[Bioperl-l] SVN and ...Re: Perltidy
rvos at interchange.ubc.ca
Sat Jun 16 13:09:18 EDT 2007
CIPRES and Bio::Phylo use svn. As for the benefits, a lot of sales talk has been expended over it already, for my own purpose I like the integration with eclipse (through subclipse plugin) and komodo, in addition to the atomic commits (so I can ctrl+c if I goof up (again)).
For standalone use on osx I didn't use the fink one, but I forgot where I did get it from. It was very easy to set up, though. On windows there is a really nice standalone one (tortoisesvn) that integrates with the explorer so you can see on the file icons what the state of a file is. I know that there's a cvs2svn utility that converts your revision history (seems a requirement).
> Date: Sat Jun 16 07:55:09 PDT 2007
> From: "Hilmar Lapp" <hlapp at gmx.net>
> Subject: Re: [Bioperl-l] SVN and ...Re: Perltidy
> To: "Sean Davis" <sdavis2 at mail.nih.gov>
> On Jun 16, 2007, at 7:21 AM, Sean Davis wrote:
> > As for access, the typical access is over http (or https).
> We're using svn+ssh here (NESCent) so the password is the same as the
> one you set for your account on the server, and you can use public/
> private key negotiation for authentication.
> I think the ability to not provide a password for every single
> interaction is a requirement. If that requires using svn+ssh or can
> be made to work through https too I don't know. On sf.net I have to
> use https for svn and it doesn't ask me for the password each time.
> Not sure how this works though, maybe some local caching?
> We should not be using http, or whatever other protocol that sends
> unencrypted passwords.
> > Access controls can be set up on the server side while allowing
> > anonymous access for checkout. There are many excellent SVN for
> > every OS, so that should not be a problem.
> On Mac OSX the most convenient way I have found is through fink. It
> does ask to install 30 other dependencies, which had me balk at
> first, but me doing it by hand is even worse than fink doing it, so I
> finally gave in and it's really a breeze. I've not had a single issue.
> From a sysadmin perspective, what might be worth keeping in mind is
> that svn is going to store everything in a database (BerkeleyDB I
> think). I.e., there is no such thing anymore as restoring individual
> source code files from backup if one gets accidentally corrupted on
> the server. It seems you have to restore the entire database, i.e.,
> the entire repository. I vaguely recall though that how svn manages
> the repository is actually configurable and that other storage than
> DB is possible too. Don't ask me for the pros and cons of one vs the
> : Hilmar Lapp -:- Durham, NC -:- hlapp at gmx dot net :
> Bioperl-l mailing list
> Bioperl-l at lists.open-bio.org
More information about the Bioperl-l