Open Stack

At least with the installation guide, I think people generally assume that
/trunk contains corrections and other improvements for the stable
documentation rather than functionality for the next release. When I first
tried installing OpenStack, most people pointed me at /trunk and I wound up
using a combination of the latter and stable branch to build my first
functional environment. We probably haven't seen many bug reports on /trunk
until now because most of the basic configuration for core components
except networking remained the same (or similar enough to work) through
Grizzly, Havana, and Icehouse. We definitely need to provide better
indication of the target audience for /trunk, but I also think several
solid releases of the installation guide will cause more people to focus on
the stable branch.


On Fri, Jun 20, 2014 at 6:37 AM, Anne Gentle <anne at openstack.org> wrote:

> Tom's patch is fine, let's see if that causes any rage.
>
> How about this: for the install guide we just don't publish to /trunk.
>
> I also like the django header but when I suggested using our existing
> DRAFT watermark in the 1191447 but someone (can't recall who) said it
> wouldn't be enough.
>
> The /trunk content was meant to be much less findable but over the years
> it has become sought. It's probably time to delete it entirely now that we
> have docs-draft links always available.
> What do you think?
> Anne
>
>
> On Fri, Jun 20, 2014 at 2:44 AM, Gauvain Pocentek <
> gauvain.pocentek at objectif-libre.com> wrote:
>
>> Le 2014-06-20 09:18, Tom Fifield a écrit :
>>
>>  On 20/06/14 15:13, Gauvain Pocentek wrote:
>>>
>>>> We've received several bug reports about the auth_strategy nova option
>>>> missing from our install procedure these last few days. These bugs are
>>>> reported against the trunk doc, by people installing icehouse. The doc
>>>> is in fact correct, both in trunk (the default for this option in juno
>>>> has changed and we don't need to set it anymore), and in the icehouse
>>>> branch (where the option is set up in the procedure).
>>>>
>>>> Would it be possible to mark the pages in the trunk/ subdirectory of
>>>> docs.o.org as "in progress" or "unstable", or anything that would
>>>> explicitly tell the user that he might not be reading the correct page?
>>>> We already have "juno" in the install-guide and config-ref titles, but I
>>>> don't think it's visible enough.
>>>>
>>>
>>>
>>> This is indeed a problem.
>>>
>>> I find it's quite similar to our Critical bug "Clearly mark outdated
>>> doc pages" https://bugs.launchpad.net/openstack-manuals/+bug/1191447
>>>
>>>
>>> I've proposed a patch @ https://review.openstack.org/#/c/101428/ that
>>> will make it more obvious that trunk is juno, but I don't think it
>>> solves the problem.
>>>
>>
>> Users following the guide will very soon realize that this is not what
>> they want with this patch. IMO that's a good thing. With the current
>> trunk/, I'm quite sure that people end up thinking "the doc is not good, I
>> can't get this thing to work". We really don't want them to think that :)
>>
>> When starting this thread I had the django documentation website in head.
>> They have a relatively visible warning on the top of dev version pages:
>> https://docs.djangoproject.com/en/dev/ . If Tom's patch is not enough,
>> we could look into implementing a similar feature.
>>
>> Gauvain
>>
>>
>> _______________________________________________
>> Openstack-docs mailing list
>> Openstack-docs at lists.openstack.org
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-docs
>>
>
>
> _______________________________________________
> Openstack-docs mailing list
> Openstack-docs at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-docs
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-docs/attachments/20140620/dcc25f4a/attachment.html>