Date: Sat, 25 Jul 1998 15:23:08 +1000 (EST) From: Peter Hawkins <peter@clari.net.au> To: Tim Gerchmez <fewtch@serv.net> Cc: freebsd-newbies@FreeBSD.ORG, freebsd-bugs@FreeBSD.ORG, brian@Awfulhak.org, davidg@FreeBSD.ORG Subject: Re: I'm leaving the FreeBSD scene. Message-ID: <Pine.BSF.3.96.980725143511.26047E-100000@dana.clari.net.au> In-Reply-To: <3.0.5.32.19980724064754.00818e60@mx.serv.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Hmmm
Let's get some perspective here. You report a problem which noone else has
been able to reproduce (I've run 2.2.7 since a few hours after its release
with no problems at all) and when asked to give send a log or (if you haven't
kept one) reinstall and retain the report the second time to permit people
to investigate, you claim harrassment?
For years, I was the systems administrator in charge of supercomputers
(all of which ran bleeding edge software and OSes all the time, so bugchasing
was a constant responsibility) and a lecturer at one of the world's largest
and most respected universities. I have reported faults and provided suggestions
to all of the major commercial systems providers.
Let me tell you this: if you filed the report you did with *ANY OF THESE*
you would have been completely ignored. At least FreeBSD took you seriously!
Most operating systems developers assume first that you "probably buggerred
somethign up" until they receive many identical complaints. They will
then follow up if the fault is properly reported (as yours was not), IF
their company believes it's worth it, with a request a few months later
that you repeat the procedure and send them the logs. The uni had staff
members who did little else *but* repeat procedures that were suspect
for the purposes of sending the logs to SGI, DEC, SUN, CRAY or whoever.
At this point, if you were *lucky* they wouldn't just say "wait for the next
release" (but we were rarely lucky and the next release usually had not fixed
the fault). In one case (SGI) their official policy is that patches are
released and it's up to you to watch their web site and get them. They also
have a policy of not checking patches with already-patched OS releases. This
means that patch #32765 (say) which is required to run your Gaussian 97 can
only be installed if you don't install patch #34544 which prevents a known NMI
hang or plugs a known security hole. Of the many thousands of patches, the
interactions are neither explored by them actively or (for the most part)
published.
Basically, you see a bug. You call them. They put you on to USA. You call
USA. USA say "try the web site". You wade through the patch list and just
maybe you find one. You d/l it and install it. You then pray that none of
your users come screaming to you about it breaking somethign else. It breaks
something else (or you think maybe the patch did, but you can't tell). You
do another round of phone calls. You read the web site. You give up and
uninstall all ptches ("SGI standard procedure") and install them (typically
a hundred) one at a time until it breaks again. (Your uni's supercomputer
has now been down for a week). When you finally find the interraction,
you decide which patch to put in based on which user has the most research money
and can make your life the most hell. You do it and wait for the other to
go for your blood by saying "sorry nothing we can do". The user finally
gives up and goes away. The next week you get a call... "something new
has broken on the SGI PowerChallenge". You weep.
Come to think of it, is there any chance of porting FreeBSD to an SGI
PowerChallenge anyone? <grin>
You say FreeBSD isn't following other OS's standards for bug tracking and
reporting? I agree! This is why I use it ;) The words "thank god they are not"
spring to mind.
The FreeBSD community:
1. didn't reject you out of hand despite you being one voice and instead
treated your report seriously
2. responded same day (as opposed to in a few months, if ever)
3. didn't say "read the manual and try again" but instead treated you
as an adult and assumed you may be right and asked for info to
see what could be wrong.
4. didn't throw itself into "don't criticise our product" mode like most
5. didn't say "wait for the next release" but were prepared to patch
this one.
6. didn't throw their least senior programmers at it while the seniors
worked on the things they took seriously
7. asked you the *same question* and for the *same information* that
all designers ask if and when they do take you seriously (ie for
either a way to reproduce it or for a log which shows what happens
on your system if you don't know how to reproduce it specifically).
8. would have worked round the clock to patch it (witness recent well
publicised security holes which FreeBSD patched 6 months before
their commercial counterparts who had been saying "solution:
don't run rlogind" all that time)
9. would have kept in contact with you
10. would have regarded any incompatability between the resulting patch
and existing patches as a further bug to be solved.
It seems to me you reported "227 is shite fix it" and assumed it's "shite"
to anyone who tries. Well you're wrong. I run it just fine and to my knowledge
noone else has yet seen your problem. Unlike most though, we do not then go
on to say "so you must have screwed up", but we say "ok so there may be
something different about your system somewhere that causes it to highlight
a fault we cannot see so can we know more please?"
If you cannot accept this it's impossible to address your problem unless someone
with greater expertise in fault reporting has the same experience as you did
and provides us a means of reproducing it in a test situation.
I believe that the FreeBSD community will not miss you and I wish you well
with your proposal to do is your right and use our "incompetant" but donated
labour to make money.
Peter
Hilink Internet Peter Hawkins
381 Swan St Richmond,
Vic, Australia Ph: +61-3-9421 2006 Fax: +61-3-9421 2007
http://www.hilink.com.au Peter@hilink.com.au
FreeBSD Project: thepish@FreeBSD.org
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-bugs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.96.980725143511.26047E-100000>