[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: release notes.

You don't sound like a dick at all; you sound like a lazy bum :-)

I hope you don't think I'm bitching, but I'm going to try to explain my
side of this.

I fully understand what real releases take in time committment.  However, I
remember Scott saying explicitly that our alpha releases should be for
real; if I still saved all my mail, I would excerpt the quote here.

These releases affect me.  The last three releases have broken my efforts
to work on the RT core, and I have no idea what's going on because there
are no real release notes.  I can't even check on the state of the system
relative to known bug reports such as the keyevents, and as a maintainer I
should be able to do that release to release.

I used to build dozens of unreleased cores in trying to make a single RT
alpha release of the old system, but that is part of the responsibility of
avoiding putting out total shit.  You can't release something to alpha
until the rest of us are ready for it, and right now there's no hand
shaking going on at all.  Sometimes I would have to get Rob or you to test
drive a core before I put it out to alpha, and that's not because some
random users subscribed to alpha.  It is also because careless releases
affect Scott's research, and it affects the rest of the group developing
the system.  I'm at a point now that I want some way of putting a core on
lisp-pmax1 and choking sup, so I can install a core when it won't screw me.

I don't mean you are careless, and I realize all any of us has to do is go
back to the previous release ... but we can't really do that right now, can
we.  Maybe I should save a copy of each good core in AFS, so I can copy it
to lisp-pmax1 whenever a release like this occurs which seems to be most

I do admit that I'm being somewhat anal retentive since I irrationally like
to believe life is an absolute knowledge game, and I also like to be in
full control.  Well, I'm not anymore, and I'm more sensitive to being
unaware of the state of the world.  It is a GREAT frustration to me to
constantly deal with "releases" that prevent me from working, or that cost
me three days of compilation when I could postpone that to when I'm more
emotionally prepared for it.  I feel like a hapless end-user, like the
people I used to shaft :-)

Once again, I just got to a point to compile some stuff for a new level of
information on my efforts, and you have changed the core.  I spent all day
gettng ready to receive that information, and now I might have to wait
until late tomorrow or Monday to see the results.  I don't even know though
at this point what I have to do -- recompile the whole backend or what.  My
compile of code: is broken now with more than two times the number of fatal
errors, and it is likely this means I have to modify my backend and
recompile the whole fucking thing again.  I was hoping to see less.

I realize good release notes wouldn't have prevented me from losing the
gratification I was hoping for tonight, but then I might have known why my
dumper was broken.  One thing a maintainer must do is anticipate how his
users will be affected, and tell them how to unfuck themselves with the
maintainer's latest and greatest.

Okay, I'm sorry to everyone for the verbosity here.  I certainly hope no
one thinks I'm relieving the tension of having my car stolen last night.
Ha ha, see I think it is funny that it was stolen between 10-11 p.m. from
right in front of my house under the street light.  I'm glad I'm an urban
dweller because it is so much more convenient now with only one vehicle
left to worry about.

Happy hacking,