Personally, I'm more of a fan of the "whatever runs the software I want to run, is the platform for me."
I've used some 15+ operating systems so far. And that is _not_ counting different versions. (E.g., I'm not counting Win'95 and Win'98 as different OSes.) And quite a few computers, ranging back to 8080 micros and PDP-11 minis. (Sadly, I couldn't afford one of those PDP-11s as a home computer, though;) Plus a bunch of game consoles.
Basically I don't give a rat's rear what OS or chip is in the computer, as long as it runs the stuff I want to run.
A lot of that stuff does happen to be games though. And between the pain of emulating those in Linux, and the 1 game every year which gets ported to the Mac... well, you could say I'm a true fan of the Windows (and PS2 and Dreamcast and...) way of life.
I really thorougly enjoy just dropping by at the store and having a truck-load of applications and games to choose from. In the Windows world alone I had some two dozen games to keep me busy in between Warcraft III and Doom 3. (Did anything else even get announced for the Mac in between those?)
Yeah, I know it's not as properly nerdy a passtime as drooling over G5 benchmarks. But what can I say? I'm damn happy with this way of life.
"and Macs dont run a *lot* of software essential to people."
Bingo. The _only_ real reason to own a computer is to run software on it. All those benchmarks, SPEC or DNA splicing or whatever, don't mean jack squat if it doesn't run the software *I* want to run.
Actually, I don't think it's about being used to paying or not. It's:
1. about perceived effort.
You wouldn't ask a total stranger to come haul your fridge up the stairs for free, would you? Yet I've had neighbours I've barely spoken to before, pop up at my door and ask that I fix their broken Windows install, or come configure some broken 5 year old scanner they picked off a flea market.
From my experience, a lot of people perceive it all as "bah, it's easy. It's stuff school kids do in their free time. I could do it all myself in half an hour, except... erm... I have better stuff to do than learn all that crap." People who can't "program" their VCR's clock, insist that _my_ job is some no brainer that a semi-trained monkey could do.
(Extra pain points for when it's a client or manager which acts like this. And insists that since he can arrange some numbers and a lame graph in Excel in an hour... you surely don't need more than that to make a full fledged enterprise system which actually collects all that data, manages it, calculates it, and can serve reports looking like his Excel mockup.)
2. it's about the underlying perception that anyone in the IT business doesn't have any life anyway, and doesn't have anything better to do with that time anyway.
This comes in various flavours. From the neighbour who thinks that you have nothing better to do than install his hard drive, to the boss who literally asked me "what do you need free Saturdays and Sundays for, anyway?"
The same boss would never think of constantly asking the janitor to do overtime. And the janitor's neighbours would never think of calling him/her to clean their floors for free.
Why? Because they can grasp the idea that the janitor would rather do something else in that time. They can grasp the idea that said janitor is only doing that work for the money. Where the work hours end, it ends.
Yet in the case of IT personnel, the perception is that "bah, these nerds would probably go spend their time in front of a computer anyway." So, hey, might as well get some unpaid work out of it, right?
"If the person is a web designer or a programmer and needs to test to see how the site renders or the program compiles/runs outside of the Mac environment?"
So how about actually testing on a PC then?
E.g., would you really trust the fonts to get rendered _exactly_ the same on your Mac as on a real PC? Emulator or no emulator? From my experience, at least in Java the font widths vary considerably between the two platforms. Are you _sure_ that that emulator does an exact emulation of PC rendering, instead of just piping it all through to MacOS?
And, no offense, but I'm sick and tired of programs and web sites designed under the assumption that "all the world is made of computers _exactly_ like mine, using _exactly_ the resolution on mine, _exactly_ the font size on mine, etc. And, oh, surely everyone has the toolbar the exact same size as mine, and set to auto-hide."
And then someone has their Windows set to use large fonts. (It _is_ a standard Windows option, you know.) Oops.
This isn't only about web sites, and not only about Macs. Linux is probably the biggest nightmare in this aspect, with 90% of the programs crapping out if you have your X set to use 100 dpi fonts instead of 72.
(Again: they're standard X fonts, and it's a standard X dot pitch option. And some of us actually use it because we want something easily readable, and not a squinting fest. So I wish people jolly well got around to actually testing their programs in it.)
But either way, back to the emulation thing, it seems to me like just the same old excuse to test everything on only one machine. Maybe a milder form of it, but nevertheless.
Doubly so when that machine is a laptop and/or has a non-standard aspect ratio LCD screen. (Which most Apple LCD screens do, laptops included.) Most of us still use CRTs with a 4:3 aspect ratio. Wouldn't it be nice if you had at least _some_ idea of how that site would really look on a real user's screen?
Don't get me wrong, I'm not against _developping_ stuff on a laptop and on the go. But if that's the only machine you _test_ it on, well, no offense, but I don't want to see the final site.
Well, that's the whole problem with patents nowadays. I don't mind patents that detail a speciffic and reproductible way of solving a problem right now (yes, including software patents for new algorithms.) But patents on vague and obvious concepts are just nuts.
_Everyone_ occasionally has one vague idea that's impossible or economically unprofitable to do right now, but might be an obvious thing for the future.
E.g., back in the late 80's, I had this idea "hey, you know, how come noone makes a game console out of a PC. You just replace the floppy and hard drive with a Nintendo style cartridge slot, and... uh... stuff like that." (There were no CD ROM based consoles back then, hence the cartridge idea.)
You'll have to admit that it's an obvious idea. Especially since consoles _are_ computers anyway. Impractical at the time, yes, but obvious. But had I patented it, now I could ask for royalties from Microsoft for their XBox.
Small computers? Ditto. Rewind further back in time to the early '80s, and there's me in high school with my Sinclair ZX-81 home computer, thinking, "man, you know what would be cool? If it had an LCD display on top and some batteries, so I can haul it around."
'Course, I had no idea how to actually design such a thing, but the idea was obvious. I don't doubt that at least half the ZX-81/ZX-Spectrum/Commodore-64/whatever owners had the exact same thought at one point or another. (Not to mention the poor buggers lugging around those luggage sized "portable PCs" at the time.)
Well, aren't you sorry that you didn't patent that concept back then? Now you could be asking for royalties from every single manufacturer of laptops, portable consoles, or PDAs. Probably from the cell phone makers too.
Well, if at least it tripped on stuff that was within the same word, it's still something.
My canonical example of a stupid censorship is from the Sega forums. So I had a question about the level where "monsters jump out of a glowing pile." Looks innocent, right?
Now let me capitalize the letters that tripped their retarded filter: "oF A Glowing". Or lemme try without the spaces: "oFAGlowing".
Yep. Their filter was "smart" enough to figure out that if it glues disparate words together and then takes a piece of the result, it ends up with "fag".
I mean, Jesus Christ, you'd think that not even Beavis and Butthead would be that obsessed as to find "fag" in that sentence. Yet it was good enough to censor for Sega.
Basically that's the problem _I_ have with automated censorship, whether on a board, or in a doll, or whatever. It's invariably (A) retarded, and (B) actually pointing out weird letter associations that no sane human would have even noticed without the program's help.
Well, yes, except noone's ever managed to _know_ what will be required of those cards in the future.
I've seen a buttload of synthetic benchmarks (all the way from the 80's), and invariably they never predict anything useful. They're actually _less_ meaningful than just running your favourite game or app or whatever.
First of all, computers are too complex to put everything into a single number. Graphics cards too. The exact mix of instructions (e.g., which shaders are used), the exact data set (e.g., overdraw vs polygon count vs texture resolution), and so on, can make it all behave completely differently.
As a simple example, we already have at least two fundamentally different speeds for graphics cards: fill rate _and_ T&L speed. In a given game and in a given resolution, you may be limited by one, while the other still is barely stressed at all.
Now add on top of it other factors like pixel and vertex shaders, FSAA, anisotropic filtering, etc.
What you're left with is: every single game, in every single setting, will act very differently, speed-wise. Each game will produce a very different result when used as a benchmark.
And those purely synthetic benchmarks will not predict future games, they will just produce purely artifficial results. Results that don't resemble the actual performance of any actual game. Past, present or future.
Which brings me to two wishes of mine, when it comes to benchmarks:
1. That people would stop using these crap synthetic benchmarks as some Supreme Truth (TM), and just face reality. As in, "yeah, but how does it behave in actual _games_"?
2. That people would use more than 2-3 games as benchmarks.
I mean, really, there's more to computer gaming than Quake 3, UT and Serious Sam. Tell me how it performs in a flight sim, for a change. Or in a racing game. Or how about an RPG? Yeah, tell me how well it does in Morrowind or DAoC or whatever. Or, hey, how about benchmarking it in ePSXe? (Yes, that PSX emulator.) Or in a strategy game? Etc.
Now those might help me more than the frame rates for some games I don't even play. (No, I'm not that much into FPS any more.)
Personally I'm not that sure that image and momentum play that much of a role.
Except for
1. a handful of nerds who are still stuck at the "Nintendo rules because of Donkey Kong on NES" point in time, and
2. the few die-hard nerds debating MHz and megabytes and "Microsoft sucks"
... the rest of us just look at the available games, and make a choice based on that. It's all about the available games. It's that simple.
Hype and past successes can only help that much. It didn't help the N64 compete with the PSX, did it? Yes, a lot of people wanted a N64 based on Nintendo's past success, but in the end they went and bought a Playstation instead. Go figure.
It makes a nice elitist story that people are just sheep which buy based on hype alone, but in practice it's just not so. In practice, most people actually think twice before throwing 179$ on a console. They read a few reviews. They look at the available titles. And even more importantly: they talk to each other, too.
And in the end what does matter are the games. And _that's_ what Sony did right all along. And that's where Nintendo and MS screwed up big time.
"Gamecube has the best games."... if your definition of "best" means about a dozen games, missing several popular genres entirely.
Just as a random example of a missing genre: you know what sold about half the Playstations? RPGs. Square is probably responsible for more Playstations sold, than Sony's own games division and Sony's whole marketing division put together.
The Playstation 2 has a whole aisle of RPGs already, _and_ it also plays every single Playstation RPG ever made. While the Gamecube has... what? A verbatim port of a Dreamcast game (PSO), a verbatim port of an N64 game which wasn't even really an RPG to start with (Gauntlet Legends), plus one or two equally non-interesting things.
Maybe, just maybe, that's where the Gamecube failed miserably, after all. Not the lack of a DVD player, but the simple fact that it leaves a lot of people (myself included) looking through its catalog of games and concluding "blah, I'm not buying a new console just for one or two games which I _might_ ever want to play."
Nintendo is so entrenched in its cattering only for a niche, that it leaves a helluva lot of potential buyers, whose preferred genres are not in that niche, completely uninterested in the Gamecube.
Sorry, as was already said: it's all about the games. Even if the Gamecube had not only a DVD player, but even if it also played minidiscs, cassettes and recorded VHS tapes... as long as I don't find myself looking through its games list and thinking "damn, I really want to play this one... and that one... and that other one..." I won't get a Gamecube.
WWI? Screw WWI, there were a lot of horses used in WW2. Everyone thinks ultra-modern tanks and jeeps and half-tracks, but a lot of the "high tech" action in WW2 consisted of ancient guns pulled by horses. Or foot infantry lugging their supplies around with horses and carriages. And even post-WW2, there were a lot of horses and carriages used in rural areas.
The comparison isn't maybe that apples to oranges, too. I mean, a lot of the problem with switching to automobiles didn't only have to do with the cost of automobiles, but also with infrastructure. Automobiles didn't work that great with horse-friendly dirt roads (same as Linux doesn't do always play nice with Windows networks). Everyone needed to invest one helluva lot of taxpayer money into national networks of paved roads, before the automobile really became that useful. They also needed gas stations, repair shops, etc. That's infrastructure too. Then people needed driver schools, traffic laws, some centralized evidence of driver licenses, license plates and traffic offenses, etc. Yep, more infrastructure investment.
So basically, yeah, you are right. Noone's going to change a huge infrastructure over-night, just because there's a new gimmic in town. No matter how revolutionary it is. Didn't happen with automobiles, and it won't happen with Linux.
Actually, I find it funny that a lot of us geeks hate Microsoft for being... a thoroughly geek company.
If you look at all they've done, it's not so much that they're evil or whatever, it's just that they're the King Kong version of a hacker. You know, the kind which:
- says "**** the standards", if he thinks he can code his more efficient standards. (A lot of the standards that MS shafted were shafted not in the name of locking the customer in, but primarily because someone at MS thought he could squeeze a millisecond out of the code by ignoring the standard. Example: why MSVC wasn't fully ANSI compatible.)
- has a chronic case of "not invented here". (See, for example, their using their own file formats wherever possible.)
- has _zero_ clue of security (see MS Bob asking you to change the password if you typed the wrong password three times in a row)
- therefore he is a firm believer in "security by obscurity". Hey, if we don't publish the API which remembers everyone's passwords, we're safe from hackers, right? (See a long series of blunders including client-side "security" and the URL exploits of Hotmail.)
- and isn't going to start checking his array bounds either, because it would, like, cost valuable microseconds (see the awful lot of such exploits in all MS products.)
- doesn't give half a damn about what the paying customer wants. Sorta, "Hey, I'm the super-star of this show, _you_ do what I say." (See, among other things, why MS and IBM went separate ways about OS/2. Microsoft just went and coded their own features, and completely ignored the features that IBM wanted and paid for.)
- isn't affraid of changing APIs and formats whenever he has another cute idea. (Quite obvious for whoever ever used Office.)
- has about zero understanding of the world outside computers (yes, see their cable and phone blunders. Or see why Microsoft must pump money with a firehose into the XBox to keep it alive.)
And so on and so forth.
Basically: take the most undisciplined and arrogant coder you know of, and now picture him Microsoft sized. What kind of programs would he do? Well, probably the same that Microsoft does.
Well, I'm a heavy duty smoker and never had a problem with my PS2 yet. Believe me, when I say heavy smoker, I mean my house is at times like a fog chamber. Forget about atomic particles. You can see _my_ trail when I move through it.:P You'd think that if it was that sensitive to smoke, it would have died a long time ago.
I would expect though that _any_ console would see a number of returns, at the hands of clueless users. Remember that these things are not marketed at veteran IT professionals.
In addition to smoke and dust issues, there's also the issue that a console is lightweight enough to carry around with you. Just toss it into any old bag and there you go. (I used to take my Dreamcast with me even to work at the old company, when I was between projects and with nothing better to do. Another guy took his PS2 with him.)
Now imagine how many such consoles had to deal with various mechanical shocks (like being smashed into a wall or into a bar in the bus while in that bag) or condensation (after carrying it through a freezing winter, I don't think everyone let it warm up before turning it on.)
Actually, it just means that you're not quite as zealous a Nintendo fan as you think. Now honestly, I've already _seen_ this mis-represented as "woohoo, Sony is losing, Nintendo was right all along." In another forum, but still...
Heck, I'd be surprised if he didn't hear from at least someone about it, even long before the game was even a beta. The game was hyped no end. (And surprisingly enough, it turned out to be a good game nevertheless.) Trailers, tech demos, publisher interviews, whatever. Plus the fact that at least two 3DMark releases were based on its engine, and the whole "it's a synthetic benchmark" vs "but it's sorta based on a beta of the engine which may be used in Max Payne, so it's 100% representative of all real games" debate.
You could ask just about anyone on a MUD or in a chat-room, and they'd _all_ know about Max Payne a year before release. Just about everyone even already had made up his/or her mind, with some people saying it'll be the Second Coming of Christ (TM), and others saying it'll surely suck gameplay-wise, because the publisher only hypes the graphics.
Basically unless the guy lived barricaded in a cave, chances are at least one of his neighbour's cousin's roommate's kid or some such had a computer and a modem, and had heard about it. And _if_ he was such a big celebrity, you know, people would notice.
I mean, hey, if the game was called "Hulk Hogan", everyone would instinctively think "hey, it's called like that wrestler/actor guy." I'd think about half his neighbours would have told Mr Hogan about that, long before the game was even in beta.
I like to think of myself as a "casual gamer". (Albeit one which "casually" spends some 40 hours a week gaming;)
A lot of what "casual gaming" is all about is what the rest of CS calls "usability". Having a clean, easy to learn interface. Flight sims which need one to remember 25 key shortcuts are not for casual gamers.
The rest includes stuff like a balanced difficulty curve.
When you think of "casual gamers" you probably think of people who play yet another incarnation of Tetris. That's not true as such. Although Tetris is an example of a casual-gamer-friendly game, it's not the only game (or kind of game) in that category.
The Sims was already mentioned, so lemme give you another example: Diablo. It had a clean intuitive interface, a well balanced skill system, a decent difficulty curve (not a total cakewalk, but very much possible to get past any stage without mad skillz or cheating), and for a PC game it was remarkably bug-free. It was pretty much the ideal game for casual gamers, and it sold like crazy because of that.
Let me give you another example of the kind of interface design decisions which make a game more casual-gamer-friendly. Fallout versus Arcanum. Neither was completely for casual gamers, but Fallout got a few things better in that direction. Take aimed shots for example. Fallout had a neat little dialog for that, where you selected where to shoot, while Arcanum needed you to RTFM and know that ctrl+click is a head shot. Fallout had a big clearly labelled "end turn" button, Arcanum had an unlabelled oval that many people didn't even know it was a button. Etc.
Console games also tend to be a good bet for a casual gamer. Not all of them, mind you, but generally it's a bit safer a bet than PC games.
"Nor does it say that "sound management policies" and yoga classes are mutually exclusive, something your rant seems to imply."
"That said, AOL's decison to use meditation to counteract the effects of working twelve hour days is mind-boggling. Cause dude, lemme tell you, working 12 hour days for extended periods is just brutal."
Well, bingo. You seem to have answered your own question perfectly.
Working 12 hours a day for more than a week, _maybe_ two, is well past the poing where it becomes a problem. Those people are going to be so tired and jaded, that their productivity will hit rock bottom. I'm talking as in: actually getting less from them in one of those weeks, than in a 40 hour week.
FFS, no amount of meditation (or medication or anything else) is going to get someone to actually be productive and creative with that kind of a daily schedule.
If that's "sound management policy", I rest my case.
Well, bingo. It's simply that everyone occasionally needs to take a break and do something different.
E.g., at the old company, we had a few rounds of Counter-Strike in the lunch break. Worked wonders.
E.g., at the new company, some people go have a walk in the park when they're overloaded. Others just take a break on the balcony and watch the rabbits outside. Works wonders too.
So it's not just that those activities do the trick for you, it's that they really do the trick for anyone. Maybe reaching that conclusion on your own is what enlightenment is all about.
The only thing that Yoga offers, and these simple activities don't, is the mystical bulls**t. Gives people some wise sounding pseudo-science crap to believe in. (After all, everything coming from the East must be soaked in ancient wisdom, while a plain old walk in the park just sounds so mundane.)
Of course, you're probably right about the turnover rate too.
The whole system where one accepts to work 12 hours, 7 days a week, plus a load of stress from a PHB, is precisely based on pointless ambitions and peer pressure. People fear that they might be judged as failures if they get a job which pays even 1% less... even if that job is only 40 hours a week and it means being treated like a human.
So they get a big house, but don't actually live in it, because all their life is really spent in a cubicle. That house is just a sad show. They get a big plasma TV, but don't actually get any time left to actually watch it. It's an even sadder show. Etc.
If they actually actually achieved that clarity and inner peace and freedom from pointless ambitions, as most eastern religions preach, they wouldn't feel a need to submit to that kind of crap. Nor would they need some Eastern (read: New Age) hocus-pocus to believe in.
'Course, some of us came to the same conclusions without needing the mystical bulls**t as part of the deal. But, hey, if it takes some navel gazing to put some sense into some people, it's not that bad a price after all:)
Yeah, that's what I'm wondering too. And I too instinctively thought of Dilbert.
Way I see it, is something like this:
PHB: Our employees are complaining about stress.
Catbert: Let's give them Yoga classes to teach them to cope with stress.
Catbert: then we can give them even more s**t, and make them work even more hours a day
Basically it strikes me like it's not as much a "benefit" for the employees, as something done for the sole benefit of the employer. Yeah, let's teach them to cope with more stress, instead of giving them less stress in the first place.
The whole "patents are bad" idea is such a simplistic black-amd-white view of the world, it's not even funny.
I seem to recall some non-software patents that were plainly ridiculous and useless. Such as having a van with a horse on a treadmill in the middle, instead of an engine. Or a horse pushing the car, and the gas pedal moving the food closer to or farther from the horse. Literally. Someone patented that.
Does that mean that all automobile-related patents can be tarred with the same brush? Well, no. And well, same for software patents.
And you're right about the "it's obvious" factor, too. A lot of stuff that the anti-patent gang is screaming against as "obvious stuff", is only obvious because you've read about it already. Some of it was taught in universities.
E.g., compression algorithms and the infamous LZW patent, may not be as high-tech as security... does anyone really believe that any kid could reinvent it from scratch, without having ever read about it? It only took... what? A few _decades_ of computing to come up with that algorithm? If it was that obvious, how come noone coded something like that before?
E.g., security stuff is even funnier. The average programmer has _zero_ clue of security. (I should know, I'm one.) If you asked a programmer to invent a new encryption algorithm from scratch, the best he could come up with would be some form of ROT13 or XOR that anyone can crack in _milliseconds_.
_That_ is the obvious stuff. The real stuff actually needs extensive maths and extensive experience in cryptography.
A lot of those "I only did it so because it was obvious" arguments are just plain BS. No, they did it so by shamelessly copying someone else's work and putting their own name on it. Whether it's literally copy-and-paste, or writing one's own code for an algorithm described in a book, it does _not_ count as inventing it. It counts as using someone else's work.
Basically when I read stuff like "it's obvious stuff and unavoidable to do the same", about 99% of the time it really means "I copied someone's work, which is unavoidable because I have no bloody clue how to invent something like that on my own."
Maybe the real problem isn't only with patents. Maybe a lot of the problem is with the "I'm a super-star, I can do whatever I damn please (including, but not limited to, plundering anyone else's work), and noone dare hold me accountable for anything" attitude that some people have.
I did say "A lot of graphics artists", and not "all graphics artists".
I do know that there are people with plenty of clue about accessibility. Graphic artists, marketroids, PHBs, you name it. There are bright people among any of those.
I'd also like to say that there's a distinction between trained "web designer" and trained only as "graphics artist". Hence, why I spoke about the second category, not the first. The first usually tends to have clue about good web design. The second often tends to have the issue that his competence ends where Photoshop ends.
Basically the whole point I was trying to make is that that exaample commitee deciding the layout... actually _doesn't_ have a qualified web designer. They have pure graphics artists and marketeers instead. That's the whole problem. Noone in that example commitee has the competence to make a good site.
Sure, marketeers and graphics artists _are_ necessary in the team anyway. But as workers, not as decision makers. And yes, that goes for the programmers too. (No, I'm not any good at Photoshop.)
Let a proper designer make the final call, and maybe pair him/her with an usability expert.
Far as I can tell, it's not even a German issue, but simply an issue of "what happens when you get one PHB, two clueless marketroids and a couple of retarded graphics artists in charge of making the site."
Neither of the types in that mix have half a clue that people go to the site for the information, and not for the bullshit. Taken in random order:
- Graphics artists. A lot of graphics artists see it all as making art. They don't know, nor care, about usability of that site, or about the visitors' need to get to the actual information, or such. They want their work of art reproduced as such at all cost, regardless of how much their favourite colours offend the eye, their font is unreadable, and that funky illogical page arrangement leaves the user disoriented and confused.
Point in case: such a graphics artist turned PHB dictated that a page have dark-ish orange on lighter orange as a colour scheme. He also wanted it displayed in a 640x480 pixel window that the user can't resize. And it had 250K worth of graphics, plus about 150k worth of rollovers. (I'm not making this up!)
We called him "the Antichrist."
- Marketroids. These are often nice people, but are trained to do one thing: get people's attention to a product. E.g., they're great at making posters and brochures and other ads.
But they often fail to understand a very basic thing: web sites are _not_ ads. If people already came to your page, you already _have_ their attention, FFS. Now they want the actual info. If you only force them to wade through more marketing bullshit, they'll go away.
This is usually the kind of people who insist that the user _must_ click through ten marketing bullshit pages before they get to what interests them.
- PHB. Even when he's not really retarded, life as a PHB in a corporation isn't as easy for him as you think. The name of the game is: ass covering. Corporations don't like personal initiative. They like mindless droids who unquestioningly obey the rules and "strategic decisions" from above. Even when the PHB is really a bright guy, to stay a PHB he'll have to convince his superiors that he's a mindless droid.
Whenever said PHB actually makes a decision, he has to have the paper trail and everything to show his superiors that he did everything by the book. Or better yet, that he _didn't_ in fact take any decision himself, and merely supervised that everything be done _exactly_ as someone else decided.
So he'll just take the retarded design from the graphics artists and marketroids, and insist that everything must match those artists' screenshots with pixel accuracy. No PHB ever got fired for counting pixels, as their _only_ contribution to the project.
In case he does actually request a change of his own, it will be an easily defendable decision. E.g., insisting that the web site obeys the same style guidelines as the company's brochure. Never mind that it's a retarded decision, since the two don't serve the same purpose and aren't in the same medium. (The brochure doesn't make the user wait while a megabyte of graphics downloads on dialup.) But it's an easily defensible decision. If any superior questions it, the PHB can argue that he was merely obeying the rules and regulations from above. Noone ever got fired for that.
If you absolutely need to do something legitimate on short notice, that the program won't do, I do believe that most DBAs will have understanding. There'll be some assholes, but most will actually be reasonable guys.
If your SQL statement looks legitimate and harmless enough, I don't think many people will give you a too hard time for that. E.g., let's say you tripped the security trigger with a "select count(*) from patients where patient_id in (select patient_id from diagnostic where disease like '%Alzheimer%')". It will probably trip a few honeytokens all right, because it goes through classified diagnostic information. However it only gives back a count, and nothing actually linking the diagnostic to a name and address. And I don't think anyone will fire you or anything for that.
On the other hand, if you do a "select distinct patient_name, patient_address, credit_card_nr, credit_card_expiry from billing_data where credit_card_expiry > SYSDATE" then you damn better have a _very_ good explanation of why you needed that data. And your company damn better have a honeypot/honeytoken to catch that kind of a statement.
On the other hand, if you actually have a program to write, I can't see what's wrong with actually using those stored procedures that your DBA writes.
Contrary to the "I'm a super-star, I'll make my own rules and write my own simplistic SQL statements" attitude some people have, databases are a science in their own right. Expecting everyone to be an expert in SQL because they skimmed through a book is like expecting someone to be an expert in C because they skimmed through a book.
Most databases offer very flexible triggers. E.g., at least in Oracle you get to execute a whole script if you want to. And if you go through some middleware, you've got even more freedom.
So a "select * from names where last_name = 'Smith'" can be made to trigger exactly nothing. Assuming that the names table really contains nothing but names and ids.
On the other hand let's say it's something like "select * from PATIENTS where last_name = 'Smith'", where the PATIENTS table also contains house address, private phone number, etc. That is already retrieving private data for every single patient with the last name 'Smith'. With a well programmed honeytoken for a bogus patient called 'Random J Smith', it won't trip because the statement scans for 'Smith', it will trip because it retrieves that kind of private data.
If you do get your trigger tripped by something like that, you probably at least have an incompetent programmer (should have selected only the fields needed anyway), or at worst someone mining data about the patients. (No doctor treats all patients named 'Smith', so they have _no_ business retrieving the data for _all_ of them.)
And precisely _because_ it's easy for beginners to write bad programs, I do expect that programs dealing with such sensitive data be thoroughly tested. Yes, including by honeytokens and whatnot.
When you deal with that kind of sensitive data, taking the usual "oh well, we'll just write bad buggy code and patch it later" approach is plain old irresponsible. Letting any newbie code directly against the live hospital data without any safety checks, is as irresponsible as letting any newbie reprogram an airplane's systems in flight. A program which has to work on that kind of data should be thoroughly tested for any possible flaws, and have a competent team trying to hack into it too.
And yes, you'll never be 100% sure that it's bug free, but the honeytokens sound like a great extra way in which you can test it. And I fail to see why more bugs caught is a horrible thing.
Re:I'd like to take this oppertunity..
on
Head First Java
·
· Score: 1
Strictly speaking, it's more complicated than that.
The code generation part, at least in IBM's implementation, is not much worse than what your average C++ compiler does. (Been using it since 1999 at work, so I'd say by now I have some idea of what its strengths and weaknesses are.)
In fact, I know a lot of commercial C++ programs which shipped with poorer code. Why? Because at release noone turned the optimizations on, nor the assertions off. Someone just stripped the debug info and shipped the sub-optimal debug version as it was.
What slows down Java are a bunch of other factors, which are not an implementation issue, but stem from its very design and philosophy.
E.g., the strong OO design, combined with the fact that _everything_ is dynamically allocated, means a flurry of small objects are allocated and dealocated. Any program more complex than "Hello World" is going literally through anywhere between _thousands_ and tens of thousands of objects per second.
In C you could just allocate an array of structs on the heap, for exactly zero penalty. In Java the best you can do is an array of _pointers_ to those structs, each struct individually malloc-ed. And even the array _itself_ is malloc-ed and all you have on the heap is a pointer to it.
So if for you had an array of 1000 structs, the Java version will have an extra 1001 malloc and free operations, and an extra two levels of indirection through pointers. You can see how it can't possibly be as fast as the C program, even if the code generator is identical.
Now add the fact that a garbage collector has to go through the whole mess periodically, and figure out which of those gazillion little objects are still in use.
Then there's stuff like the string processing. In C you could do wonderful things with byte arrays, and it was damn quick. In Java's strict object oriented world, even putting together a small web page brings you back to problem 1: a flurry of small objects being malloc-ed and destroyed.
Now if you want to go even further, Java also has other features, like mandatory array size checking. Even if you _know_ that a buffer overflow can't possibly happen (e.g., because you copy between two buffers and allocated both at the same time and with the same size), Java still keeps the training wheels on. Every single byte copied will go through two array index bounds checks: once when reading it from the source array, once when writing it to the destination array.
On the other hand, unfortunately, this is precisely the stuff which makes Java nice in the first place.
E.g., sure, you could do much better performance-wise in C with malloc, than relying on Java's garbage collector. On the other hand, this also means that in Java you won't have to go through months of debugging memory leaks. (You can still get a memory leak in Java, but far less often.)
E.g., sure, you could do much better performance-wise in C by cleverly removing unneeded array size checks. But then you can and _will_ go to far and remove checks which are actually needed. See the ton of Windows and IE exploits based on buffer overflows. By contrast, Java's forcing you to keep the training wheels on, tends to produce far more robust code and far less debugging work.
Etc.
Basically Java does have a bunch of nice features, and it does trade some speed to get them.
So basically it's all about using an exploit, to bypass the game's normal rules.
Well, gee. I suppose soon we'll see games or mods with integrated aimbot and wallhack, for those who are nostalgic about using _those_ exploits/cheats/whatever-you-want-to-call-them.
Slashdot Mirror
Personally, I'm more of a fan of the "whatever runs the software I want to run, is the platform for me."
I've used some 15+ operating systems so far. And that is _not_ counting different versions. (E.g., I'm not counting Win'95 and Win'98 as different OSes.) And quite a few computers, ranging back to 8080 micros and PDP-11 minis. (Sadly, I couldn't afford one of those PDP-11s as a home computer, though;) Plus a bunch of game consoles.
Basically I don't give a rat's rear what OS or chip is in the computer, as long as it runs the stuff I want to run.
A lot of that stuff does happen to be games though. And between the pain of emulating those in Linux, and the 1 game every year which gets ported to the Mac... well, you could say I'm a true fan of the Windows (and PS2 and Dreamcast and...) way of life.
I really thorougly enjoy just dropping by at the store and having a truck-load of applications and games to choose from. In the Windows world alone I had some two dozen games to keep me busy in between Warcraft III and Doom 3. (Did anything else even get announced for the Mac in between those?)
Yeah, I know it's not as properly nerdy a passtime as drooling over G5 benchmarks. But what can I say? I'm damn happy with this way of life.
"and Macs dont run a *lot* of software essential to people." Bingo. The _only_ real reason to own a computer is to run software on it. All those benchmarks, SPEC or DNA splicing or whatever, don't mean jack squat if it doesn't run the software *I* want to run.
Actually, I don't think it's about being used to paying or not. It's:
1. about perceived effort.
You wouldn't ask a total stranger to come haul your fridge up the stairs for free, would you? Yet I've had neighbours I've barely spoken to before, pop up at my door and ask that I fix their broken Windows install, or come configure some broken 5 year old scanner they picked off a flea market.
From my experience, a lot of people perceive it all as "bah, it's easy. It's stuff school kids do in their free time. I could do it all myself in half an hour, except... erm... I have better stuff to do than learn all that crap." People who can't "program" their VCR's clock, insist that _my_ job is some no brainer that a semi-trained monkey could do.
(Extra pain points for when it's a client or manager which acts like this. And insists that since he can arrange some numbers and a lame graph in Excel in an hour... you surely don't need more than that to make a full fledged enterprise system which actually collects all that data, manages it, calculates it, and can serve reports looking like his Excel mockup.)
2. it's about the underlying perception that anyone in the IT business doesn't have any life anyway, and doesn't have anything better to do with that time anyway.
This comes in various flavours. From the neighbour who thinks that you have nothing better to do than install his hard drive, to the boss who literally asked me "what do you need free Saturdays and Sundays for, anyway?"
The same boss would never think of constantly asking the janitor to do overtime. And the janitor's neighbours would never think of calling him/her to clean their floors for free.
Why? Because they can grasp the idea that the janitor would rather do something else in that time. They can grasp the idea that said janitor is only doing that work for the money. Where the work hours end, it ends.
Yet in the case of IT personnel, the perception is that "bah, these nerds would probably go spend their time in front of a computer anyway." So, hey, might as well get some unpaid work out of it, right?
"If the person is a web designer or a programmer and needs to test to see how the site renders or the program compiles/runs outside of the Mac environment?"
So how about actually testing on a PC then?
E.g., would you really trust the fonts to get rendered _exactly_ the same on your Mac as on a real PC? Emulator or no emulator? From my experience, at least in Java the font widths vary considerably between the two platforms. Are you _sure_ that that emulator does an exact emulation of PC rendering, instead of just piping it all through to MacOS?
And, no offense, but I'm sick and tired of programs and web sites designed under the assumption that "all the world is made of computers _exactly_ like mine, using _exactly_ the resolution on mine, _exactly_ the font size on mine, etc. And, oh, surely everyone has the toolbar the exact same size as mine, and set to auto-hide."
And then someone has their Windows set to use large fonts. (It _is_ a standard Windows option, you know.) Oops.
This isn't only about web sites, and not only about Macs. Linux is probably the biggest nightmare in this aspect, with 90% of the programs crapping out if you have your X set to use 100 dpi fonts instead of 72.
(Again: they're standard X fonts, and it's a standard X dot pitch option. And some of us actually use it because we want something easily readable, and not a squinting fest. So I wish people jolly well got around to actually testing their programs in it.)
But either way, back to the emulation thing, it seems to me like just the same old excuse to test everything on only one machine. Maybe a milder form of it, but nevertheless.
Doubly so when that machine is a laptop and/or has a non-standard aspect ratio LCD screen. (Which most Apple LCD screens do, laptops included.) Most of us still use CRTs with a 4:3 aspect ratio. Wouldn't it be nice if you had at least _some_ idea of how that site would really look on a real user's screen?
Don't get me wrong, I'm not against _developping_ stuff on a laptop and on the go. But if that's the only machine you _test_ it on, well, no offense, but I don't want to see the final site.
Well, that's the whole problem with patents nowadays. I don't mind patents that detail a speciffic and reproductible way of solving a problem right now (yes, including software patents for new algorithms.) But patents on vague and obvious concepts are just nuts.
_Everyone_ occasionally has one vague idea that's impossible or economically unprofitable to do right now, but might be an obvious thing for the future.
E.g., back in the late 80's, I had this idea "hey, you know, how come noone makes a game console out of a PC. You just replace the floppy and hard drive with a Nintendo style cartridge slot, and... uh... stuff like that." (There were no CD ROM based consoles back then, hence the cartridge idea.)
You'll have to admit that it's an obvious idea. Especially since consoles _are_ computers anyway. Impractical at the time, yes, but obvious. But had I patented it, now I could ask for royalties from Microsoft for their XBox.
Small computers? Ditto. Rewind further back in time to the early '80s, and there's me in high school with my Sinclair ZX-81 home computer, thinking, "man, you know what would be cool? If it had an LCD display on top and some batteries, so I can haul it around."
'Course, I had no idea how to actually design such a thing, but the idea was obvious. I don't doubt that at least half the ZX-81/ZX-Spectrum/Commodore-64/whatever owners had the exact same thought at one point or another. (Not to mention the poor buggers lugging around those luggage sized "portable PCs" at the time.)
Well, aren't you sorry that you didn't patent that concept back then? Now you could be asking for royalties from every single manufacturer of laptops, portable consoles, or PDAs. Probably from the cell phone makers too.
Well, if at least it tripped on stuff that was within the same word, it's still something.
My canonical example of a stupid censorship is from the Sega forums. So I had a question about the level where "monsters jump out of a glowing pile." Looks innocent, right?
Now let me capitalize the letters that tripped their retarded filter: "oF A Glowing". Or lemme try without the spaces: "oFAGlowing".
Yep. Their filter was "smart" enough to figure out that if it glues disparate words together and then takes a piece of the result, it ends up with "fag".
I mean, Jesus Christ, you'd think that not even Beavis and Butthead would be that obsessed as to find "fag" in that sentence. Yet it was good enough to censor for Sega.
Basically that's the problem _I_ have with automated censorship, whether on a board, or in a doll, or whatever. It's invariably (A) retarded, and (B) actually pointing out weird letter associations that no sane human would have even noticed without the program's help.
Well, yes, except noone's ever managed to _know_ what will be required of those cards in the future.
I've seen a buttload of synthetic benchmarks (all the way from the 80's), and invariably they never predict anything useful. They're actually _less_ meaningful than just running your favourite game or app or whatever.
First of all, computers are too complex to put everything into a single number. Graphics cards too. The exact mix of instructions (e.g., which shaders are used), the exact data set (e.g., overdraw vs polygon count vs texture resolution), and so on, can make it all behave completely differently.
As a simple example, we already have at least two fundamentally different speeds for graphics cards: fill rate _and_ T&L speed. In a given game and in a given resolution, you may be limited by one, while the other still is barely stressed at all.
Now add on top of it other factors like pixel and vertex shaders, FSAA, anisotropic filtering, etc.
What you're left with is: every single game, in every single setting, will act very differently, speed-wise. Each game will produce a very different result when used as a benchmark.
And those purely synthetic benchmarks will not predict future games, they will just produce purely artifficial results. Results that don't resemble the actual performance of any actual game. Past, present or future.
Which brings me to two wishes of mine, when it comes to benchmarks:
1. That people would stop using these crap synthetic benchmarks as some Supreme Truth (TM), and just face reality. As in, "yeah, but how does it behave in actual _games_"?
2. That people would use more than 2-3 games as benchmarks.
I mean, really, there's more to computer gaming than Quake 3, UT and Serious Sam. Tell me how it performs in a flight sim, for a change. Or in a racing game. Or how about an RPG? Yeah, tell me how well it does in Morrowind or DAoC or whatever. Or, hey, how about benchmarking it in ePSXe? (Yes, that PSX emulator.) Or in a strategy game? Etc.
Now those might help me more than the frame rates for some games I don't even play. (No, I'm not that much into FPS any more.)
Except for
1. a handful of nerds who are still stuck at the "Nintendo rules because of Donkey Kong on NES" point in time, and
2. the few die-hard nerds debating MHz and megabytes and "Microsoft sucks"
... the rest of us just look at the available games, and make a choice based on that. It's all about the available games. It's that simple.
Hype and past successes can only help that much. It didn't help the N64 compete with the PSX, did it? Yes, a lot of people wanted a N64 based on Nintendo's past success, but in the end they went and bought a Playstation instead. Go figure.
It makes a nice elitist story that people are just sheep which buy based on hype alone, but in practice it's just not so. In practice, most people actually think twice before throwing 179$ on a console. They read a few reviews. They look at the available titles. And even more importantly: they talk to each other, too.
And in the end what does matter are the games. And _that's_ what Sony did right all along. And that's where Nintendo and MS screwed up big time.
"Gamecube has the best games."... if your definition of "best" means about a dozen games, missing several popular genres entirely.
Just as a random example of a missing genre: you know what sold about half the Playstations? RPGs. Square is probably responsible for more Playstations sold, than Sony's own games division and Sony's whole marketing division put together.
The Playstation 2 has a whole aisle of RPGs already, _and_ it also plays every single Playstation RPG ever made. While the Gamecube has... what? A verbatim port of a Dreamcast game (PSO), a verbatim port of an N64 game which wasn't even really an RPG to start with (Gauntlet Legends), plus one or two equally non-interesting things.
Maybe, just maybe, that's where the Gamecube failed miserably, after all. Not the lack of a DVD player, but the simple fact that it leaves a lot of people (myself included) looking through its catalog of games and concluding "blah, I'm not buying a new console just for one or two games which I _might_ ever want to play."
Nintendo is so entrenched in its cattering only for a niche, that it leaves a helluva lot of potential buyers, whose preferred genres are not in that niche, completely uninterested in the Gamecube.
Sorry, as was already said: it's all about the games. Even if the Gamecube had not only a DVD player, but even if it also played minidiscs, cassettes and recorded VHS tapes... as long as I don't find myself looking through its games list and thinking "damn, I really want to play this one... and that one... and that other one..." I won't get a Gamecube.
The comparison isn't maybe that apples to oranges, too. I mean, a lot of the problem with switching to automobiles didn't only have to do with the cost of automobiles, but also with infrastructure. Automobiles didn't work that great with horse-friendly dirt roads (same as Linux doesn't do always play nice with Windows networks). Everyone needed to invest one helluva lot of taxpayer money into national networks of paved roads, before the automobile really became that useful. They also needed gas stations, repair shops, etc. That's infrastructure too. Then people needed driver schools, traffic laws, some centralized evidence of driver licenses, license plates and traffic offenses, etc. Yep, more infrastructure investment.
So basically, yeah, you are right. Noone's going to change a huge infrastructure over-night, just because there's a new gimmic in town. No matter how revolutionary it is. Didn't happen with automobiles, and it won't happen with Linux.
Actually, I find it funny that a lot of us geeks hate Microsoft for being... a thoroughly geek company.
If you look at all they've done, it's not so much that they're evil or whatever, it's just that they're the King Kong version of a hacker. You know, the kind which:
- says "**** the standards", if he thinks he can code his more efficient standards. (A lot of the standards that MS shafted were shafted not in the name of locking the customer in, but primarily because someone at MS thought he could squeeze a millisecond out of the code by ignoring the standard. Example: why MSVC wasn't fully ANSI compatible.)
- has a chronic case of "not invented here". (See, for example, their using their own file formats wherever possible.)
- has _zero_ clue of security (see MS Bob asking you to change the password if you typed the wrong password three times in a row)
- therefore he is a firm believer in "security by obscurity". Hey, if we don't publish the API which remembers everyone's passwords, we're safe from hackers, right? (See a long series of blunders including client-side "security" and the URL exploits of Hotmail.)
- and isn't going to start checking his array bounds either, because it would, like, cost valuable microseconds (see the awful lot of such exploits in all MS products.)
- doesn't give half a damn about what the paying customer wants. Sorta, "Hey, I'm the super-star of this show, _you_ do what I say." (See, among other things, why MS and IBM went separate ways about OS/2. Microsoft just went and coded their own features, and completely ignored the features that IBM wanted and paid for.)
- isn't affraid of changing APIs and formats whenever he has another cute idea. (Quite obvious for whoever ever used Office.)
- has about zero understanding of the world outside computers (yes, see their cable and phone blunders. Or see why Microsoft must pump money with a firehose into the XBox to keep it alive.)
And so on and so forth.
Basically: take the most undisciplined and arrogant coder you know of, and now picture him Microsoft sized. What kind of programs would he do? Well, probably the same that Microsoft does.
I would expect though that _any_ console would see a number of returns, at the hands of clueless users. Remember that these things are not marketed at veteran IT professionals.
In addition to smoke and dust issues, there's also the issue that a console is lightweight enough to carry around with you. Just toss it into any old bag and there you go. (I used to take my Dreamcast with me even to work at the old company, when I was between projects and with nothing better to do. Another guy took his PS2 with him.)
Now imagine how many such consoles had to deal with various mechanical shocks (like being smashed into a wall or into a bar in the bus while in that bag) or condensation (after carrying it through a freezing winter, I don't think everyone let it warm up before turning it on.)
Actually, it just means that you're not quite as zealous a Nintendo fan as you think. Now honestly, I've already _seen_ this mis-represented as "woohoo, Sony is losing, Nintendo was right all along." In another forum, but still...
You could ask just about anyone on a MUD or in a chat-room, and they'd _all_ know about Max Payne a year before release. Just about everyone even already had made up his/or her mind, with some people saying it'll be the Second Coming of Christ (TM), and others saying it'll surely suck gameplay-wise, because the publisher only hypes the graphics.
Basically unless the guy lived barricaded in a cave, chances are at least one of his neighbour's cousin's roommate's kid or some such had a computer and a modem, and had heard about it. And _if_ he was such a big celebrity, you know, people would notice.
I mean, hey, if the game was called "Hulk Hogan", everyone would instinctively think "hey, it's called like that wrestler/actor guy." I'd think about half his neighbours would have told Mr Hogan about that, long before the game was even in beta.
I like to think of myself as a "casual gamer". (Albeit one which "casually" spends some 40 hours a week gaming;)
A lot of what "casual gaming" is all about is what the rest of CS calls "usability". Having a clean, easy to learn interface. Flight sims which need one to remember 25 key shortcuts are not for casual gamers.
The rest includes stuff like a balanced difficulty curve.
When you think of "casual gamers" you probably think of people who play yet another incarnation of Tetris. That's not true as such. Although Tetris is an example of a casual-gamer-friendly game, it's not the only game (or kind of game) in that category.
The Sims was already mentioned, so lemme give you another example: Diablo. It had a clean intuitive interface, a well balanced skill system, a decent difficulty curve (not a total cakewalk, but very much possible to get past any stage without mad skillz or cheating), and for a PC game it was remarkably bug-free. It was pretty much the ideal game for casual gamers, and it sold like crazy because of that.
Let me give you another example of the kind of interface design decisions which make a game more casual-gamer-friendly. Fallout versus Arcanum. Neither was completely for casual gamers, but Fallout got a few things better in that direction. Take aimed shots for example. Fallout had a neat little dialog for that, where you selected where to shoot, while Arcanum needed you to RTFM and know that ctrl+click is a head shot. Fallout had a big clearly labelled "end turn" button, Arcanum had an unlabelled oval that many people didn't even know it was a button. Etc.
Console games also tend to be a good bet for a casual gamer. Not all of them, mind you, but generally it's a bit safer a bet than PC games.
"That said, AOL's decison to use meditation to counteract the effects of working twelve hour days is mind-boggling. Cause dude, lemme tell you, working 12 hour days for extended periods is just brutal."
Well, bingo. You seem to have answered your own question perfectly.
Working 12 hours a day for more than a week, _maybe_ two, is well past the poing where it becomes a problem. Those people are going to be so tired and jaded, that their productivity will hit rock bottom. I'm talking as in: actually getting less from them in one of those weeks, than in a 40 hour week.
FFS, no amount of meditation (or medication or anything else) is going to get someone to actually be productive and creative with that kind of a daily schedule.
If that's "sound management policy", I rest my case.
Well, bingo. It's simply that everyone occasionally needs to take a break and do something different.
:)
E.g., at the old company, we had a few rounds of Counter-Strike in the lunch break. Worked wonders.
E.g., at the new company, some people go have a walk in the park when they're overloaded. Others just take a break on the balcony and watch the rabbits outside. Works wonders too.
So it's not just that those activities do the trick for you, it's that they really do the trick for anyone. Maybe reaching that conclusion on your own is what enlightenment is all about.
The only thing that Yoga offers, and these simple activities don't, is the mystical bulls**t. Gives people some wise sounding pseudo-science crap to believe in. (After all, everything coming from the East must be soaked in ancient wisdom, while a plain old walk in the park just sounds so mundane.)
Of course, you're probably right about the turnover rate too.
The whole system where one accepts to work 12 hours, 7 days a week, plus a load of stress from a PHB, is precisely based on pointless ambitions and peer pressure. People fear that they might be judged as failures if they get a job which pays even 1% less... even if that job is only 40 hours a week and it means being treated like a human.
So they get a big house, but don't actually live in it, because all their life is really spent in a cubicle. That house is just a sad show. They get a big plasma TV, but don't actually get any time left to actually watch it. It's an even sadder show. Etc.
If they actually actually achieved that clarity and inner peace and freedom from pointless ambitions, as most eastern religions preach, they wouldn't feel a need to submit to that kind of crap. Nor would they need some Eastern (read: New Age) hocus-pocus to believe in.
'Course, some of us came to the same conclusions without needing the mystical bulls**t as part of the deal. But, hey, if it takes some navel gazing to put some sense into some people, it's not that bad a price after all
Yeah, that's what I'm wondering too. And I too instinctively thought of Dilbert.
Way I see it, is something like this:
PHB: Our employees are complaining about stress.
Catbert: Let's give them Yoga classes to teach them to cope with stress.
Catbert: then we can give them even more s**t, and make them work even more hours a day
Basically it strikes me like it's not as much a "benefit" for the employees, as something done for the sole benefit of the employer. Yeah, let's teach them to cope with more stress, instead of giving them less stress in the first place.
Well said.
The whole "patents are bad" idea is such a simplistic black-amd-white view of the world, it's not even funny.
I seem to recall some non-software patents that were plainly ridiculous and useless. Such as having a van with a horse on a treadmill in the middle, instead of an engine. Or a horse pushing the car, and the gas pedal moving the food closer to or farther from the horse. Literally. Someone patented that.
Does that mean that all automobile-related patents can be tarred with the same brush? Well, no. And well, same for software patents.
And you're right about the "it's obvious" factor, too. A lot of stuff that the anti-patent gang is screaming against as "obvious stuff", is only obvious because you've read about it already. Some of it was taught in universities.
E.g., compression algorithms and the infamous LZW patent, may not be as high-tech as security... does anyone really believe that any kid could reinvent it from scratch, without having ever read about it? It only took... what? A few _decades_ of computing to come up with that algorithm? If it was that obvious, how come noone coded something like that before?
E.g., security stuff is even funnier. The average programmer has _zero_ clue of security. (I should know, I'm one.) If you asked a programmer to invent a new encryption algorithm from scratch, the best he could come up with would be some form of ROT13 or XOR that anyone can crack in _milliseconds_.
_That_ is the obvious stuff. The real stuff actually needs extensive maths and extensive experience in cryptography.
A lot of those "I only did it so because it was obvious" arguments are just plain BS. No, they did it so by shamelessly copying someone else's work and putting their own name on it. Whether it's literally copy-and-paste, or writing one's own code for an algorithm described in a book, it does _not_ count as inventing it. It counts as using someone else's work.
Basically when I read stuff like "it's obvious stuff and unavoidable to do the same", about 99% of the time it really means "I copied someone's work, which is unavoidable because I have no bloody clue how to invent something like that on my own."
Maybe the real problem isn't only with patents. Maybe a lot of the problem is with the "I'm a super-star, I can do whatever I damn please (including, but not limited to, plundering anyone else's work), and noone dare hold me accountable for anything" attitude that some people have.
I did say "A lot of graphics artists", and not "all graphics artists".
I do know that there are people with plenty of clue about accessibility. Graphic artists, marketroids, PHBs, you name it. There are bright people among any of those.
I'd also like to say that there's a distinction between trained "web designer" and trained only as "graphics artist". Hence, why I spoke about the second category, not the first. The first usually tends to have clue about good web design. The second often tends to have the issue that his competence ends where Photoshop ends.
Basically the whole point I was trying to make is that that exaample commitee deciding the layout... actually _doesn't_ have a qualified web designer. They have pure graphics artists and marketeers instead. That's the whole problem. Noone in that example commitee has the competence to make a good site.
Sure, marketeers and graphics artists _are_ necessary in the team anyway. But as workers, not as decision makers. And yes, that goes for the programmers too. (No, I'm not any good at Photoshop.)
Let a proper designer make the final call, and maybe pair him/her with an usability expert.
That's all I'm saying.
Far as I can tell, it's not even a German issue, but simply an issue of "what happens when you get one PHB, two clueless marketroids and a couple of retarded graphics artists in charge of making the site."
Neither of the types in that mix have half a clue that people go to the site for the information, and not for the bullshit. Taken in random order:
- Graphics artists. A lot of graphics artists see it all as making art. They don't know, nor care, about usability of that site, or about the visitors' need to get to the actual information, or such. They want their work of art reproduced as such at all cost, regardless of how much their favourite colours offend the eye, their font is unreadable, and that funky illogical page arrangement leaves the user disoriented and confused.
Point in case: such a graphics artist turned PHB dictated that a page have dark-ish orange on lighter orange as a colour scheme. He also wanted it displayed in a 640x480 pixel window that the user can't resize. And it had 250K worth of graphics, plus about 150k worth of rollovers. (I'm not making this up!)
We called him "the Antichrist."
- Marketroids. These are often nice people, but are trained to do one thing: get people's attention to a product. E.g., they're great at making posters and brochures and other ads.
But they often fail to understand a very basic thing: web sites are _not_ ads. If people already came to your page, you already _have_ their attention, FFS. Now they want the actual info. If you only force them to wade through more marketing bullshit, they'll go away.
This is usually the kind of people who insist that the user _must_ click through ten marketing bullshit pages before they get to what interests them.
- PHB. Even when he's not really retarded, life as a PHB in a corporation isn't as easy for him as you think. The name of the game is: ass covering. Corporations don't like personal initiative. They like mindless droids who unquestioningly obey the rules and "strategic decisions" from above. Even when the PHB is really a bright guy, to stay a PHB he'll have to convince his superiors that he's a mindless droid.
Whenever said PHB actually makes a decision, he has to have the paper trail and everything to show his superiors that he did everything by the book. Or better yet, that he _didn't_ in fact take any decision himself, and merely supervised that everything be done _exactly_ as someone else decided.
So he'll just take the retarded design from the graphics artists and marketroids, and insist that everything must match those artists' screenshots with pixel accuracy. No PHB ever got fired for counting pixels, as their _only_ contribution to the project.
In case he does actually request a change of his own, it will be an easily defendable decision. E.g., insisting that the web site obeys the same style guidelines as the company's brochure. Never mind that it's a retarded decision, since the two don't serve the same purpose and aren't in the same medium. (The brochure doesn't make the user wait while a megabyte of graphics downloads on dialup.) But it's an easily defensible decision. If any superior questions it, the PHB can argue that he was merely obeying the rules and regulations from above. Noone ever got fired for that.
If your SQL statement looks legitimate and harmless enough, I don't think many people will give you a too hard time for that. E.g., let's say you tripped the security trigger with a "select count(*) from patients where patient_id in (select patient_id from diagnostic where disease like '%Alzheimer%')". It will probably trip a few honeytokens all right, because it goes through classified diagnostic information. However it only gives back a count, and nothing actually linking the diagnostic to a name and address. And I don't think anyone will fire you or anything for that.
On the other hand, if you do a "select distinct patient_name, patient_address, credit_card_nr, credit_card_expiry from billing_data where credit_card_expiry > SYSDATE" then you damn better have a _very_ good explanation of why you needed that data. And your company damn better have a honeypot/honeytoken to catch that kind of a statement.
On the other hand, if you actually have a program to write, I can't see what's wrong with actually using those stored procedures that your DBA writes.
Contrary to the "I'm a super-star, I'll make my own rules and write my own simplistic SQL statements" attitude some people have, databases are a science in their own right. Expecting everyone to be an expert in SQL because they skimmed through a book is like expecting someone to be an expert in C because they skimmed through a book.
Most databases offer very flexible triggers. E.g., at least in Oracle you get to execute a whole script if you want to. And if you go through some middleware, you've got even more freedom.
So a "select * from names where last_name = 'Smith'" can be made to trigger exactly nothing. Assuming that the names table really contains nothing but names and ids.
On the other hand let's say it's something like "select * from PATIENTS where last_name = 'Smith'", where the PATIENTS table also contains house address, private phone number, etc. That is already retrieving private data for every single patient with the last name 'Smith'. With a well programmed honeytoken for a bogus patient called 'Random J Smith', it won't trip because the statement scans for 'Smith', it will trip because it retrieves that kind of private data.
If you do get your trigger tripped by something like that, you probably at least have an incompetent programmer (should have selected only the fields needed anyway), or at worst someone mining data about the patients. (No doctor treats all patients named 'Smith', so they have _no_ business retrieving the data for _all_ of them.)
And precisely _because_ it's easy for beginners to write bad programs, I do expect that programs dealing with such sensitive data be thoroughly tested. Yes, including by honeytokens and whatnot.
When you deal with that kind of sensitive data, taking the usual "oh well, we'll just write bad buggy code and patch it later" approach is plain old irresponsible. Letting any newbie code directly against the live hospital data without any safety checks, is as irresponsible as letting any newbie reprogram an airplane's systems in flight. A program which has to work on that kind of data should be thoroughly tested for any possible flaws, and have a competent team trying to hack into it too.
And yes, you'll never be 100% sure that it's bug free, but the honeytokens sound like a great extra way in which you can test it. And I fail to see why more bugs caught is a horrible thing.
Strictly speaking, it's more complicated than that.
The code generation part, at least in IBM's implementation, is not much worse than what your average C++ compiler does. (Been using it since 1999 at work, so I'd say by now I have some idea of what its strengths and weaknesses are.)
In fact, I know a lot of commercial C++ programs which shipped with poorer code. Why? Because at release noone turned the optimizations on, nor the assertions off. Someone just stripped the debug info and shipped the sub-optimal debug version as it was.
What slows down Java are a bunch of other factors, which are not an implementation issue, but stem from its very design and philosophy.
E.g., the strong OO design, combined with the fact that _everything_ is dynamically allocated, means a flurry of small objects are allocated and dealocated. Any program more complex than "Hello World" is going literally through anywhere between _thousands_ and tens of thousands of objects per second.
In C you could just allocate an array of structs on the heap, for exactly zero penalty. In Java the best you can do is an array of _pointers_ to those structs, each struct individually malloc-ed. And even the array _itself_ is malloc-ed and all you have on the heap is a pointer to it.
So if for you had an array of 1000 structs, the Java version will have an extra 1001 malloc and free operations, and an extra two levels of indirection through pointers. You can see how it can't possibly be as fast as the C program, even if the code generator is identical.
Now add the fact that a garbage collector has to go through the whole mess periodically, and figure out which of those gazillion little objects are still in use.
Then there's stuff like the string processing. In C you could do wonderful things with byte arrays, and it was damn quick. In Java's strict object oriented world, even putting together a small web page brings you back to problem 1: a flurry of small objects being malloc-ed and destroyed.
Now if you want to go even further, Java also has other features, like mandatory array size checking. Even if you _know_ that a buffer overflow can't possibly happen (e.g., because you copy between two buffers and allocated both at the same time and with the same size), Java still keeps the training wheels on. Every single byte copied will go through two array index bounds checks: once when reading it from the source array, once when writing it to the destination array.
On the other hand, unfortunately, this is precisely the stuff which makes Java nice in the first place.
E.g., sure, you could do much better performance-wise in C with malloc, than relying on Java's garbage collector. On the other hand, this also means that in Java you won't have to go through months of debugging memory leaks. (You can still get a memory leak in Java, but far less often.)
E.g., sure, you could do much better performance-wise in C by cleverly removing unneeded array size checks. But then you can and _will_ go to far and remove checks which are actually needed. See the ton of Windows and IE exploits based on buffer overflows. By contrast, Java's forcing you to keep the training wheels on, tends to produce far more robust code and far less debugging work.
Etc.
Basically Java does have a bunch of nice features, and it does trade some speed to get them.
So basically it's all about using an exploit, to bypass the game's normal rules. Well, gee. I suppose soon we'll see games or mods with integrated aimbot and wallhack, for those who are nostalgic about using _those_ exploits/cheats/whatever-you-want-to-call-them.