Why Linux on the Desktop is Wrong
Too many Linux enthusiasts continue to tout, "this is the year of Linux on the desktop." I understand that Google's Chrome OS as well as other announcements make it seem like a pending reality. I am quite hopeful myself. |
|
|
Comments (7)
OSS has a handicap. It gets produced to do what who ever started the project wants, to some extent, locked into interface designs that make no sense, then gets added to, not by people trying to make it easier, but people trying to make it do "more", without necessarily ever doing one thing very very well. And, even when it does, half the time the documentation is the last thing ever bothered with, so you get idiot things like, "Offset - holds the offset rotation.", but with a) no example of how to use it, b) no idea how to set it, etc. There is an entire page in the OSS Imprudence client for OpenSim/Second Life, which contains a list of math functions like that, which you can use in setting values, and ***one*** simplistic example of how to use like 2 of them, which fails to address the most common reason for using them at all, which is things like setting all three values at once, base on an *existing* rotation. (The client uses Euler to show what its set to, but the client uses Quaternion internally, which makes a damn mess, because something like ends up as , when you are done, but ***only*** if you have some way to take and *multiply* it with . This has to be done to "all" rotations at once, in principle, but there is nothing in the examples that tells you how the hell to do that. Euler, by itself, gimble locks, which means setting a rotation never come out "right". The Offset is worse, since there is no way to even enter that, so if you can't enter it, how the client knows what it is, when you use it... And so on.)
If you are talking about OSS in terms of shifting people from Office to OpenOffice, or MS Outlook to Firebird, sure.. The rest of us have to work for a damn living, and that means having usable help files, sane documentation, and not having to bloody guess what the hell a parameter actually does, or how to use it, especially when its the feature we *need*. OSS, in this respect, often seriously pisses me off. If its not hidden some place that makes no sense to me, its not documented properly. As bad as MS is, except for something moving things for no damn reason, they, and most of the proprietary people, know that people need to be able to figure out how to use the damn program, not just the 4 people that came up with it on sourceforge, and anyone willing to spend weeks asking each other questions about, "Ok, so what they hell does XGWET do exactly?", never mind the new guy that has to hunt through 4 years of forums posts, error reports, and people going, "I haven't a clue.", before finding the post that explains it.
I would love to keep some of the proprietary I have. Sadly, its shackled, using unobtanium and krpytonite, to MS' wrist. And that is a much bigger problem than, "Not enough people use OSS."
Depending on which "next guy" they're talking about, that isn't saying much.
Quoting:Problem is.. OSS doesn't always have the same incentives as the customer. I use my machine about 80% of the time for games....+ rant, rant, rant
Perhaps another type of system that would better suit this customer here is Mac OS X running on one of Apple's high-end product lines. Not OSS and certainly proprietary, but yet quite effective.
Quoting:The rest of us have to work for a damn living, and that means having usable help files, sane documentation, and not having to bloody guess what the hell a parameter actually does, or how to use it, especially when its the feature we *need*. OSS, in this respect, often seriously pisses me off....+ rant, rant, rant
And yet another type of system that would perhaps better suit this particular customer is one of the BSD OS's -- e.g., FreeBSD -- running on standard commodity hardware. Plusses: - a VERY usable help system - "sane documentation" - other plusses This customer and similar readers might do well to visit Greg Lehey's 'Explaining BSD' found at http://www.freebsd.org/doc/en/articles/explaining-bsd/ This could be a good place to start off, given the concerns this customer so vehemently expresses.
Quoting: Problem is.. OSS doesn't always have the same incentives as the customer. I use my machine about 80% of the time for games.
I use my machine(s) about 0% of the time for (the usual conception of) games. If you've not going to run games, and you want to figure things out, then you must have usable documentation, customizable interfaces, efficient interfaces, secure computing, stable APIs (not at the whim of marketing or CxOs) and occasionally you have to look at the source code.
There's also this little matter of speed being part of usability. The Windows box I have to use at work gradually slowed under the weight of the "security" features to the point of un-usability. It's replacement is on the same path.
And all in the name of extracting what the market will bear from the consumer.
That depends upon their not having done this deliberately the entire time.
Post a comment
A free account is needed to post. Write your comment here first: your text is kept in this browser while you sign up or log in, and is waiting for you when you come back.
Your message:Sign up and post (it is free) or Log in and post

A better pull quote would have been:
Now that's more like it! No special pleading for MSFT, no dragging out oft-refuted urban linux legends, no "Now I like Linux as much as the next guy." Just plain old emergent behavior from corporations that don't have the same incentives as the consumers.