript which generates the HTML. I have an old MacBook Air which I use when I have to work while travelling. This doesn&#x27;t happen often, so I didn&#x27;t replace it. It&#x27;s mind-boggling how slow the web has gotten on it.<p>If developers developed on older machines, they would get so frustrated by the bloat that it would get solved overnight.<p>I think the industry (generalizing) doesn&#x27;t care about this, because bandwidth, loading-times etc. are external costs and if one &quot;moves fast and breaks things&quot; in hope of getting another round of VC funding, then there is no time to care about this. In my view, the web gets worse with every new JS dependency&#x2F;framework and what not.<p>Oh and if your website or web-app doesn&#x27;t work without javascript (generalizing) then it is part of the problem.

I think this was posted less than 3 months ago.

TLDR: exploring new ways of doing things or optimising for developer experience is stupid.<p>&quot;The entry barrier to programming needs to be high!&quot;