They're a bunch of goofballs. It's the same thing as with GMO's except that they have even LESS scientific basis for alarm. I wrote a sarcastic letter to the editor a couple weeks ago, but I spose I was too rude, it didn't get published. Bottom line is: 1) Cell Phones are more dangerous; closer proximity, higher signal strength. 2) Are parents going to sue other parents who have wi-fi if their children visit? How about Starbucks, McDonald's, and several hospitals? 3) I can get a handful of scientists who claim that UFO's are real too, does that mean they really exist and I can sue the government successfully for covering up?
When are they going to apologize for requiring Internet Exploder 5.5 for Quicken 2004 to work? In Q2000 I could force it to use Netscape with 128 bit encryption, but no such luck with the new one. Of course by the time I figured this out the box was open, and there's no chance CompUSA is going to take a return....
Oh please. Problems where University admins can't keep their systems up (been there done that, have the scars to prove it, so don't give me grief about slamming University admins) are not the same thing as being slashdotted. I never said that Princeton never goes down, I said that the server isn't even remotely slashdotted, and right now is even up just fine period, so posting the article here is WHORING.
Sorry to break your bubble, but most of the time, large corporations have IT people to do stuff like this. It's not like Joy has the root password to the servers to go remove whatever email he wants. He doesn't, I don't (I work for Sun too), and while both of us are quite capable of managing that aspect, since it has nothing to do with what Sun pays us to do, we don't do it.
The kernel was re-written to be modular 15 years ago, which you'd know if you had read more than the first page of the article. As for the other layers of the OS, much of that is modular too; it's called "packages", quite similar to "rpms" that you might have heard of.
"Anytime Sun has a slip up, Merrill Lynch analyst Steven Milunovich comes running to make sure everyone knows how dire the situation really is. He puts on his cute, schoolgirl outfit, gets giddy and starts to ramble on about the end being near."
So tell us again how Sun is doomed? Just because this one guy keeps saying it doesn't make it true.
Non kernel layers are irrelevant TO MY POINT, not to users. Duh. It's open source, you can make the non-kernel layers identical without changing the kernel.
As for your installation concerns, I beg to differ. I've installed SuSE, Red Hat, and Solaris repeatedly, and the major difference between the Linuces and Solaris installation is that Solaris doesn't have all the pretty pictures, and it doesn't have a long involved post-installation configuration gui. If you think that gui is impossible to make for Solaris, you're deluding yourself, and if you're expecting to configure "a bunch" of any machine without making use of automated tools like jumpstart (not quite familiar enough with Linux to know what the equivalent is, but I'm sure someone has already scratched that itch) instead of some gui, you're doing it wrong.
And if you proceed to whine about how hard it is to set up jumpstart I'm just going to laugh at you. The first two times you do a jumpstart config it's confusing, but the time invested to figure it out pays off in the long run. Could jumpstart be better? Absolutely! But so could many things about most of the commercial Linux distro's too.
Seems to me that for Sun to abandon both SPARC and Solaris, two of three key technologies and basic compentencies, would be insane. While I'm dubious about the means of execution, the message that "the kernel you run doesn't !@#$ing matter" is a good one. I can make a Solaris box, for 85-90% of the things that most of any group of users needs to do, look just exactly like Linux. What does the Linux kernel have that Solaris doesn't? (non-kernel layers are irrelevant, because with sufficient work all that code can be ported, and I don't believe that "sufficient work" is a huge amount).
As for SPARC, that's an interesting concern, and I at least see some point to the argument that Intel won, despite not agreeing.
As others have pointed out, 1) it would be utterly stupid to run a company the way some sideline quarterback thinks it ought to be done. If he was so great at running companies, why isn't he running his own? And 2) it ain't over till it's over. I'm amazed at how many people here are talking like Sun is dead. As if such things haven't happened before. There are high barriers to overcoming the market position Sun is in, but they're hardly insurmountable.
Did you see the story about that, where the Tories sent out a press release saying exactly what's in my sig? They meant it as a joke, but it's pretty over the top....
I can say I've met my share as well. But it's unadulterated arrogance to say 1) "Most" don't get the current system and 2) My system will be so much easier (as long as you know Python, I guess).
So project goals have no relationship to design criteria? They're not, say, top-level criteria? If you don't meet a goal, doesn't that mean you haven't met a criterion?
Pick nits all you like, but I have to say that this in no way sounds like "good stuff". Given what a bastard mess gnome-config and trying to manage the billion daemons that make up gnome's state are, I cannot imagine that this project is going to do ANYTHING but 1) make a mess out of boot and 2) what it doesn't break will perform badly The only good news is that most people are likely to recognize this and ignore it..
I'm sorry, but he only includes that as a bone to those who he already realizes are going to resist what he wants to do. The fact that two of four primary design criteria mention the graphical desktop makes it clear that he has no idea of what init is doing, and is very gui-centric.
As far as "most" sysadmins not understanding run levels, he's out of his mind. Maybe he doesn't get it, but it's a long standing thing that works well. In fact, it works SO well, that Linux adopted it from System V after using the older monolithic rc scripts for a long while.
Provide a full services framework (including handling "boot up")
Integrate well with a desktop interface
Start X, and then allow login ASAP
Allow daemon binaries to directly contribute services rather than requiring each distro/vendor to write shell script wrappers
Items 2 & 3 are the killer. This is clearly a guy who thinks that the only reason to run Linux is to support an X environment, which is absolutely wrong.
Our university had variable speed vcr's in the library. This was a trivial hack. I can't see why it qualifies as "news" or "stuff that matters" in any sense.
I should have written this in the first place, but so I'm following myself up: You don't need a manual to use a hammer. Why should you need one to use a CD player or a video player?
By "use" I mean "get the basic functionality from". Obviously there will always be additional capabilities that require manual reading to become a serious user of them--even my DVD player required some reading to get to setup details etc. But I was able to use the DVD player to watch a DVD without having to read the manual! It doesn't sound like you can say that for xine/mplayer.
That's really what I think this thread's author started talking about.
Preeecisely. The KISS principle is critical to UI design; Larry Wall's dictum of "make easy things easy, and hard things possible" is also a good one, but you can't let that get in the way of KISS. A power user interface that is intimidating to everyone but the author and those who are masochistic enough to force themselves to think like the author is a bad interface.
I'm a power user of enough things to agree that there has to be somewhere to tweak all the tiny little knobs, but it shouldn't be the first page you see, and it should be reasonably easy to find nonetheless. From the sound of it, mplayer and xine, by requiring you to read the manual first, are not good UI designs, regardless of how well they're written or how well they meet their functional goals.
I don't mean to imply that a power interfaces is a bad interface, just that a bad interface is a bad interface, regardless of how "powerful" it appears to someone who designed it.
I have to say that unless I am using some tool that is mandated by work, if I have to spend more than about 5 to 10 minutes trying to figure out your user interface, I'm going to go find another solution to my problem. Web sites and web tools in particular are subject to this.
I do some web design for work, for people who *have* to use my tool to accomplish a particular task, and I have spent a lot of time thinking about how to make the tool work best for them, simply out of consideration. I hate it when work tools force me to twist my head around some horribly byzantine interface, and I don't want to do that to anyone else.
As a side note, _Don't Make Me Think_ by Steve(n?) Krug is one of the best introductions I've seen to the topic, and his coverage is quick and to the point. I'd be curious how the book reviewed here compares to it, as described by someone who's read both.
Slashdot Mirror
Wrong. They already got the money in a referendum last year.
It makes me ashamed that I live here sometimes.
Well, duh... You've got to know how to do something before you can automate it, right?
When are they going to apologize for requiring Internet Exploder 5.5 for Quicken 2004 to work? In Q2000 I could force it to use Netscape with 128 bit encryption, but no such luck with the new one. Of course by the time I figured this out the box was open, and there's no chance CompUSA is going to take a return....
how precisely does the 1x1 gif collect my email address for AT&T?
Wow. Can you read the minds of individuals as well as corporations?
Last time I checked, enterprises had desktops too.
Oh please. Problems where University admins can't keep their systems up (been there done that, have the scars to prove it, so don't give me grief about slamming University admins) are not the same thing as being slashdotted. I never said that Princeton never goes down, I said that the server isn't even remotely slashdotted, and right now is even up just fine period, so posting the article here is WHORING.
Where's my -1 Whoring moderation pick? I mean, really, their server is holding up JUST FREAKIN FINE.
Sorry to break your bubble, but most of the time, large corporations have IT people to do stuff like this. It's not like Joy has the root password to the servers to go remove whatever email he wants. He doesn't, I don't (I work for Sun too), and while both of us are quite capable of managing that aspect, since it has nothing to do with what Sun pays us to do, we don't do it.
The kernel was re-written to be modular 15 years ago, which you'd know if you had read more than the first page of the article. As for the other layers of the OS, much of that is modular too; it's called "packages", quite similar to "rpms" that you might have heard of.
"Anytime Sun has a slip up, Merrill Lynch analyst Steven Milunovich comes running to make sure everyone knows how dire the situation really is. He puts on his cute, schoolgirl outfit, gets giddy and starts to ramble on about the end being near."
So tell us again how Sun is doomed? Just because this one guy keeps saying it doesn't make it true.
As for your installation concerns, I beg to differ. I've installed SuSE, Red Hat, and Solaris repeatedly, and the major difference between the Linuces and Solaris installation is that Solaris doesn't have all the pretty pictures, and it doesn't have a long involved post-installation configuration gui. If you think that gui is impossible to make for Solaris, you're deluding yourself, and if you're expecting to configure "a bunch" of any machine without making use of automated tools like jumpstart (not quite familiar enough with Linux to know what the equivalent is, but I'm sure someone has already scratched that itch) instead of some gui, you're doing it wrong.
And if you proceed to whine about how hard it is to set up jumpstart I'm just going to laugh at you. The first two times you do a jumpstart config it's confusing, but the time invested to figure it out pays off in the long run. Could jumpstart be better? Absolutely! But so could many things about most of the commercial Linux distro's too.
As for SPARC, that's an interesting concern, and I at least see some point to the argument that Intel won, despite not agreeing.
As others have pointed out, 1) it would be utterly stupid to run a company the way some sideline quarterback thinks it ought to be done. If he was so great at running companies, why isn't he running his own? And 2) it ain't over till it's over. I'm amazed at how many people here are talking like Sun is dead. As if such things haven't happened before. There are high barriers to overcoming the market position Sun is in, but they're hardly insurmountable.
Did you see the story about that, where the Tories sent out a press release saying exactly what's in my sig? They meant it as a joke, but it's pretty over the top....
I can say I've met my share as well. But it's unadulterated arrogance to say 1) "Most" don't get the current system and 2) My system will be so much easier (as long as you know Python, I guess).
Pick nits all you like, but I have to say that this in no way sounds like "good stuff". Given what a bastard mess gnome-config and trying to manage the billion daemons that make up gnome's state are, I cannot imagine that this project is going to do ANYTHING but 1) make a mess out of boot and 2) what it doesn't break will perform badly The only good news is that most people are likely to recognize this and ignore it..
I'm sorry, but he only includes that as a bone to those who he already realizes are going to resist what he wants to do. The fact that two of four primary design criteria mention the graphical desktop makes it clear that he has no idea of what init is doing, and is very gui-centric.
As far as "most" sysadmins not understanding run levels, he's out of his mind. Maybe he doesn't get it, but it's a long standing thing that works well. In fact, it works SO well, that Linux adopted it from System V after using the older monolithic rc scripts for a long while.
SystemServices has four major goals:
Items 2 & 3 are the killer. This is clearly a guy who thinks that the only reason to run Linux is to support an X environment, which is absolutely wrong.
Our university had variable speed vcr's in the library. This was a trivial hack. I can't see why it qualifies as "news" or "stuff that matters" in any sense.
By "use" I mean "get the basic functionality from". Obviously there will always be additional capabilities that require manual reading to become a serious user of them--even my DVD player required some reading to get to setup details etc. But I was able to use the DVD player to watch a DVD without having to read the manual! It doesn't sound like you can say that for xine/mplayer.
Preeecisely. The KISS principle is critical to UI design; Larry Wall's dictum of "make easy things easy, and hard things possible" is also a good one, but you can't let that get in the way of KISS. A power user interface that is intimidating to everyone but the author and those who are masochistic enough to force themselves to think like the author is a bad interface.
I'm a power user of enough things to agree that there has to be somewhere to tweak all the tiny little knobs, but it shouldn't be the first page you see, and it should be reasonably easy to find nonetheless. From the sound of it, mplayer and xine, by requiring you to read the manual first, are not good UI designs, regardless of how well they're written or how well they meet their functional goals.
I don't mean to imply that a power interfaces is a bad interface, just that a bad interface is a bad interface, regardless of how "powerful" it appears to someone who designed it.
I do some web design for work, for people who *have* to use my tool to accomplish a particular task, and I have spent a lot of time thinking about how to make the tool work best for them, simply out of consideration. I hate it when work tools force me to twist my head around some horribly byzantine interface, and I don't want to do that to anyone else.
As a side note, _Don't Make Me Think_ by Steve(n?) Krug is one of the best introductions I've seen to the topic, and his coverage is quick and to the point. I'd be curious how the book reviewed here compares to it, as described by someone who's read both.