Just to keep the mortals in the loop,
This date to day, on Tuesday the 13th of August 2002, Theo had another fit
and kicked out all the OpenBSD developers for a couple of days or so:
Subject: Re: dealing with security issues when Theo is away
Date: Tue, 13 Aug 2002 10:25:08 -0600
From: Theo de Raadt<dera...@cvs.openbsd.org>
None of this that you posted changes a single thing.
I DID say who was responsible.
Those people were not contacted.
It seems you still don't understand the level of not caring that
happened.
I am taking a holiday next week. For that time, I think cvs will be
turned off.
Good god, reading even further, you are so fucking out of touch.
There are only 3 machines on at my house at the moment, and you start
talking about OTHER machines?
NOONE PHONED ME.
And:
Subject: And
Date: Wed, 14 Aug 2002 17:35:30 -0600
From: Theo de Raadt<dera...@cvs.openbsd.org>
If I don't get answers from the evasive developers soon, I am going to
take this to misc, and I will be very open with naming names.
This is now days of people trying to hide from what happened.
------ snip ---- snip ----
So Theo shut down all machines in his basement and none of the developers
had any access to the work they doing.
I'd like to remind people that at this point we lost valuable developers
like Niels Provos which turns out the be one of the few who fully understood
crypto and the security improvements like separation of privileges. Not to
forget Hugh, Aaron and a few others.... Others had their account re-enabled
after groveling. And all that over a misunderstanding that is to blame to
the fact that Theo had no written procedures on how to deal with 'issues'.
When Theo is away, you just 'wing it'.
Today, we see the same shit all over again... Theo just announced the
following:
----- snip ---- snip ----
To: hack...@cvs.openbsd.org
Subject: Tree locked
Date: Fri, 13 Aug 2010 10:03:05 -0600
From: Theo de Raadt<dera...@cvs.openbsd.org>
I am locking all the trees until the development community decides
how future releases will be done.
Yes, we all have to do our part. We write code, and some people go
further by building, and some people go even further by building
during the release cycle.
But everyone also has to test, or we will ship crap. Yet on random
releases this process totally falls over, and we end up shipping crap.
Three architectures did not have one of their boot methods checked --
yes, they are listed in the TESTS file! -- and the bugs were found
very very late in the process. Basically 1 week after the TEST file
went up.
pkg_add turns out to have a major bug which would have been spotted if
just a few other people had tested another line item in the TESTS
file.
That is ridiculous.
I cannot accept all this pressure being on me; I want recognition that
all the people who thus far have accused me for not being clear are
wrong.
we have developers in the group who cannot by themselves recognize --
even ANTICIPATE -- that we are going into the same 6-month release
cycle, EVERY feb/march, and EVERY august/sept, and then participate to
identify the 10 last stupid bugs that we should fix. Is there that
little desire to ship a good release?
It will not be fixed by sending more mails out. I did send out mails
and they were ignored. Communication coming from me is not the
problem; it is clear that developers are NOT LISTENING.
The problem is not new developers either. Anyone accusing them has
got it all wrong. New developers are supposed to learn the ropes from
old developers, and it is the old developers who are not doing their
part. Yes, that means you.
31 people tested, meaning 140 people did not. Any suggestions for
people who have idled out and don't want to be involved any more?
When we ship a crap release, it is not my fault. It is YOUR fault.
So tell me how we are going to fix this. Don't reply just to me.
As I said, I will not accept responsibility for what went wrong here.
And if anyone wants their account disabled, please accuse me just once
more.
----- snip ---- snip ----
And he picks on a few individuals:
----- snip ---- snip ----
To: hack...@cvs.openbsd.org
Subject: Testing
Date: Fri, 13 Aug 2010 09:39:12 -0600
From: Theo de Raadt<dera...@cvs.openbsd.org>
I would like to see some tests for the upcoming release from Henning.
I hope this communication is clear enough.
----- snip ---- snip ----
To: henn...@cvs.openbsd.org
cc: hack...@cvs.openbsd.org
Subject: Apology
Date: Fri, 13 Aug 2010 09:44:45 -0600
From: Theo de Raadt<dera...@cvs.openbsd.org>
I find myself unable to take back-talk from people discussing testing
and then accusing me of having not done my job asking for testing,
when they did not do their job. You did zero tests (everyone can
check ~jsing/TESTS) over the 3 week period, yet you feel you can
lecture me on why the testing procedure failed.
I will enable your account when you find the time to apologize for
your statements.
I hope this mail is clear enough.
----- snip ---- snip ----
And now he comes up with this brilliant gem:
----- snip ---- snip ----
To: hack...@cvs.openbsd.org
Subject: Testing idea
Date: Fri, 13 Aug 2010 10:16:01 -0600
From: Theo de Raadt<dera...@cvs.openbsd.org>
How is this for a testing idea:
- if you are part of the release build process (ie. building binaries
for the release), all is good
- if you work hard on finding and fixing a bug found during the
release / testing process, all is good.
- if your name is in the TESTS file, all is good
- if you send me a mail describing circumstances that led you to not
be able to test, all is good
Otherwise, on the day the CDRs go to the plant your account will be
disabled
'due to inactivitity'.
Is that where we should go?
----- snip ---- snip ----
Yes Theo, let's just go that road and you'll find yourself again in a same
spot as 8 years ago: wondering why everybody started bailing out on you.
Not all of us accept not that you are doing the right thing, so we'll find
another sand box to play in.
Enough of the bullshit... when is this project grow up and be run by a team
that is accountable (voted for?) and not a guy sitting in his livingroom in
his underwear ranting on ICB?
H.
