Open Source Worse than Flying

george writes "In an article published on TheRegister, Otto Z. Stern makes the bold statement that "The only thing as goat-rendering awful as flying has to be the progression of open source code." Accusing Open Source of being buggy and its devolopers of preoccupation with mudane details."I'm sitting here...wondering when the Linux freaks are going to solve their Ubuntu versus Mandriva color scheme debate or maybe even write a printer driver so that something I buy actually works with my open sores PC.""

Comments

Re:Linux will never progress very far

[Milo77]

Wow. I kinda think that this post proves the previous guy's point exactly. Specifically, where you say that you can configure KDE to basically look like OSX. If you think that a KDE theme is all you need to get the user experience of OSX you're just being silly. If it's "good enough" for you, then you are exactly the person too close to things to see how bad they suck. Further backing up the original guys post is the fact that you are modded so high. I am not sure there is an easy way to cure what appears to be an epidemic of "bad taste" among *nix users. I don't think there is a pill or anything :)

Re:Wow, what an ass

[Osty]

There really is no reason for every damn product on the market to need a custom driver though. There should be one interface for a printer, one for a camera, one for a video card, one for a joystick, one for a modem, etc. The consumer needs to demand this.

There already are such standards: printer = postscript, camera = FAT (filesystem for flash memory), video card = VESA, joystick = USB HID, modem = Hayes, etc. The problem is that these either cost way too much (postscript printers, real hardware modems) to be viable in the current consumer market (different from the business market, which is why you should have no problem using multi-thousand dollar "enterprise" printers but can't use your $50 inkjet), or they don't let you use the advanced functionality of the device (video card, joystick). In the first case, consumers aren't going to go back to paying $500 for a printer or $100 for a modem when they can get a $50 printer and $10 modem that work with the 90%+ majority OS. In the second case, while you may get your hardware working, you're going to bitch that you can't use higher resolutions at proper refresh rates or take advantage of all of that hardware acceleration in your $200 video card, or that you can only use two of the ten buttons on your joystick. There's simply no way to design a standard driver that will allow designers to continue to advance their product and still remain competitive (even "standards" like OpenGL allow for extensions, because if it didn't it would've been dead years ago).

Not only will it give us choices as to what OS and software we use with these products but it'll also make computers a lot more stable. A lot of crashes and other common problems are the result of minor incompatibilities between different drivers on the system. Standard drivers can be well tested. A mish mash of random drivers can't be tested well at all.

We'd also be stuck in the early 90s, technology-wise, because nobody could or would advance the state of the art. Standards are all well and good, but you have to be able to extend them for them to remain viable. Look at HTML for example -- the deliberate snubbing of standards by Microsoft and Netscape forced the standard to move forward. Yes, it resulted in crap like <blink> and <marquee>, and it caused a lot of compatibility pain (do you use iframes or layers? IE events or Netscape events?), but if that hadn't happened we'd still be stuck in the days of HTML 3.x, using tables for layout and not having anything close to CSS (or worse, we'd have Netscape's javascript-based style sheet language instead).

Standards are defined by committee, which the absolute worst way to innovate.