People blame developers but it's all driven by a product mentality that favors rapid iterations and technical debt to run business experiments on customers. Slow-and-steady, carefully written software isn't tolerated within many product orgs these days.
No greater mistake can be made than to imagine that what has been written latest is always the more correct; that what is written later on is an improvement on what was written previously; and that every change means progress. Men who think and have correct judgment, and people who treat their subject earnestly, are all exceptions only. Vermin is the rule everywhere in the world: it is always at hand and busily engaged in trying to improve in its own way upon the mature deliberations of the thinkers. So that if a man wishes to improve himself in any subject he must guard against immediately seizing the newest books written upon it, in the assumption that science is always advancing and that the older books have been made use of in the compiling of the new. They have, it is true, been used; but how? The writer often does not thoroughly understand the old books; he will, at the same time, not use their exact words, so that the result is he spoils and bungles what has been said in a much better and clearer way by the old writers; since they wrote from their own lively knowledge of the subject. He often leaves out the best things they have written, their most striking elucidations of the matter, their happiest remarks, because he does not recognise their value or feel how pregnant they are. It is only what is stupid and shallow that appeals to him. An old and excellent book is frequently shelved for new and bad ones; which, written for the sake of money, wear a pretentious air and are much eulogised by the authors’ friends. In science, a man who wishes to distinguish himself brings something new to market; this frequently consists in his denouncing some principle that has been previously held as correct, so that he may establish a wrong one of his own. Sometimes his attempt is successful for a short time, when a return is made to the old and correct doctrine. These innovators are serious about nothing else in the world than their own priceless person, and it is this that they wish to make its mark.
Third-party delivery platforms, as they've been built, just seem like the wrong model, but instead of testing, failing, and evolving, they've been subsidised into market dominance.
It's really sad that in 2020, 10k+ engineers can't make a photo, video, post and message sharing website that is not a pain to use. We collectively failed as a profession. If one needs 2MB of CSS for such a website, there is clearly a problem.
Just as an aside, to give you an interesting benchmark—on roughly the same system, roughly optimized the same way, a benchmark from 1979 at Xerox PARC runs only 50 times faster today. Moore’s law has given us somewhere between 40,000 and 60,000 times improvement in that time. So there’s approximately a factor of 1,000 in efficiency that has been lost by bad CPU architectures.
The myth that it doesn’t matter what your processor architecture is—that Moore’s law will take care of you—is totally false.
The worst programs are written by people who know how to plug a million and one things together, but can't drill down and analyse the algorithmic implications of what they're doing. Electron runs like shit and inhales RAM is because it was programmed by people who don't have solid understanding of fundamentals. They understand a huge number of horizontal abstractions but they have no concept of how it looks vertically.
And therein lies the problem. How many interviewers dock marks for iterating over columns, instead of rows? Because that matters, a huge amount. How many interviewers would give credit for "how can you speed this up?" if the interviewee said, "write it in C, and simplify the datastructures you want me to use so we maximise sequential lookups over basic arrays, to maximise cache usage." They'll look at you like you have three heads.
"Don't you know Big N complexity is the only thing that really matters if you're looking for speed?" - then you get Electron.
2 years ago
2 years ago
Nothing fucking works. Nothing. Turning it off and back on again isn't a cute ritual, it's the cornerstone of all modern electronics. Everything ships with zero day patches. My $3000 TV crashes when you navigate an OSD menu the wrong way. Not the unnecessary smart features that it shipped with - that I of course augmented with a separate $300 purchase - but the actual 'treat me like a display' menu.
I work for a SaaS company and just as if not more work goes in to deciding how we measure uptime as goes in to designing for it. "Well, no customer incidents were reported, so that doesn't count as being down", "We have 1 hour of scheduled maintenance every week, but we still achieved 99.99 uptime" - it's creative, I'll give them that.
We talk about the network being unreliable as if a 200km 28ghz link and a trunk connection in a data center are the same thing. It's unqualified, and unhelpful, and nobody really knows what they are doing.
We "dismantle" waterfall as if it's not the same type of people who misunderstood the original publication doing the same thing with every other methodology and fad. (If you have not read "the leprechauns of software engineering" yet, it's an interesting read and worth a little bit of your time).
My house is full of devices, my history is full of purchases, that are a disappointment. I can't remember the last time I went a single. god. damn. day. without the things that are suppose to be helping me misbehaving in some way. And the worst part, is many of them can't even be fixed. They will putter along, the occasional patch, until they lose the attention of some swim lane on a plan of record somewhere and become e-waste.
I have been programing since I was eight. It was the most obvious passion I have ever found in life, but it feels like we're stuck. The arguments all feel the same boring old rehashed ones from over the last 20 years, probably longer. I'm bored. Is anybody else just tired of it all? Everything is amazing and crappy at the same time.
My workstation (E5-2640) has seen multiple generations of operating systems, video editing software, DAWs.
Browsers and web browsing in general is the only thing that I can tell it's getting consistently worse year after year.
I know it's an odd metric but 10-15 seconds to fully render a newspaper homepage is more than it takes for my full DAW setup (Cubase + FL Studio as VST plugin) to fully come up with tracks loaded and play button ready. I don't even recall dialup being this bad.