Fading Memories

On Friday I noticed a problem with my Kubuntu installation on my Thinkpad X61 Tablet. Now only would the wireless driver regular drop dead (and then iwconfig would aver that no network device had any wireless extensions), it would also try to take 100% CPU, leaving little for my compilation processes.

On Saturday, I received the latest issue of Linux Magazine which came with an OpenSUSE 10.3 dvd. Taking this as a omen, I decided to install it on my laptop, together with a fresh Kubuntu partition and a Project Indiana OpenSolaris partition.

I've been using SuSE since version 5.0, and I'm still running 10.2 on my server. I couldn't keep my server up without Yast, which I find great for system administration. But it's been a few years since I last used SuSE, SUSE or suse on my desktop.

Some things are good, nay great: the 64 bit version is very fast. Yast is still great for system administration. Plugging in my laser printer or my digital camera pops up a helpful printer installation dialog or digikam. Not so with Kubuntu. The fonts somehow are much nicer, much smoother, especially after disabling manually all hinting. And the Intel graphics driver does work with Krita 2.0, which didn't work with Kubuntu. Wireless networking doesn't drop dead anymore.

On the bad side, and this is very bad, my expensive tablet screen doesn't work. And I cannot figure out how to fix that. The XEmacs packages with OpenSUSE is a beta and crashes on the kde-emacs scripts. No suspend-to-ram on lid close that I can figure out. And, really, the package management system is awful. Dependency resolution simply does not work out nine times out of ten. It's still incredibly slow, although much faster than 10.2. Whether I use apt4rpm, zypper or yast doesn't matter: just remove a single package can take minutes. I'm not an expert, so there may be something great about rpm that I don't see? I just want deb, dpkg and apt-get and be productive. None of the myriad non-deb package systems are worth it.

Still, it works, it's fast and I can develop again. After I've compiled XEmacs from scratch.

And so are the three stooges. I have been wondering what Microsoft would try after having failed buying a fast-track standardization of OfficeOpen XML (and I've been wondering why nobody has sued Microsoft for break of trademark for that name). It's getting clearer: in the past few days all over the it-related web stories have started sprouting that spread the meme that ODF supporters are leaving the sinking ship, that ODF isn't a good enough standard for all document needs now and in the future and that since we'll need to interoperate with OOXML anyway, why not have it standardized. All backed up by statements from siome ODF Foundation spokesperson. But while ODF Foundation has a very grand sounding name, it's just two or three crackpots who failed to make money out of ODF and are now trying to make money out of something else.

But even though I'm not a conspiracy nut, I do think I'm detecting a pattern here. If you cannot convince people your standard is good enough, try to convince them other standard sucks, too.

And for anyone wondering what exactly is wrong with OOXML (apart from the problem that even Microsoft doesn't implement it in its own office applications), please look at the Eooxml Objections Clearinghouse.