This points out the fact that when criminals (or potential criminals) can get their hands on source code they'll sift through it looking for exploits.
Of course they will. All the more reason to accept peer reviewers, as they do the same thing. Why should criminals have an advantage?
Unless a piece of software is released under an OpenSource(tm) license, and mechanisms are in place for peer-review to result in rapid fixes (i.e. there is a body accepting open submissions, etc.) the public release of the source code DOES represent a security risk.
What you say here is not quite accurate. The software does not have to be released under and open source license to retain security, and there need be no body to accept (code) submission. At minimum, we would like:
Availability of the source code for perusal. Redistribution can be restricted. We just need to see it.
People willing to listen to suggestions. They do not have to accept code. Any sort of bug-tracking system would provide this.
It's sort of an all-or-nothing situation.
Not at all. What I've outlined above is clearly not Open Source, but it can improve security.
Regardless, the silliness of Novell's statement is that they imply security through obscurity is inherently better than open peer review, which has been proven time and again to be false.
While it's true that executing both paths following a branch means work will be thrown away, predication relies on the assumption that wasting some execution units to execute unnecessary code will be faster than executing into one path, finding out it is wrong, flushing the pipeline and starting all over back at the branch. The idea is to predicate hard-to-predict branches.
Researchers have also toyed with the idea of dual execution, which is a similar idea, only it is done dynamically by the hardware. A simultaneous multithreading machine might do this, for example.
Predication can also be used to eliminate the loop prologue and epilogue when performing software pipelining, resulting in smaller code size.
The slides at the site give some good examples of how to use predication.
Nowhere did I claim NATO is not using propaganda. I would expect Milosevic to target such sources, hopefully not with violence. Shutting down satellite links is the perfect way to cut off propaganda without the destruction of property or killing of people. It seems entirely reasonable to me.
CNN, etc. are often used as propaganda weapons against NATO. Reporters are allowed to see only what Milosevic wants them to see.
Did I say anything about blowing up a roomfull of students? My goodness, please read the article summary before you post. Apparently, satellite links were shut down. AFAIK, nothing more (I couldn't find the reference on the site.
No, I don't believe everything in U.S. News. But Beograd.com seems to confirm the report.
Of course NATO uses propaganda. Every side in a war does. Each side should also work to degrade the ability of the other to distribute such materials.
This action makes good sense to me. When you're fighting a war, you limit the enemy's ability to communicate and distribute propaganda.
A recent U.S. News had a short article about Yugoslavia's electronic propaganda war. They literally have an entire room full of computers run by people (mostly students, IIRC) solely for the use of sending propaganda over the internet (and they've been doing a much more effective job than NATO!). Far from the bulwark of free and independent news, the internet is being used a a tool to further Milosevic's own political goals.
The Linux team could set up clusters of workstations, etc., but i don't think that gets us any useful information. The more I think about it, the more I like the idea that the hardware may favor NT. If it does, it indicates a problem with Linux and we need to fix it.
Why doesn't samba work well with Win95? These are the questions we need to answer.
There are lots of tradeoffs like this you can make in an experiment. The Mindcraft setup is one choice. What you have suggested is another.
Who will pay for it? Are the various commercial Linux vendors willing to pool resources and commission Mindcraft or some other third party to run the tests?
All of these tests should be run, not to prove anything about Linux vs. NT, but to make Linux better. That should be our goal. NT should be an afterthought.
You have to be very, very careful doing this, however, because you're introducing another variable into the experiment. Better to allow both parties to come to an agreement on a hardware setup.
On the other hand, there is some merit to letting a third party pick the hardware. In this case, we will have the opportunity of finding problems with drivers rather than simply using things we already know work well.
If poorly supported hardware is used, however, it must be made absolutely crystal clear in the press release. Nothing less is acceptable.
If the Linux community is serious about doing this, it's going to take a lot of planning. Just because you may not be taking the trip doesn't mean you can't contribute to the effort. After all, the community has always been trumpeted as a strength of Linux. We ought to set up a web site, mailing list or something to coordinate all of this.
It seems pretty clear to me now that Mindcraft really does want to produce unbiased results. This statement is quite telling:
Mindcraft has withheld the publication of our second Linux and Windows NT Server benchmark results (the test for which Linus and others provided some suggestions for tuning) pending the response to this Open Benchmark invitation.
Looks like Linux did not come out on top in their second test, and Mindcraft is willing to bring in Linux experts to verify the results.
My biggest fear with this is not that Mindcraft will try some underhanded trickery, it's that the Linux community will not accept the results unless linux comes out on top. That is just as bad as if Mindcraft tailored its benchmarks to favor NT.
This open benchmark is a wonderful opportunity for the Linux community to benchmark itself. AFAIK, no controlled experiments on high-end servers have been done to see how Linux stacks up against the heavyweights. If Linux comes out on top, great. But if not, we as a community have to accept that and learn from the process.
This is best chance yet to discover the bottlenecks in the kernel and several critical pieces of software. IMHO, the Linux experts should not go in with the goal of beating NT. They should go in with the goal of squeezing every last bit of performance out of the machine and using the resulting data to fix the problems.
Mindcraft and Microsoft (!) are donating resources to the Linux community in an effort to help us improve the OS. Let's grab the opportunity!
In order to become a licensed architect, one usually needs an Masters of Architecture degree from an acredited school, and several years of internship, working for a licensed architect.
Such a system seems reasonable in the field of software engineering. No one is precluded from working just because they don't have a degree. But at the same time, those who are held responsible, are the ones who have completed graduate coursework in the field.
This seems like a quite reasonable approach to things, if people feel it is necessary to license software developers. I still disagree with the need, however.
(Note: there is an inherit assumption in all this that writing code is, in fact, an engineering science and not an art form.)
I don't agree. All engineering is art. All engineering is science. That's why it's engineering!
Engineering is the art of using science to pick a sub-optimal solution for the sake of politics (to please everyone).;)
It seems to me we already have a certification process: a degree from a University, community college, etc.
Now some may object that this is isn't good enough. Well, if that's true, I'd have to ask, "Why?" Let's make it good enough. A degree in the field of computer science, computer engineering or electrical engineering should say something. If it doesn't, it needs to be fixed.
If companies are going to hire "local music dropouts," that's their problem. Don't go blaming the engineers who have an advanced degree in the field.
Furthermore, the pushing out of the older generation of programmers is only going to make things worse. IMHO, these are the people you need the most since they have the experience of getting products out the door (on time) and have a wealth of practical experience to draw on. This needs to be passed on to the younger crowd.
I've always found the whole idea of licensed programmers absurd. A combination of advanced education and on-the-job mentoring and experience is much more effective than passing a silly exam. Corporations should be responsible enough to conduct regular reviews of their employees and address any problems that crop up.
Re:Importance of a good compiler on IA-64
on
Gcc for the IA-64.
·
· Score: 1
What I was trying to convey is that not only is Digital/Compaq not committing to use IA-64, they are also not developing their own VLIW. They did research into it back in the early days of Alpha (after 21064 IIRC).
Draw your own conclusions.
Re:Importance of a good compiler on IA-64
on
Gcc for the IA-64.
·
· Score: 3
For RISC vendors, they are used to dealing with extremely complex compilers.
How complex is "extremely complex?" Nowhere near the level of IA-64, I would guess. I found it odd that Taschek would bash HP for their compiler support. I think they have the best chance of producing a decent VLIW-style compiler, given all the research that has come out of HP labs.
What's interesting is that Digital/Compaq hasn't jumped on the Merced bandwagon, nor have they announced a VLIW architecture of their own. Hmm...
Having GCC working well on IA-64 will go a long way to further the impact of Linux, especially if GCC works much better than any compiler out of Redmond. Even if Windows 2000 runs on IA-64, if the applications crawl, not too many people are going to choose that OS over another.
I disagree with this and the meat of Taschek's article. If businesses choose Windows, they will choose Windows. I don't think a compiler will make or break an operating system. Established use and ingrained practices are much larger factors. If a good compiler targeting Windows on IA-64 is not available, Merced will be relegated to a niche market. If the free software community is counting on Merced being the end of proprietary operating systems, we've lost.
It's also interesting that people assume Microsoft will be providing the IA-64 compiler for Windows. Intel is producing their own compiler. They're having problems, but chances are there will be more than one IA-64 compiler for Windows.
But that still doesn't answer the quesion. Why haven't the offers been made? It costs nothing for Corel to make an offer that isn't taken up. In fact, if they aren't taken up, it's more fodder for the DOJ, which would also benefit Corel.
If Corel could break the MS establishment in the U.S., it would, I think, result in far greater benefits.
I'm not trying to sound like an isolationist here, but why South Africa? I understand the offer came as part of an interview, but why hasn't Corel offered to supply, say, the Chicago school district with lots of goodies? Seems to me the U.S. would be a rather large and important market for Corel.
What about England, Germany or Russia? Corel has made offers to Mexico. Why not Canada or Japan?
Moreover, the IBM ESA/390 (G5) executes complex instructions in "millicode." Opcodes are mapped to addresses containing the millicode specification of the instruction. But the kicker is, every opcode is assigned an address. It's just that most of them are empty and the instruction executes natively. It's possible to upgrade the millicode on-the-fly (while the machine is running) to patch bugs. You could think of using this to completely re-map the instruction set.
This information is taken from a presentation on the G5, so I don't have the complete story. Please post corrections if necessary.
This has been possible for some time (possibly as far back as the 360, but I'm not sure). Transmeta is certainly not the first.
Seriously, what instabilities do you see in Red Hat? I've been using it a while, and I currently run RH 5.2 with kernel 2.2.5. I have never had a problem.
You're quite lucky, then. I installed RH 5.1 on two machines. The blasted control-panel/linuxconf system simply does not work. For example, I try to fire up ppp with linuxconf and I get loads of disk activity (probably dumping core).
Configuration is a nightmare for anyone remotely familiar with the "traditional" method of system administration (i.e. editing text files). Because the configuration files are auto-generated with invocations of other scripts (sometimes several layers deep), it's tough to wrap your mind around what is supposed to happen. But perhaps that's just me and I'm not "with it." It wouldn't be the first time.:)
Compiling everything by hand gets old fast, especially on my system [p166/48mb].
No argument there. Debian, IMHO, has a much better package system that Red Hat. apt is the main reason.
BTW: I've tried Debian. In fact, its currently on my secondary system [486/66, 64mb]. However, I still haven't gotten it fully configured the way I like it. Stability, yes, but I still have problems getting it to do what I want it to. Its always fun when dselect decides to skip over half the packages you told it to install.
I've had similar problems with dselect, but I attribute that to my lack of reading on the subject.:) I think it's fair to say that most people agree dselect is not exactly the package interface you want to have. I'm not claiming Debian is the be-all end-all Linux distribution. I said it was the best (always IMHO) in terms of features and stability.
Seriously, the Linux community seems to have split into two groups: people who acknowledge that Red Hat has done a lot of good for the community, and those who constantly throw FUD at Red Hat.
It should not be US fighting each other over who has the better distro. I've yet to use a distro that was significantly better or worse than another.
I am not throwing FUD. The Linux community in general and the Slashdot community in particular (I'm including the *BSD, Be, etc. folks here as well) are going to have to face the fact that their favorite system is not perfect. There are problems, and those problems need to be faced head-on. It does no good to cry "FUD" as a defense against criticism. Call what is FUD FUD. Call what is constructive criticism the road to better software.
I don't recall saying that Red Hat has not done good things. They've done a lot. And I would never criticize them for their status as a successful corporation. But one must not get complacent. We can only pat ourselves on the back for so long.
Hmm...the sad thing is, Red Hat really doesn't provide Q/A. At least they haven't proven it yet.
It's rather ironic that the distribution that seems to provide the highest level of quality (in terms of features and stability) is Debian: a completely free distribution put together by volunteers.
Exactly what is the random box supposed to be? It looks like a fortune, but that's rather disappointing. I get one of those at the bottom of each page now.
I was hoping for a box that would show a random box item. That is, randomly select from the box list and put it on the page. That way I can get glimpses of all the available boxes without having to go and enable or test each one individually.
The install docs don't mention anything about this. It's something that should be in there, though.
That's one other thing I forgot to mention about Debian: lack of documentation. Not for the software (there's lots in/usr/doc), but for the newbie, it's a little confusing. The install manual is pretty bare-bones. And to be fair, it looks like the Debian Website is getting some docs together (though most are in the early stages of development).
Oh, and one other thing: kernel-package is pretty neat. It lets you compile kernel source, make a kernel image package and install it all with just a few commands. It even updates LILO for you. This way you can keep up with new kernels while still maintaining a sane package environment. And with multiple image packages, switch back and forth is a breeze.
As a recent Debian convert, let me see if I can describe my experience.
Previously I used Red Hat 5.1-5.2 The nice thing about Red Hat is that it's easy to install.
But Debian is not much more difficult.
The problem I ran into with Red Hat is that the blasted config files are all auto-generated and any effort to hand-edit them is either be wasted (because they get overwritten at boot-up) or results in a broken system because various tools rely on a strict format for the files.
Now perhaps I just did not follow all of the correct procedures, but having come from Slackware (and originally *cough* SLS), I'm used to rolling my own. Often I just don't want to be bothered with linuxconf or the [expletive] control-panel.
I'd describe Debian as Slackware with really, really nice packages. Packages include scripts that get run after installation to configure everything. There's a down side to this, however. You can't just start the install and walk away, because some scripts are interactive and ask you questions. Also, when I installed, the machine must have compiled the emacs math elisp code about 20 times. And this is not a quick process. I'm still not sure why deselect felt it should re-install elisp code several times.
That said, the package manager (dselect) is not as horrible as is generally thought around these parts. Yes, it takes a bit of getting used to, but the on-line help is really fabulous. I did not need to read the "printed" documentation to do simple things like select and install packages. Apparently a new and improved package manager is on the way.
But the truly great thing about Debian is apt. Apt lets you completely update your system with two simple commands:
% apt-get update % apt-get upgrade
The first updates the list of available packages. Apt knows how to interface through ftp and http so you can scan packages from lots of remote places. The second command pulls down any new versions of packages you have installed. When a bugfix for Debian is announced, I run these two commands and my system is brought completely up-to-date with the latest rev. Apt is smart enough to resolve all package dependencies and also bring over any new packages that are necessary.
The downside to apt is that it relies on the package source location to have directories organized just like they are on the distribution CD. This isn't a problem at debian.org, but installing packages from most other places means you have to ftp and install "manually" (ala RPM). But I figure it's a good sign when this is one of the few "complaints" I have.
Now Red Hat's rpm can also install or upgrade over the net, but is there a tool for updating the package list (over the net) like apt? Perhaps someone can enlighten me. I find it hard to believe it could be any simpler than apt.
One of the few problems I have with Debian is that it starts every bloody daemon in existence by default. I still haven't had the time to go through and shut them all off. Also, my network setup is a little hosed, but I think that's my fault.
I've been very pleased with Debian so far. It's much less buggy than Red Hat (imagine that, quality assurance really does work!), upgrades are a breeze and it just feels rock-solid.
GNUPro is still Open Source. Users pay for Cygnus support, NOT the tools. The tools can be had from ftp.cygnus.com.
What version of gcc is this based on? I remember working with GNUPro a couple of years ago. It was great because all of the most recent C++ improvements were incorporated into the tools LONG before the official gcc team was able to make a release. It let me get my work done. This was pre-egcs and was one of the triggers for the egcs project.
> Outside of programming classes, I could see having maybe ONE writing class.
The more, the better. You would not believe some of the writing I've seen from advanced undergrads or graduates. Expressing yourself clearly and professionally is one of the most important skills to have in any field.
> Otherwise, the Humanities, Management, Statistics, and a few others were just filler.
Stats? I wish I'd had more classes in that area!
It's important to be well-rounded. I took a music class my senior year. Even though I was already heavily into the piano and trumpet, it opened up a whole new world. I can't imagine what I'd be doing without it.
I've gained far too much enjoyment from literature, music and the arts to just throw it all away.
If you go read "The Soul of a New Machine," you'll find an interesting discussion of the job interviews for the Data General Eagle team. One of the questions asked concerned hobbies outside of work. It was very important to the team leaders that their workers do something outside the computer world when away from the lab.
If this is all too wishy-washy and non-geeky, then so be it. I'm happy.
Slashdot Mirror
Of course they will. All the more reason to accept peer reviewers, as they do the same thing. Why should criminals have an advantage?
Unless a piece of software is released under an OpenSource(tm) license, and mechanisms are in place for peer-review to result in rapid fixes (i.e. there is a body accepting open submissions, etc.) the public release of the source code DOES represent a security risk.
What you say here is not quite accurate. The software does not have to be released under and open source license to retain security, and there need be no body to accept (code) submission. At minimum, we would like:
It's sort of an all-or-nothing situation.
Not at all. What I've outlined above is clearly not Open Source, but it can improve security.
Regardless, the silliness of Novell's statement is that they imply security through obscurity is inherently better than open peer review, which has been proven time and again to be false.
And as a student here pointed out, you can compile some of your routines in IA32 as a poor-man's program compression mechanism. ;-)
Researchers have also toyed with the idea of dual execution, which is a similar idea, only it is done dynamically by the hardware. A simultaneous multithreading machine might do this, for example.
Predication can also be used to eliminate the loop prologue and epilogue when performing software pipelining, resulting in smaller code size.
The slides at the site give some good examples of how to use predication.
CNN, etc. are often used as propaganda weapons against NATO. Reporters are allowed to see only what Milosevic wants them to see.
No, I don't believe everything in U.S. News. But Beograd.com seems to confirm the report.
Of course NATO uses propaganda. Every side in a war does. Each side should also work to degrade the ability of the other to distribute such materials.
A recent U.S. News had a short article about Yugoslavia's electronic propaganda war. They literally have an entire room full of computers run by people (mostly students, IIRC) solely for the use of sending propaganda over the internet (and they've been doing a much more effective job than NATO!). Far from the bulwark of free and independent news, the internet is being used a a tool to further Milosevic's own political goals.
Right on.
The Linux team could set up clusters of workstations, etc., but i don't think that gets us any useful information. The more I think about it, the more I like the idea that the hardware may favor NT. If it does, it indicates a problem with Linux and we need to fix it.
Why doesn't samba work well with Win95? These are the questions we need to answer.
Who will pay for it? Are the various commercial Linux vendors willing to pool resources and commission Mindcraft or some other third party to run the tests?
All of these tests should be run, not to prove anything about Linux vs. NT, but to make Linux better. That should be our goal. NT should be an afterthought.
On the other hand, there is some merit to letting a third party pick the hardware. In this case, we will have the opportunity of finding problems with drivers rather than simply using things we already know work well.
If poorly supported hardware is used, however, it must be made absolutely crystal clear in the press release. Nothing less is acceptable.
If the Linux community is serious about doing this, it's going to take a lot of planning. Just because you may not be taking the trip doesn't mean you can't contribute to the effort. After all, the community has always been trumpeted as a strength of Linux. We ought to set up a web site, mailing list or something to coordinate all of this.
To those who make the trip: take good notes!
Looks like Linux did not come out on top in their second test, and Mindcraft is willing to bring in Linux experts to verify the results.
My biggest fear with this is not that Mindcraft will try some underhanded trickery, it's that the Linux community will not accept the results unless linux comes out on top. That is just as bad as if Mindcraft tailored its benchmarks to favor NT.
This open benchmark is a wonderful opportunity for the Linux community to benchmark itself. AFAIK, no controlled experiments on high-end servers have been done to see how Linux stacks up against the heavyweights. If Linux comes out on top, great. But if not, we as a community have to accept that and learn from the process.
This is best chance yet to discover the bottlenecks in the kernel and several critical pieces of software. IMHO, the Linux experts should not go in with the goal of beating NT. They should go in with the goal of squeezing every last bit of performance out of the machine and using the resulting data to fix the problems.
Mindcraft and Microsoft (!) are donating resources to the Linux community in an effort to help us improve the OS. Let's grab the opportunity!
Such a system seems reasonable in the field of software engineering. No one is precluded from working just because they don't have a degree. But at the same time, those who are held responsible, are the ones who have completed graduate coursework in the field.
This seems like a quite reasonable approach to things, if people feel it is necessary to license software developers. I still disagree with the need, however.
(Note: there is an inherit assumption in all this that writing code is, in fact, an engineering science and not an art form.)
I don't agree. All engineering is art. All engineering is science. That's why it's engineering!
Engineering is the art of using science to pick a sub-optimal solution for the sake of politics (to please everyone). ;)
Now some may object that this is isn't good enough. Well, if that's true, I'd have to ask, "Why?" Let's make it good enough. A degree in the field of computer science, computer engineering or electrical engineering should say something. If it doesn't, it needs to be fixed.
If companies are going to hire "local music dropouts," that's their problem. Don't go blaming the engineers who have an advanced degree in the field.
Furthermore, the pushing out of the older generation of programmers is only going to make things worse. IMHO, these are the people you need the most since they have the experience of getting products out the door (on time) and have a wealth of practical experience to draw on. This needs to be passed on to the younger crowd.
I've always found the whole idea of licensed programmers absurd. A combination of advanced education and on-the-job mentoring and experience is much more effective than passing a silly exam. Corporations should be responsible enough to conduct regular reviews of their employees and address any problems that crop up.
Draw your own conclusions.
How complex is "extremely complex?" Nowhere near the level of IA-64, I would guess. I found it odd that Taschek would bash HP for their compiler support. I think they have the best chance of producing a decent VLIW-style compiler, given all the research that has come out of HP labs.
What's interesting is that Digital/Compaq hasn't jumped on the Merced bandwagon, nor have they announced a VLIW architecture of their own. Hmm...
Having GCC working well on IA-64 will go a long way to further the impact of Linux, especially if GCC works much better than any compiler out of Redmond. Even if Windows 2000 runs on IA-64, if the applications crawl, not too many people are going to choose that OS over another.
I disagree with this and the meat of Taschek's article. If businesses choose Windows, they will choose Windows. I don't think a compiler will make or break an operating system. Established use and ingrained practices are much larger factors. If a good compiler targeting Windows on IA-64 is not available, Merced will be relegated to a niche market. If the free software community is counting on Merced being the end of proprietary operating systems, we've lost.
It's also interesting that people assume Microsoft will be providing the IA-64 compiler for Windows. Intel is producing their own compiler. They're having problems, but chances are there will be more than one IA-64 compiler for Windows.
If Corel could break the MS establishment in the U.S., it would, I think, result in far greater benefits.
What about England, Germany or Russia? Corel has made offers to Mexico. Why not Canada or Japan?
I'm just curious.
This information is taken from a presentation on the G5, so I don't have the complete story. Please post corrections if necessary.
This has been possible for some time (possibly as far back as the 360, but I'm not sure). Transmeta is certainly not the first.
You're quite lucky, then. I installed RH 5.1 on two machines. The blasted control-panel/linuxconf system simply does not work. For example, I try to fire up ppp with linuxconf and I get loads of disk activity (probably dumping core).
Configuration is a nightmare for anyone remotely familiar with the "traditional" method of system administration (i.e. editing text files). Because the configuration files are auto-generated with invocations of other scripts (sometimes several layers deep), it's tough to wrap your mind around what is supposed to happen. But perhaps that's just me and I'm not "with it." It wouldn't be the first time. :)
Compiling everything by hand gets old fast, especially on my system [p166/48mb].
No argument there. Debian, IMHO, has a much better package system that Red Hat. apt is the main reason.
BTW: I've tried Debian. In fact, its currently on my secondary system [486/66, 64mb]. However, I still haven't gotten it fully configured the way I like it. Stability, yes, but I still have problems getting it to do what I want it to. Its always fun when dselect decides to skip over half the packages you told it to install.
I've had similar problems with dselect, but I attribute that to my lack of reading on the subject. :) I think it's fair to say that most people agree dselect is not exactly the package interface you want to have. I'm not claiming Debian is the be-all end-all Linux distribution. I said it was the best (always IMHO) in terms of features and stability.
Seriously, the Linux community seems to have split into two groups: people who acknowledge that Red Hat has done a lot of good for the community, and those who constantly throw FUD at Red Hat.
It should not be US fighting each other over who has the better distro. I've yet to use a distro that was significantly better or worse than another.
I am not throwing FUD. The Linux community in general and the Slashdot community in particular (I'm including the *BSD, Be, etc. folks here as well) are going to have to face the fact that their favorite system is not perfect. There are problems, and those problems need to be faced head-on. It does no good to cry "FUD" as a defense against criticism. Call what is FUD FUD. Call what is constructive criticism the road to better software.
I don't recall saying that Red Hat has not done good things. They've done a lot. And I would never criticize them for their status as a successful corporation. But one must not get complacent. We can only pat ourselves on the back for so long.
It's rather ironic that the distribution that seems to provide the highest level of quality (in terms of features and stability) is Debian: a completely free distribution put together by volunteers.
I was hoping for a box that would show a random box item. That is, randomly select from the box list and put it on the page. That way I can get glimpses of all the available boxes without having to go and enable or test each one individually.
That's one other thing I forgot to mention about Debian: lack of documentation. Not for the software (there's lots in /usr/doc), but for the newbie, it's a little confusing. The install manual is pretty bare-bones. And to be fair, it looks like the Debian Website is getting some docs together (though most are in the early stages of development).
Oh, and one other thing: kernel-package is pretty neat. It lets you compile kernel source, make a kernel image package and install it all with just a few commands. It even updates LILO for you. This way you can keep up with new kernels while still maintaining a sane package environment. And with multiple image packages, switch back and forth is a breeze.
Previously I used Red Hat 5.1-5.2 The nice thing about Red Hat is that it's easy to install.
But Debian is not much more difficult.
The problem I ran into with Red Hat is that the blasted config files are all auto-generated and any effort to hand-edit them is either be wasted (because they get overwritten at boot-up) or results in a broken system because various tools rely on a strict format for the files.
Now perhaps I just did not follow all of the correct procedures, but having come from Slackware (and originally *cough* SLS), I'm used to rolling my own. Often I just don't want to be bothered with linuxconf or the [expletive] control-panel.
I'd describe Debian as Slackware with really, really nice packages. Packages include scripts that get run after installation to configure everything. There's a down side to this, however. You can't just start the install and walk away, because some scripts are interactive and ask you questions. Also, when I installed, the machine must have compiled the emacs math elisp code about 20 times. And this is not a quick process. I'm still not sure why deselect felt it should re-install elisp code several times.
That said, the package manager (dselect) is not as horrible as is generally thought around these parts. Yes, it takes a bit of getting used to, but the on-line help is really fabulous. I did not need to read the "printed" documentation to do simple things like select and install packages. Apparently a new and improved package manager is on the way.
But the truly great thing about Debian is apt. Apt lets you completely update your system with two simple commands:
% apt-get update
% apt-get upgrade
The first updates the list of available packages. Apt knows how to interface through ftp and http so you can scan packages from lots of remote places. The second command pulls down any new versions of packages you have installed. When a bugfix for Debian is announced, I run these two commands and my system is brought completely up-to-date with the latest rev. Apt is smart enough to resolve all package dependencies and also bring over any new packages that are necessary.
The downside to apt is that it relies on the package source location to have directories organized just like they are on the distribution CD. This isn't a problem at debian.org, but installing packages from most other places means you have to ftp and install "manually" (ala RPM). But I figure it's a good sign when this is one of the few "complaints" I have.
Now Red Hat's rpm can also install or upgrade over the net, but is there a tool for updating the package list (over the net) like apt? Perhaps someone can enlighten me. I find it hard to believe it could be any simpler than apt.
One of the few problems I have with Debian is that it starts every bloody daemon in existence by default. I still haven't had the time to go through and shut them all off. Also, my network setup is a little hosed, but I think that's my fault.
I've been very pleased with Debian so far. It's much less buggy than Red Hat (imagine that, quality assurance really does work!), upgrades are a breeze and it just feels rock-solid.
GNUPro is still Open Source. Users pay for Cygnus support, NOT the tools. The tools can be had from ftp.cygnus.com.
What version of gcc is this based on? I remember working with GNUPro a couple of years ago. It was great because all of the most recent C++ improvements were incorporated into the tools LONG before the official gcc team was able to make a release. It let me get my work done. This was pre-egcs and was one of the triggers for the egcs project.
> Outside of programming classes, I could see having maybe ONE writing class.
The more, the better. You would not believe some of the writing I've seen from advanced undergrads or graduates. Expressing yourself clearly and professionally is one of the most important skills to have in any field.
> Otherwise, the Humanities, Management, Statistics, and a few others were just filler.
Stats? I wish I'd had more classes in that area!
It's important to be well-rounded. I took a music class my senior year. Even though I was already heavily into the piano and trumpet, it opened up a whole new world. I can't imagine what I'd be doing without it.
I've gained far too much enjoyment from literature, music and the arts to just throw it all away.
If you go read "The Soul of a New Machine," you'll find an interesting discussion of the job interviews for the Data General Eagle team. One of the questions asked concerned hobbies outside of work. It was very important to the team leaders that their workers do something outside the computer world when away from the lab.
If this is all too wishy-washy and non-geeky, then so be it. I'm happy.