I just want to share my experiences with Slackware64, why I upgraded to -current, and why am I currently back on -stable.
The most obvious reason for upgrading to -current is KDE 4.3.3, as it has many improvements. Most relevant for me are these:
1. Complete translation
to my language.
2. Dolphin and KMail are even more stable
, though I haven't really have stability problems with 4.2.4 that would be worth mentioning. Just, in very special situations I was able to provoke a crash, where now the programs just "behave".
3. Integration with HAL got a lot better.
Most of my external USB hard disks are LUKS encrypted, now. When I plug them to my computer, the device notifier opens, and when I click on "Volume (crypto_LUKS)" I am asked for the LUKS passphrase. This second step never works in KDE 4.2.4.
4. Look and feel:
Faster, fancier, better.
Nevertheless I am back on Slackware64-13.0 -stable now, for the following reasons.
1. Problem in -current with multilib setup
[EDIT: These probs are gone with newer version of the multilib stuff!]
After applying Alien Bob's excellent (!) patches and massconverting the 32-bit Slackware-current packages as described in the README, I found replacements for some files in /etc/rc.d. For example, there was an rc.hald.new next to rc.hald, but with a size of 0 bytes. And there was a new rc.inet1.conf, with a different size than the one that has been there before. Plus two or three more files with extension .new, either with 0 bytes or with a different size than the original file.
The result was an error message "can not start dbus; can you?" or so, when I tried to start KDE.
Replacing the existing files with the .new versions prevented my system from booting, the next time, even when I did it only with the Non-0-Byte files.
I am interested: Did anyone else observe this?
Here it is reproducable!
2. -stable is just that!
While currently the updates in -current wouldn't hurt, as they don't affect any core libraries, this will sooner or later change, and the likelihood that something essential will break will increase, at least in the weeks or months before the next stable release.
However, I want to make my 64-bit machine my main machine, now, because thanks to the latest updates to multilib support provided by Alien Bob, finally I can use the PCFAX functionality of my Brother MFC-290C, which was actually the last issue that prevented me from doing so. Now this works, and I don't want to take the risk of breaking anything relevant on my main machine.
I therefore decided to live with incomplete translations and the little inconvenience regarding LUKS encrypted USB hard disks, among a few other, minor snags, all cosmetic in nature.
I learned to like Slackware64-current mainly because of KDE 4.3.3 a lot, as all the (minor) problems I have in KDE 4.2.4 are solved, there. It only seems to conflict a bit with a multilib setup.
My hope now is, that we'll have to wait for 13.1 not nearly as long as we had to wait for a 64-bit version (which is so good in quality, that the long time of waiting was really worth it, of course) or 13.0. In fact, I think, the time for 13.1 could be now... or around Easter 2010. This would be much sooner than usual, and break the release rhythm everyone is used to, but the many improvements I have seen in KDE 4.3.3 would well justify such a decision, IMHO.
What do you think?