I really don't think Pat gives a rip about the development process of the linux kernel. It affects him in no way as Slackware's maintainer. I have never seen him make any postings resenting Linus. I'm not sure how else Pat could have given you that impression *except* verbally. I'm not one to trust psychics myself. Plus, Linus does not know every line every every code in the kernel or everything thats in it. If you knew about the linux kernel development process you would know he has his "lieutenants", or people he has designated as filters for patches and such to help him. He, like most people, are not much into micro-managing. Second, bsd can keep their development process. It works for them and thats great. What works for them doesn't mean it works for everyone else. Linus does things his way because it's what he's most comfortable with, if you don't like it, fork. He really could care less as I'm sure most of the developers as well. You may not like him for that but thats not his job or desire. If it was a popularity contest nothing productive would get done. As Linus has said, he likes to leave politics out of it as much possible. Even though they usually find their way into things.

Oh and about Linus running a mac. I know the article you read. He got the mac for free first of all. Second he said, "Part of it is simply that I wanted to try something else". Here's the article link. Also, the G5 he he has *only* runs linux, so no he wasn't testing out OS X either.

Well that is a comforting thought, if it is indeed the case.

Oh and If you think he's on an ego trip he's really not. If you read his biography he says it in plain English, he's an asshole. I also have never seen him pretend to know everything about everything kernel related. If he doesn't know anything about something then he will ask people to explain it to him. If you remember when the stuff about reiser4 on the LKML happened, linus asked hans to explain some things a few times. He doesn't pretend to know about things he doesn't know about.

Pat has never said nor do I think he would say anything about the developmnet process of linux. That is not his style. However, I was refering to his actions. His slow acceptance of new versions of software speaks volumes, especially the kernel. People's actions speak louder than words.

Even when he dropped gnome, he was incredibly polite about it in his changelog.

Wow. Well first, 2.6 is on 10.1 cd's. Also, Pat said he just wanted to wait till 11.0 to release with 2.6 as official kernel because 10.1 is a *minor* release. *Major* releases usually change large things such as a mighty big kernel jump. I don't understand how you could read into that as Pat not trusting Linus and the development process. If you noticed 10.1 came out pretty fast. Hardly enough time for testing out a new kernel of that magnitude.

Pat's slow to accept a whole lot of things, and that shouldn't be taken as a slight against them. Apache 2.0 has been out for over 2 years, now, and Slack still ships with 1.3. Apache's development model is extremely different from that of Linux, but Pat's stated reasons for staying with 1.3 are exactly the same as his stated reasons for staying with Linux 2.4: Proven track record. Technically, Apache 2.0 is still a development branch, just like Linux 2.6 is. While the stability and security in 2.0.54 are far removed from 2.0.1, just as 2.6.11 is far removed from 2.6.1, Apache 2 is still a .0 tree, which in most programming histories means development. There's an unspoken rule in computer circles that advises the user to wait for the .1 before buying a product, because by then the Beta is over.

Likewise, Pat's sticking with PHP4 and MySQL 4.1. It's simply because he's conservative when it comes to those things, and quite frankly, when dealing with a distribution that's known for its stability and reliability above all else, that's a good thing.



EDIT: And yes, Slack 11, which I think is due by the end of the summer, is slated to ship with a 2.6 kernel.

I'm sorry but yeah I do read into it like that. If he had complete trust that the development process of the kernel was rock solid... slack would be on 2.6 already. A lot of software version are a version behind in slack... I think the devel process plays a critical role in pats mind and he is therefore relunctant to jump into wagen like every one else.

Stability and progress are always competing agianst each other. Linux seems to chose the progress path more frequently then stability. That is not to say linux is unstable, just that stability testing is done mostly in the "field". Pat relizes this and is slow in adopting this new versions.

Take some of the more bleeding edge distro's out thier. They use all officially released version of software... the latest by the way. And I can speak from experience that stability is not thier strong suite.

Yes but I don't think that means Pat somehow has some kind of dislike for Linus or the kernel development method. *ALL* software goes through a maturity period. It doesn't just go from 0.0.1 to stable the next day. The kernel is still changing it's development model as we know it also. For right now instead of Even/Odd kernel releases its now this 2.6.x.y thing. That could change if they find something even more efficient or better suited to them. Linux kernel development moves at an incredibly fast pace. I think it was something like the kernel developers write 30,000 lines of code a MONTH. If it was me in charge I would have absolutely no idea how to contain that kind of a project. Linus and the developers manage it though and they manage it quite well. Besides does it really affect you or me in any infinitely large way? I don't see how.


Some things you can't test unless it is in the field. How do you test if network code can handle large volumes of traffic? By using it. If you think thats a bad idea then you should read books on Linux's earlier years. Much of the bugs fixed in the network code was never fixed or found for that matter until it was used in large environments. Thats how you find those hard bugs. Now I'm not saying use the newest kernel for muli-million dollar business servers. I am saying that not testing things in the "real world" will teach you a lesson real quick.


You don't have to tell me other distro's aren't stable. Thats why I haven't touched another distro other then slack in almost 4 years.

Take gnome for example. The development process is none existent. Pat knows this and dropped it becuase he was getting to many bug reports about supposedly "stable and official" release versions. So he dropped it. But still he was very polite about it in his changelog. Saying it has always been a "moving target" and not to be taken as a slight against gnome. I read into that meaning he thinks the development process of gnome is rather sh*&tty and he does not want to deal with it anymore.

The major reason he dropped gnome is because it is the biggest pain in the ass EVER to compile. I have done it on LFS once and I vowed never again. That was awhile ago and apparently things have gotten worse. Pat was most getting annoyed because he uses the same build scripts for each package. When a new release comes out he uses the same script and just makes the package and tests it out. Makes managing a whole distro by oneself a great deal easier. Gnome kept changing their build process and Pat kept having to change his scripts and adding more packages. Pat decided his time would be better spent on other things and that the community could take care of Gnome for slackware if it chose.

I'm refering not to the actually development of code but to how it gets tested. Quality control. Pat was getting I'm sure thousands or millions of bug reports. Things he didn't cause and therefore couldn't fix for gnome. I'm reffering to how they test the code before they decide to have an officail stable release. Development on gnome is continueing sure. But how do they look at the code? Who looks at it. Who decides if it is stable. how do they decide when it is stable for a release. These are the things i'm talking about.

I really hate to be the contradicting person I seem to be right now. I'm not sure why Pat would get bug reports. He's not the maintainer or developer of Gnome. I'm sure he got plenty of emails saying what a piece of crap Gnome is. I'm sure if he was getting that volume of emails he would request people to stop and send them to the Gnome maintainers. I know he answers most of his mail and I'm sure he would put a stop to something like that immediately. Also, in the changelog he even says it's been an idea he's been considering for more then 4 years. So it's nothing new with the latest release(s) of Gnome. Gnome's always been a piece of crap to put together and he finally just dropped it.

Getting back on topic, one thing i'd really like to see is GCC being updated...

Having the 2.4 kernel is one thing.. Although its stable, but i can vouch for the stability of 2.6.9 and above.. But anyway, Slack 10.1 is still on GCC 3.3.5. gcc.gnu.org says that 3.3.6 There's the 3.4.x release, which is up to 3.4.4 now, And there's even GCC 4.0.0 out.

I'm sure there's a good reason for the continuing use of 3.3.x, be it performance or security, stability, whatever.. but isn't it time we see a jump to atleast 3.4.x?

Also, how come Pat announces updates for Slack/390, but not for Slamd64, Minislack, or Splack ?

because slack/390 is an official slackware port...