Thank you all for your replies, you have been very helpful.
Special thanks to Frank Gurman, who after drinking a nice cup of wine
killed my doubt without mercy, intead of pointing me to the manual I read
over and over again to read the same lines that made me post here. After
all, I know how to read =o)
So now the problem has itīs cause discovered: I was reluctant to open my
mind to a new concept. A concept where itīs all related and revision
numbers donīt help anymore as CVS to tell how a file changed. A concept
where the repository is a project. And a concept that I didnīt feel
confortable with.
Yes, It bothers me. Iīm with Ron Gilbert on this one. Maybe I have to
become more familiar with it and (as I always loved to do) stick to the
standard. But I canīt possibly see the advantage of having my
"can-be-useful-and-meaningful-release" trasformed into a
"internal-ignore-me-please" revision.
I will have to study a lot (and I already am, making numerous tests)
because, as I see it, my destiny is to use multiple repositories.
Until then, I really appreciate the comments and healthy discussion (after
all, thatīs how big projects became successfull: because of discussion)
that this list sustains on this matter.
I really loved the way Subversion promises the moon to programmers and I am
convinced itīs 450% better than CVS, but hey, Are you serious that the
Project1 programmers can freely change the Project45 revision? I donīt see
any meaning here. Even if you say: "Hey, itīs a technical gory deatil so
you donīt have to use dates." Ok. I buy it. But does it make sense that
yesterday you were working with revision 342 of Project2 and suddenly today
in the morning you are working with revision 784? I mean, the guys from
Project45 have gone wild in their nightly work, but why the heck should
that mess with me? Project45 is a Web page, they commit every second. And
as an organized person, I preffer to know that my project increases itīs
"not-to-be-considered" release as a helper, and not an exposed technical
piece that we should share among all... I know it shouldnīt be used and de
industry standard code growth indicator, but it helps... it really helps.
To me, sharing it with other projects is like sharing your wife with your
neighboors.
Anyways, Iīll go drink some coffee now... I would drink Franks suggestion
but here in the office itīs quite regulated. Believe me, Iīm needing it.
Thanks to all in the list,
Rodrigo
Ron Gilbert
<lists@rzweb.com Para: users@subversion.tigris.org
> cc:
Assunto: Re: Unclear: CVS and Subversion repository difference.
17/06/2005 00:59
> This isn't a problem; you just imagine to be. Lots of subversion
> newbies think that it matters, and it just doesn't. Here's a FAQ
> about it:
Well, it does matter to me. Sorry we'll just have to respectably disagree.
And don't get me wrong, I like the global version number for the project
much better than a per-file version number.
Anyway, non-issue because you can simply create multiple repositories, and
I find that a better solution anyway, as the thread originator might as
well.
Maybe my situation is different from yours, but I am dealing with several
projects with many offsite contract programmers and artists and they can
not always have read access to all the projects.
> If you're using svn://, then you can control per-directory write-
> access via a pre-commit hook script.
But read access is exactly what I need to control, hence the multiple
repositories.
Ron
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Fri Jun 17 16:02:41 2005