This is the mail archive of the
mailing list for the Cygwin project.
Re: Website: Package listing not working + cygport query
- From: Christopher Faylor <cgf-use-the-mailinglist-please at cygwin dot com>
- To: cygwin at cygwin dot com
- Date: Wed, 18 Jun 2014 01:35:13 -0400
- Subject: Re: Website: Package listing not working + cygport query
- Authentication-results: sourceware.org; auth=none
- References: <5395FA98 dot 5050005 at tiscali dot co dot uk> <20140617143111 dot GF7138 at ednor dot casa dot cgf dot cx> <20140617153200 dot GA6910 at ednor dot casa dot cgf dot cx> <53A11DDD dot 6050102 at cygwin dot com>
- Reply-to: cygwin at cygwin dot com
On Wed, Jun 18, 2014 at 12:04:29AM -0500, Yaakov Selkowitz wrote:
>On 2014-06-17 10:32, Christopher Faylor wrote:
>> I'm surprised that I haven't seen this before: The name that you used to
>> generate your packages was "David Stacey/None". My parser wasn't
>> expecting a name with a space and I didn't detect the fact that the
>> parsing failed (#($& just realized that I forgot to fix that). So it
>> was just outputting a '0' size, one per line.
>> Yaakov, wouldn't it make sense not to just use 0/0 as user/group when
>> generating packages with cygport by forcing --owner=0, --group=0 to the
>> tar command? I'm not suggesting that this is the right way to fix this
>> particular problem but I don't think it's a good idea to leak people's
>> names and groups out onto the interwebs if we can help it. It will also
>> make the packages slightly smaller.
>That would undermine any legitimate specification of gid/uid. I would
>like to see how this is handled in the Linux distros first.
I don't see how the random uid/gids of a user would ever be legitimate.
Some released rpm packages do contain (incorrectly, imo) the uid/gid of
the build system that created them but rpm does have a mechanism for
specifying ownership. If it was really important that arbitrary
ownership must be assigned for some package then it seems like that
would be another cygport enhancement since there is no easy way for
a package maintainer to force their uid to some other value.
Problem reports: http://cygwin.com/problems.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple