Where is FreeBSD going?

[Brad Knowles]

At 11:28 PM +0000 2004/01/06, Paul Robinson wrote:

>  Accepted. It came from paul at iconoplex.co.uk and therefore can
>  only represent my own opinion.

	In the future, may I suggest that you make this sort of statement 
more clear at the beginning?  It sounded to me like you were standing 
up as a self-appointed champion of the rest of the world.

>  But I know a lot of people who are looking at deploying 5- who
>  aren't just pissed off - they're *scared*.

	FreeBSD-5 was always going to be problematical.  There have 
probably been more things changed for this major version than for any 
previous major version in history, maybe even for all previous major 
versions combined.  They bit off a great big honking whackload with 
this version, and they knew it.  That's why we're so far behind the 
original release timetable (one year?  two years?).

	Any reasonable production-oriented plan would have been to stick 
with 4.x until such time as 5.x has been declared "STABLE", and then 
wait for another minor release or two after that.  Timetables can 
(and do) slip, so you'd have to build that into the picture.

>                                             I don't think many of the
>  developers understand this.

	My personal opinion is that I believe many of them understand 
this better than you know.  See above.

>  To us (yes, I'm not speaking for Brad Knowles), FreeBSD is not a
>  project we spend our spare time on and love and adore. Well, it is,
>  but it's also a lot more. It defines our careers. We roll out
>  something that isn't "quite right", our jobs are finished.

	I've been there.  I was the only FreeBSD guy bringing in machines 
into the largest ISP in Belgium, where everyone else was a Linux 
fanatic.  They learned to respect the machines I brought in and how 
rock-solid they were, and my co-workers have since taken over and 
rolled out even more FreeBSD servers since I left.  I believe they 
have the largest USENET news servers in the country, and the machines 
are also some of the most robust in the facility.

>  Right now, if somebody asks me what our roll-out strategy is for
>  the next 18 months, I have to respond "don't know", whereas the
>  Linux guys are just laughing... don't even start me on what the
>  Windows guys are doing to my career right now....

	See above.  Roll out 4.x for now, wait for 5.x to stabilize. 
That should have been the plan since 5.x first became -CURRENT years 
ago.

	The Linux guys have a lot to deal with, too.  Red Hat licensing 
is now looking nearly as expensive as Sun, Mandrake is bankrupt, SuSE 
is being bought by Novell (in preparation to kill it?), and who else 
is left?  They've always had a schizophrenic situation, with the 
dichotomy between the kernel developers versus the distribution 
creators.

	Windows?  Well, Longhorn just got pushed out yet another year -- 
wait until 2005 or 2006, at least.  Nothing to worry about there.

>  OK, so it has got personal... I accept it is not the FreeBSD
>  development team's job to look after my career, and to date I've
>  looked after that by myself OK, but all I'm asking is you try and
>  at least understand where some people are coming from on this.

	I understand, and I believe that the vast majority of the FreeBSD 
developers understand.

>  Mark has mailed me off-list. His tone isn't great. I probably
>  deserve the "Fuck off. Go away." I'l deal with that seperately. :-)

	In my original draft of my response, I basically told you to STFU 
myself.  I decided that discretion was the better part of valor, and 
toned down that remark.  But I can certainly understand the 
frustration resulting from having seen your post.

>  OK, I've never run into that. Over on the DragonFly stuff, he seems
>  pleasant enough and his ideas are innovative, strong, if sometimes...
>  *cough*... eccentric (e.g. replacing sysinstall with an Apache server
>  and a load of PHP...), but I'll accept I haven't seen that, and I
>  know others have had their problems there.

	Well, since it's his project, I'm sure he feels a lot more 
secure.  Perhaps he's taken some lessons from previous mistakes with 
the FreeBSD project, and he's working to avoid re-living them with 
DragonFly.

>                                              I did see the fall-out
>  on these lists with the argument that caused it all to kick off
>  about a year ago though, and I don't think others on the project
>  dealt with him (in public at least) fairly. Again, just my opinion,
>  I wasn't involved, don't know what happened in private.

	I don't think that we saw more than the slightest bit of what 
really happened.  I will agree that I think this matter could have 
(and should have) been better handled with regards to the public 
aspects, but anyone who was watching the lists at the time should 
have noted that this was not a new type of problem, and there were 
multiple references to previous situations of a similar nature.

>  Ooooh, no. That isn't what I want at all. I just want end-users to
>  feel they have a voice. That's all. Maybe they do, and I don't see
>  it. Maybe they don't *and that's for the good for the project* but
>  in my opinion, it just seems odd.

	People of all sorts will get that voice, if they find a way to 
take some sort of ownership in the project.  That could be 
development, QA, documentation, or something else.

	But if you don't want to join the party, then you can't 
reasonably complain when you are unable to vote when the primary 
rolls around.  Moreover, you can't reasonably complain about the 
person who gets elected, if you didn't vote.

>  Actually, no, I suspect 4.9 will keep me going for at least another
>  18 months, by which point hopefully 5- stable will be back where
>  everybody wants it.

	Which should have been your plan from Day One.

>  And yes, I was having a bad day, and my tone was rotten to those
>  of you who put so much time into FreeBSD, and all I ask in future
>  is that you realise that some points about bitrot, bloat, bad
>  performance and a lack of *feeling* the end user gets heard is
>  enough to cause real problems for a lot of people.

	FreeBSD is a very unique OS.  Unlike virtually any other OS 
(except other *BSD variants), the user has the option of being more 
involved in the project and helping to determine its future.

	You don't get anything remotely like this with Windows -- you 
just get whatever dreckage and bletchery that Redmond chooses to 
foist upon you.

	You don't get anything like this with Linux -- you could get 
involved with a particular distribution, but that doesn't give you 
any potential for involvement with the kernel.  You could get 
involved with the kernel, but that has little direct impact on the 
user environment.

	You don't really get anything like this anywhere else.


	It's entirely your option, but if you choose to not get involved 
in the project, then you really don't have any grounds to complain 
that it's not going the way you want.

-- 
Brad Knowles, <brad.knowles at skynet.be>

"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
     -Benjamin Franklin, Historical Review of Pennsylvania.

GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E-(---) W+++(--) N+
!w--- O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++)
tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)