ITPro on Linux, Slackware and BSD

Normally I ignore the trade rags. There’s been a bunch of Linux-come-lately publicans which were really nothing more than advertising. Even Linux Journal was briefly on my shit-list. Of course, if you itemized your taxes this year, you’ll know why packing a sheaf of advertisements into a “magazine” is appealing to advertisers and consumers are generally apathetic. IT Pro recently rose above the level of crap put forward by most of these and had a fairly decent article on the hows and whys of Linux versus BSD. If you’re curious as to how much of an accidental hero Linux is, this is it. Frankly Linux should never have been and it was only because Torvalds was doing more with less that it stuck around. That and it showcases the power of advertising. Frankly there never would have been a Linux or a BSD had AT&T and IBM and everyone else did the whole Microsoft thing before Microsoft. UNIX was supposed to work on POSIX and a series of standards, and it never did. There was always lockout, but the advertising for The Next UNIX always said “Cheaper, and STILL WORKS!” Cheaper maybe, still working never. Linux answered the call for something you could make work and still be cheap (free), and were it not for the accidental oversight of BSD’s fanbase missing the math co-processor requirement there wouldn’t have ever been a Linux in the first place. Keep in mind the Linux kernel has been paying for it ever since – BSD is usually faster by a factor of two, but BSD also has an ungodly complicated kernel linking system (compared to Linux) and it’s heavilly wedded to the hardware. The GNU/HURD project which was supposed to be a marriage to Linux to the BSD (“mach”) kernel never got off the ground simply due to differing ideas about how userland is supposed to talk to the kernel. Best intentions and all that. The spiritual successor to HURD is Gentoo.

IT PRO also did a story on Slackware. Slack has a special place in my (and a lot of people’s) heart by being the first commercial Linux distribution in English that didn’t suck. SLS always was a mess of bad assumptions, slackware was supposed to be SLS without the associated preconfigured crap. SLS never quite got it right, it was somewhere between BSD and Yggdrasil with the latter getting automatic configuration done correctly for most hardware. Unfortunately 1995 saw people trying to buy hardware to run Linux, Linux didn’t have the critical mass to go the other way with nVidia and ATI tailoring their drivers to run on Linux, so Yggs went the way of the buffalo.

Slack breifly made headlines by skipping a version number in the SuSE/Redhat pissing contest of “Linux 7, 8, 9, etc”. Slack simply opted to skip numbers to get ahead, it was an epic troll. Unfortunately the article points out that slack dropped packages which were too complicated to configure (something constantly plaguing the GNOME community, I personally wish I could just march them into the ovens) and really this is the reasoning behind slack. It either is stable and builds or it’s not stable and doesn’t build so it gets cut. Contrast this to GNOME where the build scripts are straight out of BSD’s ports system and contains all sorts of strange build hacks if you want an illustration of the problem. It touches on something much deeper though – most Linux distros use XOrg instead of X11, and slack seriously missed the boat by staying on X11 for a long time. Granted, there’s some really cool stuff out there in X11 missing from XOrg, but you probably won’t miss it if you’ve never rolled-your-own from source. By contrast, most distributions don’t have looking glass or enlightenment because they require esoteric hardware access and the X11 framework. E only recently updated, but there’s a point to using X11 or XOrg and I think a lot of Linux newbies miss their obligations in understanding what is what and why it’s there in a particular distribution.

There’s a quote which I think is a good ending – “Use Ubuntu Linux and learn Ubuntu. Use Slackware and learn Linux”.