Posts by Caspian Prince 106 publicly visible posts • joined 16 Jul 2007
Well not the RMI classloader as such ... just the ability to load classes dynamically at runtime, even ones generated by the application itself. Strange how they never quite managed to get security locked down on this aspect in a satisfactory manner and in fact they've even removed the security APIs in the latest releases, presumably conceding defeat.
Then again - one of the greatest strengths of the JVM is dynamic classloading. So... hm.
Also depends how you value growth, as in, if you value growth above all else, great but a bit of a shame it came at the total collapse of the environment that support all life on Earth currently. As for myself I believe that growth as an end in itself is an utterly disastrous metric for success by any measure.
Tried quite hard to like C# here and at first it beguiles with many "finally! I can do X" moments (coming from Java-land) ... only later to discover actually it's not been that brilliantly thought through at all and just tacked on by the whims of whatever software paradigm was flavour of the month.
Just as a recap, here's Mode7's writeup on industry opinions on the matter a few months back. For better or worse, I'm quoted in it. (I make games with an online aspect).
https://modecollapse.substack.com/p/stop-killing-games-will-become-an
The pundits on the "gamers" side of the fence repeatedly show themselves to have an utter lack of understanding of even the most trivial technical aspects of handover and support. Let alone the legal framework under which this all has to exist. This is sad, because whenever someone who does actually know say, "yes, but you haven't actually thought about <area I am domain expert in>" the pro-SKG lobby does the argumentative equivalent of putting their fingers in their ears and shouting "LA LA LA!" followed by, "And anyway you're LYING!" There are even people in this very thread who claim to know something about B2C services and online games or even games development just showing their utter ignorance of the massive infrastructure within which the whole ecosystem operates.
Readers, the people who make these games are not lying. They are trying to warn you that if the EU attempts to legislate any of this stuff, you will just end up with no online games any more, or hopelessly crippled games, or really expensive games. The actually very small proportion of players that want to continue playing a game many years beyond the point where everybody else has gotten bored of it and left - which is why the games are shut down remember - are just going to make online gaming, for everyone as a whole, much, much worse, in one or more aspects.
Fortunately the EU are, it seems, actually aware that if they try and meddle this time, they will probably break everything. If they're going to succeed in anything legislative, the only hope they can have of anything workable is by using existing contract law, and a legally binding service level agreement. In practice this means that they will maybe enforce a contract between the publisher of a game and a customer such that a game is guaranteed to function as it did at the point of the contract's execution for a minimum defined period of time. And that is *all* you can expect them to be able to do, without completely ruining it for everyone.
It ain't exactly lacking for performance, and indeed the only thing really holding it back is the constant dereferencing, which Valhalla fixes. You can work around it today with non-idiosyncratic code but it looks hideous and is hard to maintain and you might as well just be using C++ if you go down that route.
Algorithms are king to the extent that you're choosing O(N) or O(Nsquared) etc but 1. in many cases the actual magnitude of N has a significant bearing on overall performance within the context of the overhead in which the algorithm operates (eg. interpreted vs JITted vs compiled vs compiled and optimised) and 2. when you implement the same algorithm implemented in different languages you're still looking at *orders of magnitude* in performance differences. Show me a fancy algorithm written in Python and I'll show you it running 50 times faster in Java, and then someone else will show me it running twice as fast in C, and then some smartass will make that 50% quicker in hand-crafted assembly language. This is probably Python's real problem ... it keeps getting bullied into lifting heavy weights it shouldn't be asked to lift and was never designed to cope with.
actually sitting there right now, applying over and over to jobs on LinkedIn (lol), and they're not even getting *replies*, let alone interviews. And on the rare occasion you get an interview, only 1 in 100 is maybe being made an offer, the rest rejected for entirely specious reasons. The same job will then resurface, unfulfilled, over and over, for months on end, despite the huge pool of people perfectly adequately skilled and experienced enough to do the work applying for it. UK businesses: get a grip. Stop being dicks in the hiring process.
I created something known as LWJGL. It flourished and attracted other maintainers and eventually became quite popular. I must have put in maybe a couple of years of entirely unpaid labour into it, and the current maintainers have put in decades more. I say the current maintainers, because eventually a certain Markus Persson became unbelievably, unfathomably, extraordinarily rich off the back of my work, at a time when I was literally selling possessions just to pay the rent. Did he ever even think of sending a bean my way? Did he f---. And that was the last time I bothered making my work free.
An unholy trinity of shit implementations of shit ideas that they repeatedly try to fix with a shit wrapper du jour every couple of years to try and pretend that is isn't all shit. I'm not touching that stuff even with yours, and I don't care that I'm going to end up unemployed and destitute because of it, I'd rather keep what sanity and hair I have left and remember fondly when you could just write applications that did useful stuff, using programming languages.
... is that it's easy to remember and read out a dotted quad but essentially impossible to remember or speak ff:12:34:56:78:89:0a:ff ... and it probably really is that simple.
If only they'd just added a couple more bytes to the address and left it with room to add more as needed.
That's how they still work - if you want to do it that way. These days the favoured way is to dynamically link a shared object library, the reason being that hundreds of applications will all be using the same code so you might as well have the OS only load it into RAM once. In other words after not-very-much linking it becomes hugely more efficient, not less efficient, to load the entire library instead of just bringing in bits and pieces.
Only about 1 in 10 people I've ever worked with over the last 35 years or so have really been software engineers... the rest do it for various reasons but not because *it's their thing*, they just kind of accidentally fell into it like accountancy or sysadmin or estate agents, and trudge along in a mediocre fashion, never really caring much for what they do.
That's not quite the case though: because RAM is not reserved by a single application it's easy to conflate RAM usage with the simple requirement of disk space. The reality is that RAM usage is amortised over the usage of multiple applications simultaneously, and not all applications are loaded at once. So while you are totally correct to say that it is way cheaper to bung 4GB of RAM into a server rather than shave 4GB off of the memory usage of the service running on it, it's also still true of an application that runs on a million installations: because 4GB of RAM installed on a million computers is amortised over the varying usage of a thousand applications. Your application isn't the only application to benefit from the 4GB of RAM: the other 1000 different applications also benefit from it, and that makes all 1000 applications quicker and easier to build and maintain, and that makes all 1000 applications cheaper to make, and that means that those million users pay correspondingly less for their software, and it *still* works out cheaper for them to just buy a 4GB stick than it is to pay for the extra cost of the engineering time to make 1000 applications a bit more efficient.
There are other concerns that the commissioners of software have long been concerned about need to be considered yet it's fashionable amongst us older devs to pretend that back in my day, we had to lick t' road clean before fatha would even let us go to school and we made every byte count, etc. etc.
The problem is if you cast your mind back to software development in the 70s and 80s and even 90s ... it was fucking awful. The tools were primitive. There wasn't a lot in the way of useful abstraction, which means stuff took ages to develop, was usually fragile and tied to specific bits of hardware, and riddled with the sorts of potential security bugs that would have you hung by the foreskin until sorry for in modern day computing (but fortunately back then everything was air-gapped).
Not only did stuff take ages to develop, it didn't actually do anything much that everyone takes for granted these days. Software almost certainly wouldn't have worked with any other character set than US ASCII. It wouldn't have handled RTL script. It very likely didn't have any undo feature, or any cut and paste or global clipboard. Your fonts would have been shitty bitmaps on a low-res monochrome screen instead of beautifully rendered glyphs in 32 bit colour rendered slightly larger on your 4K monitor for your tired miserable old eyes. It might not, if you go far enough back, even have had a GUI. It won't have been able to even access more than 2GB of RAM until it got a 64 bit OS. It would likely have just crashed when it ran out of physical memory. All of these things have been added and demonstrably made using software vastly better than it used to be... and it's all cost space. Everything's come at a cost in space, and CPU power.
And while it was taking ages to develop, very few people had much of the required patience and autistic attention to detail to actually do it properly, and so they commanded a very high price, which they charged for a very long time, and this vexes people who pay for these things to get developed, so they're very keen to make it a) easier and therefore less of an exclusive and hence expensive club and b) quicker so it costs even less to make and gets to market faster. And these two last drivers of market forces are the full force of what's driven software development for the last ... 3 decades or so? Make it quicker. Make it cheaper.
It's still vastly cheaper to buy RAM than it is to optimise software. Vastly, vastly, cheaper. Even at todays slightly higher prices. And ... I'm fine with that, because I can concentrate on the first, and most difficult bit of software development - make it work - for longer before I have to worry about the next bit - making it fast.
Today my crappy £200 phone has about 10 times the processing power of, and unfeasibly more memory than, the Windows PC I first used Microsoft Office on. I think it's a fairly low bar to reach to get a word processor running on something "low-powered", or perhaps everyone has become so inured to the incredible bloat of Windows products that it has become some sort of fantastical dream.
"This is one of the perils of trying to automate the assessment of an employee's value."
Are you kidding me ... honestly, this is one of the key symptoms of internal rot in a company when it starts to try and automate things like this. On the one hand - lol, Canonical have just shot themselves in the foot with facepalmingly accurate corporate stupidity. And on the other hand - would you want to remain at a company that actually set any merit behind this sort of treatment.
So long and thanks for all the work, Till.
Having had the pleasure of many languages over the years, including .net, I'd rather stick with Java than almost any other flavour-of-the-month.
Currently doing C code. I forget sometimes how wonderfully simple it is. And just how long it takes to get anything at all of any significance working.
... by which I mean, the same as if they were in the industry using their skills to do the job. Only then will you have a chance to attract people good enough to teach it well.
I would have loved to teach computer science... but not for less than one third of my current income.
I have absolutely no idea. I just install every update that comes my way from the Software Update centre. And then one day it just refused to boot after one of those updates - I forget the exact failure but it was obscure enough that I couldn't actually find a solution on the interwebs.
I have had a lot of trouble with hanging as well - some kernels are not so well tested as they should be perhaps. And don't get me started on Nvidia driver failures.
