perl-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From sebb <>
Subject Re: Please delete old releases from mirroring system
Date Mon, 27 Jan 2014 23:38:41 GMT
On 27 January 2014 21:12, Steve Hay <> wrote:
> On 27 January 2014 18:12, sebb <> wrote:
>> On 27 January 2014 09:21, Steve Hay <> wrote:
>>> On 31 December 2013 17:53, sebb <> wrote:
>>>> It's also quite unusual to have a browsable copy of the source on the
>>>> ASF mirrors.
>>>> This doubles the amount of space taken, and is not scalable.
>>>> It would be better to provide a link to the SVN tag.
>>>> Also the download page should have links to the KEYS file, and to the
>>>> hashes and sigs for each downloadable item.
>>>> Have a look at the httpd download pages for examples of how it can be presented.
>>> This is all now done in
>>> and
>> It looks like the you are not yet using svnpubsub for your release directory.
>> So either you need to copy the md5 and sha1 files to
>> /www/ or you need to raise a JIRA for infra to
>> convert you to svnpubsub.
>> Or you could do both; copy the missing files from dist/release/perl
>> now, and then svnpubsub will take over.
> I believe we are using SvnPubSub for releases - see JIRA INFRA-6707


I was going by the fact that the copy on minotaur still shows files
owned by phred rather than svnwc.
I assumed that would also be updated when the main mirrors were
changed to svnpubsub, but it seems not.

> However, I noticed myself that the files at
> do not reflect the changes I have
> made in SVN, so I have raised another JIRA regarding this already:
> INFRA-7241
>>> We just need to regenerate
>>> from
>>> the latter and we should be all good then.
>>> So does anyone know how/where (on as perlwww?) we
>>> run the script to regenerate the site from
>> As far as I can tell, you are using svnpubsub for the website, so that
>> will depend on whether you are also using CMS.
> Yes, we are using SvnPubSub for the website too (see INFRA-7055), but
> we still intend to generate the contents of the SVN folder from
> like we
> always used to, so that's what I've edited so far... I'm just not yet
> sure how/where to regenerate the website files from that. A cron job
> on used to do it. I think we need to run largely the
> same script (but outputting to a different location) and then check
> the output into SVN, but I've yet to figure out the details.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message