> This idea comes up and gets knocked down over and over again on this
> list. I guess there isn't much to be done about that.
There is a wealth of unstated information behind that sentence. Since
I myself have raised this issue at least twice, spawning 50+ message
threads, I feel I can empathise quite strongly with those that have come
after me, asking for the same thing. Here's the information that's
unstated. Since MANY people see the text-base as an obvious penalty,
(and it really is in large scale products, all arguments about how
cheap disk space is aside), the one thing I have not yet seen from any
of the core designers is any attempt to address the issue. Just about
every time this discussion comes up, the real point gets lost behind
everyone pointing out all the reasons why its hard.
Many things in life are hard, but to make them easier they need to be
addressed. The issue of the text-base not being the same as the working
copy (due to newline translation and keyword expansion) is not a
justification for keeping the text base, it's a justification for how
things like newline handling were implemented. Theres a big difference
between the two. In a sense, I see a lot of the solution justifying
the problem, rather than the problem justifying the solution. There
have to be other ways to deal with newline conversion, if it is even
that much of an issue in the first place. There most certainly are
ways to deal with keyword expansion. cvs doesn't keep a text base and
deals with it just fine thank you.
I think svn is an absolutely wonderful project. I respect the team
that's done such a bang-up job immensely. All I see here is professional
discipline and that's more refreshing than you imagine. However, and
I mean this with ALL due respect and this is meant as a constructive
criticism, I have noticed a fairly heavy trend towards "that's the way
it is we'll deal with it later". That's a bit disturbing for the long
term future of svn, unless large re-write or re-design cycles are
anticipated post 1.0, in which case its no big deal at all.
Come the day someone wants to really address all the problems the
current text-base raises (not least of which is space wasted), I would
be very happy to participate in any discussions, and even happier to
help churn out the code, but I know that's not going to happen till
after 1.0, and there are probably more important things to worry about,
like alternate FS solutions to DB. But please, remember me if and when
this really does come up for review.
Sincerely,
Kean Johnston.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Apr 28 05:41:14 2003