On 9/14/06, Juliusz Chroboczek <[EMAIL PROTECTED]> wrote:
I will not use this functionality myself, I think tand hat the
submitter is using the wrong tool for the job (he wants to use
Unison).

I disagree.  I perfectly saw his point here, and I absolutely don't
think it is Unison that would solve his problem.  There are often
files in a source repository that have a small mixture of global stuff
and local settings.  It is great to be able to commit the larger
changes that everyone needs and then quickly tell darcs to ignore the
file (because of local settings) until I find that I need to update
something with a larger global update, in which case I could do just
as quick an unignore, record, reignore.

I can use a particular example, even:  I've been using Django, a
python web framework.  It uses a settings file that contains
everything from the INSTALLED_APPS (code modules to activate) to
DATABASE_PASSWORD.  Best practices in this example aside (it is a
python file, so if I weren't lazy I could always break local settings
into a seperate python file and import them), it would be nice to not
have darcs ask me if I want to record the changed local database
password every time I record, but yet I can unignore that file when I
add a new application I'd like installed on every repository.  Ignore
patches might be another way around this issue (ie, if I record the
local DATABASE_PASSWORD and mark it that it should never, ever, under
any circumstance even be looked at much less pulled from my
repository), but most likely ignored files would be in this case, as
in possibly others, less likely to cause a security risk.

The original submitter's goals shouldn't be so easily dismissed as the
needs for another tool.

--
--Max Battcher--
http://www.worldmaker.net/
All progress is based upon a universal innate desire on the part of
every organism to live beyond its income. --Samuel Butler

_______________________________________________
darcs-users mailing list
[email protected]
http://www.abridgegame.org/mailman/listinfo/darcs-users

Reply via email to