svn fetch multiple directories
Daniel J. Luke
dluke at geeklair.net
Mon Oct 13 14:47:21 PDT 2014
On Oct 13, 2014, at 5:27 PM, Ryan Schmidt <ryandesign at macports.org> wrote:
> On Oct 13, 2014, at 4:15 PM, Mark Brethen wrote:
>> On Oct 13, 2014, at 4:10 PM, Ryan Schmidt wrote:
>>> Why do you need to fetch from multiple URLs?
>> The repository contains files for windows, linux, etc. (everything but the kitchen sink) which is unnecessary for macports (the install takes an hour because I'm limited to checking out the whole trunk for each subport).
> OK. So if we did cache svn checkouts, that would help in that you wouldn't have to spend a lot of time downloading, but you would still have unnecessary files on disk.
> Recent versions of svn include pretty good support for sparse checkouts, but MacPorts doesn't have any support for that, so if you wanted to make use of that you'd have to write custom fetch code again. :/
> In this case I'd probably set svn.url to the "main" URL, i.e. whichever URL will provide the largest amount of data, and then use a post-fetch block to get the rest of them. This way if we do eventually cache the checkout from the svn.url, you'll get the biggest benefit.
it may be worthwhile to step back and figure out why the port has to checkout from svn at all. In general we should discourage (ab)using source control repositories in this way. It's /much/ better to be able to get a download file that can be mirrored/checksumed etc.
We should reserve the other ways of getting distfiles for extraordinary circumstances (which this might be one, but it's useful to remind people that it really should be a last resort).
Daniel J. Luke
| *---------------- dluke at geeklair.net ----------------* |
| *-------------- http://www.geeklair.net -------------* |
| Opinions expressed are mine and do not necessarily |
| reflect the opinions of my employer. |
More information about the macports-dev