I hate discussions like this, since they almost always start at a reasonable place and then devolve into an all or nothing argument.
In terms of diversity of distros, it's very clear that this has helped Linux to date, in terms of technical advancement and user acceptance. No argument there. But in terms of commercial success and trying to dislodge Windows from the 92% of desktop OS shipments it supposedly owns (according to the latest numbers I've seen), this is a case of "diversity == confusion", since it just makes support, training, and app selection all the more difficult.
What do you think happens when a CIO/CTO type looks online and sees both distro and version specific RPMs, debs, and tar.gz files for the same app? He/she runs for the hills of Redmond, that's what happens.
...will there be a way to turn it off? I doubt it's perfect for all people and all conditions, particularly in this early stage.
And for that matter, when will there be a Linux distro with TT font installation that consists of nothing more than copying the font to a directory? Even running a custom applet to install fonts isn't always the right answer (think: programmatic operation).
Source for the quote?
on
Qt Going GPL
·
· Score: 1
Where did that quote from Eng come from? I'm not questioning its accuracy; I just want to find it.
This is terrific news, for all the reasons others have pointed out already, but mostly, IMO, because it will remove the political barriers to using Qt, and therefore KDE. Now we can have a real, honest-to-God, head-to-head competition between KDE (particularly KDE 2.0) and GNOME, and let the better DE win.
Personally, I always thought the licensing issue was blown way out of proportion, but now that it's history we can lay that to rest. (Even though I suspect some of the more ardent/.'ers will suffer withdrawal symptoms...)
How many times do people in the computer and communications have to learn the same lessons? It's bad enough that we have the never ending flurry of browser and other security problems, but we now also have to endure DoS attacks on phones?
I would love to see the publisher who would take on a book project, including fronting a reasonable advance to the authors, paying for printing, etc., on the basis of non-binding promises to buy. My hunch is it will happen about one day after Satan starts ice skating...
I disagree, at least in part. The current dynamic is that the desktop users are running Windows. How do you get them to switch to Linux? By convincing them that they're better off by doing so. And that means convincing them that there's a good cost/benefit ratio involved--the benefit of converting to and learning the new environment will outweigh the cost to make the conversion.
The only way that will happen is if Linux has not just applications, but the right set of applications, with the right set of features and capabilities. (Notice how prominently Sun mentioned that StarOffice 5.2 has improved MS Office document compatibility? Clearly Sun understands the importance of interoperability.) But another multiplier effect is usability. If you have a rock-solid OS with a huge assortment of exactly the kinds of apps users want, but its usability stinks, then users will stay away in droves. It just won't be worth the pain. Stable + highly usable + no apps yields the same result--zippo on the mainstream desktop.
Right now, Linux has a killer stability proposition, but it doesn't yet have the apps the mainstream users want (but it's getting much closer now), and it's still falling short on the usability front. There are still far too many places in setting up and using Linux where things that "just happen" under Windows or the Mac OS require more work and/or more knowledge on the user's part. Every situation like that is another barrier to entry for the mainstream users that the Linux camp should be working hard to eliminate.
I respectfully disagree with Chris' point about one of the DE's dying off. I expect both GNOME and KDE to evolve, possibly a lot, but neither is dependent right now on traditional business concerns, like making a profit, for their survival. If they were, then I'd agree 100%, but since this is open/free source, that economics sword of Damocles isn't haning over anyone's head. (e.g. How long has the Hurd project been trudging along, and how many people outside its developers use it?)
I do think the author of the original article had a point in that Linux should be focused on the higher level issues, like user interface guidelines, and all LInux programmers (including me) should be concentrating on the end user's experience with Linux and our apps. Linux has already largely conquered the server market, and is set to lay waste to the embedded space. But the toughest market of all is the desktop, since that's where things like user preferences and overall usability are far more important than purely technical areas.
Linux earned its success in the server segment and it will do so in the embedded space by proving it can carry the load reliably, cheaply, and flexibly. Now is has to prove it can be pretty, as well. That's a radically different challenge. Personally, I think Linux and the Linux programmers are up to the challenge. But in order for them to succeed, more of them will have to make pleasing the end user an explicit goal, and leave their guru-style prejudices and preferences behind.
Any sufficiently advanced technology...
on
2Ghz P4 Shown Off
·
· Score: 4
is indistinguishable from a rigged demo.
Not to imply Intel had rigged this one, but a single demo of a single P4@2GHz doesn't mean we'll that chip at that speed for sale any time soon.
I won't say all the things that others have already in reply to this one. I agree with them completely, and I'll just add one point: IBM has been, is, and likely will be for quite some time, the best friend Linux and open source have among the Big Companies.
It's on LinuxApps and freshmeat, and it's called verInfo. It lets you store all kinds of information in Linux executables (not data files), similar to what Windows version resources do, but better.
The argument isn't centralized vs. decentralized control of distribution. In that, Ms. Dyson and everyone else taking that view is mistaken, IMO.
What we're seeing now is radically different, in that the Internet and the current level of connectivity, as well as the sheer number of people using it, means that the barriers to distributing IP have been driven to essentially zero. Yes, in the old days PC clubs swapped disks of BASIC programs (and I probably have a few on a closet shelf, in fact), but those bits didn't travel as far, as fast, or get into as many hands as bits can and routinely do today.
We're seeing the IP equivalent of the "flash riots" the SF writer Larry Niven talked about decades ago, an analogy I think of every time I hear about some site "being slashdotted".
Every time I read about some new technology that gives us yet another opportunity to trade off a little proviacy for a little more airline security, I wonder how far this trend will go. I think I've finally got it figured, though: Flying naked. By 2020, all airlines will only allow completely naked passengers onboard, and even then only after they've been through the InstantMRI (tm) machine. No luggage, no laptops, no carry-ons of any type, just you and a few hundred other buck-naked people at 30,000 for a few hours. Talk about the friendly skies...
(I also wonder about the economic impact of that. Would ticket prices go down, since modest people wouldn't want to fly, or go up because some people would like being locked in a room full of naked strangers? Hmm.)
I wish I had a cut-and-dried answer. My hunch is that we'll see widely differing levels of convergence depending on which part of the market we look at. Stock traders need one set of functions, cab drivers another, the hardcore gizmonauts another, etc. And different groups will also have different abilities and desires to take on learning curves to master combination devices. Just look at the mainstream computer users today--I know many of them who 1) think Windows is too hard to use, 2) think Linux is a company that makes air conditioners, 3) don't want their cell phone to be anything other than a cell phone, and most important of all, 4) have a justifiable distrust of new technology.
(I suppose this could turn into a whole tangential discussion about how lousy tech products [insert your own list here] are slowly turning people in neo-Luddites, but maybe that's best left for another time.)
As for what's physically possible, that's clearly going to run head-long into what's usable by human beings. I'm very skeptical that WAP will catch on with more than a very tiny percentage of cell phone users, since there's nothing they'll want to do that you can do on such a limited screen. It might be possible to surf on a WAP, but will people really want to?
...then I suggest you try StarOffice 5.2 (the latest version). So isn't everyone's cup of Jolt, thanks to its silly insistence on using its own desktop in a window, but it's definitely a full-featured package, and it seems to support MS Office interoperability very well.
...or is 100% untraceable Internet communication the cyber equivalent of perpetual motion--it would be Very Cool to invent, everyone wants it for various reasons, but the nasty truth is that you just can't get there from here?
But it's also extremely annoying. If the companies running those sites want to feel the wrath of customers (and some of us out here in net-land can be very loud and annoying), then they should feel free to play their little games.
I think we have to draw a bright line between simply trying to get into another system, which is "bad", but is the computer equivalent of trespassing, and crackers who use other systems for DOS attacks, or try to damage the systems they break into. In the latter case it's theft and/or vandalism.
Check the information on the site, and you'll see that Kylix is going to use a new framework, called CLX ("clicks"), and not VCL. They say that a project that ook 6 months to develop with VCL will take a month to convert to CLX. This sounds like a long time, given how much of the "plumbing" that VCL takes off your hands, letting you concentrate on the guts of the program.
Slashdot Mirror
In terms of diversity of distros, it's very clear that this has helped Linux to date, in terms of technical advancement and user acceptance. No argument there. But in terms of commercial success and trying to dislodge Windows from the 92% of desktop OS shipments it supposedly owns (according to the latest numbers I've seen), this is a case of "diversity == confusion", since it just makes support, training, and app selection all the more difficult.
What do you think happens when a CIO/CTO type looks online and sees both distro and version specific RPMs, debs, and tar.gz files for the same app? He/she runs for the hills of Redmond, that's what happens.
And for that matter, when will there be a Linux distro with TT font installation that consists of nothing more than copying the font to a directory? Even running a custom applet to install fonts isn't always the right answer (think: programmatic operation).
Where did that quote from Eng come from? I'm not questioning its accuracy; I just want to find it.
Personally, I always thought the licensing issue was blown way out of proportion, but now that it's history we can lay that to rest. (Even though I suspect some of the more ardent /.'ers will suffer withdrawal symptoms...)
As Dr. Evil said, "Give me a frickin' break."
I would love to see the publisher who would take on a book project, including fronting a reasonable advance to the authors, paying for printing, etc., on the basis of non-binding promises to buy. My hunch is it will happen about one day after Satan starts ice skating...
The only way that will happen is if Linux has not just applications, but the right set of applications, with the right set of features and capabilities. (Notice how prominently Sun mentioned that StarOffice 5.2 has improved MS Office document compatibility? Clearly Sun understands the importance of interoperability.) But another multiplier effect is usability. If you have a rock-solid OS with a huge assortment of exactly the kinds of apps users want, but its usability stinks, then users will stay away in droves. It just won't be worth the pain. Stable + highly usable + no apps yields the same result--zippo on the mainstream desktop.
Right now, Linux has a killer stability proposition, but it doesn't yet have the apps the mainstream users want (but it's getting much closer now), and it's still falling short on the usability front. There are still far too many places in setting up and using Linux where things that "just happen" under Windows or the Mac OS require more work and/or more knowledge on the user's part. Every situation like that is another barrier to entry for the mainstream users that the Linux camp should be working hard to eliminate.
I do think the author of the original article had a point in that Linux should be focused on the higher level issues, like user interface guidelines, and all LInux programmers (including me) should be concentrating on the end user's experience with Linux and our apps. Linux has already largely conquered the server market, and is set to lay waste to the embedded space. But the toughest market of all is the desktop, since that's where things like user preferences and overall usability are far more important than purely technical areas.
Linux earned its success in the server segment and it will do so in the embedded space by proving it can carry the load reliably, cheaply, and flexibly. Now is has to prove it can be pretty, as well. That's a radically different challenge. Personally, I think Linux and the Linux programmers are up to the challenge. But in order for them to succeed, more of them will have to make pleasing the end user an explicit goal, and leave their guru-style prejudices and preferences behind.
Not to imply Intel had rigged this one, but a single demo of a single P4@2GHz doesn't mean we'll that chip at that speed for sale any time soon.
Perhaps a Borg-ified Tux? Seems appropriate, since it just assimilated one OS and has its sights on others...
I won't say all the things that others have already in reply to this one. I agree with them completely, and I'll just add one point: IBM has been, is, and likely will be for quite some time, the best friend Linux and open source have among the Big Companies.
It's on LinuxApps and freshmeat, and it's called verInfo. It lets you store all kinds of information in Linux executables (not data files), similar to what Windows version resources do, but better.
What we're seeing now is radically different, in that the Internet and the current level of connectivity, as well as the sheer number of people using it, means that the barriers to distributing IP have been driven to essentially zero. Yes, in the old days PC clubs swapped disks of BASIC programs (and I probably have a few on a closet shelf, in fact), but those bits didn't travel as far, as fast, or get into as many hands as bits can and routinely do today.
We're seeing the IP equivalent of the "flash riots" the SF writer Larry Niven talked about decades ago, an analogy I think of every time I hear about some site "being slashdotted".
(I also wonder about the economic impact of that. Would ticket prices go down, since modest people wouldn't want to fly, or go up because some people would like being locked in a room full of naked strangers? Hmm.)
(I suppose this could turn into a whole tangential discussion about how lousy tech products [insert your own list here] are slowly turning people in neo-Luddites, but maybe that's best left for another time.)
As for what's physically possible, that's clearly going to run head-long into what's usable by human beings. I'm very skeptical that WAP will catch on with more than a very tiny percentage of cell phone users, since there's nothing they'll want to do that you can do on such a limited screen. It might be possible to surf on a WAP, but will people really want to?
...then I suggest you try StarOffice 5.2 (the latest version). So isn't everyone's cup of Jolt, thanks to its silly insistence on using its own desktop in a window, but it's definitely a full-featured package, and it seems to support MS Office interoperability very well.
...or is 100% untraceable Internet communication the cyber equivalent of perpetual motion--it would be Very Cool to invent, everyone wants it for various reasons, but the nasty truth is that you just can't get there from here?
It could probably tell what you ate for breakfast, down the level of how much salt you put on your eggs.
But it's also extremely annoying. If the companies running those sites want to feel the wrath of customers (and some of us out here in net-land can be very loud and annoying), then they should feel free to play their little games.
I think we have to draw a bright line between simply trying to get into another system, which is "bad", but is the computer equivalent of trespassing, and crackers who use other systems for DOS attacks, or try to damage the systems they break into. In the latter case it's theft and/or vandalism.
Check the information on the site, and you'll see that Kylix is going to use a new framework, called CLX ("clicks"), and not VCL. They say that a project that ook 6 months to develop with VCL will take a month to convert to CLX. This sounds like a long time, given how much of the "plumbing" that VCL takes off your hands, letting you concentrate on the guts of the program.