The Old Way Was Better

[Michel Talon]

 > For what it's worth, I and a large number of other people I know would
 > agree. The 5.x releases so far have been VERY badly received with the
 > people I've spoken to and it's not done FBSD any favours. I seem to
 > remember something similar happening when we were all running 3.x > 
 >boxes  and 4.x started getting releases, but nothing quite like this.


Anyways i am one for whom FreeBSD 5.1 works quite well, and i cannot say 
the same of the latest 4.* series. On my laptop which runs 4.8 things
have degraded with respect to what it was around 4.4. After suspend
sound doesn't work any more, pcmcia cards are not properly reset,
and i am now seeing ATA DMA errors and downgrading to PIO, a thing
that never occurred up to 4.7

Why am i ranting about that? i think that the FreeBSD developers work
full time on the 5.* series, and that it has diverged so much from
the 4.* series that they commit errors when backporting stuff from
Current. You have only to look at the recent PAE merging fiasco to
attest of that. The 5.* series began several years ago, and we cannot
reasonably ask that the developers remember what is in the 4.* code.
The net result is that both the 4.* and the 5.* series are presently
more unstable than stable. In my opinion, the 4.* series should not
have been maintained so long (except for security fixes), it stresses 
too much the available developer workforce. Choices have been made
for the 5.* series, good or bad, this is not the point, it is urgent
to concentrate on that, and only that. Otherwise you can be assured
that Linux will draw circles around what will remain of FreeBSD.
Matt Dillon disagreeed with these choices and is developing his own 
version starting from FreeBSD-4. The future will show what were the
good choices, but at present, concentrating on 4.* is suicide.