I opened some issues...

> On Feb 8, 2022, at 11:25 AM, Zoran Regvart <zo...@regvart.com> wrote:
> 
> Hi David & Cameleers,
> 
> On Tue, Feb 8, 2022 at 7:57 PM David Jencks <david.a.jen...@gmail.com 
> <mailto:david.a.jen...@gmail.com>> wrote:
>> We have a problem in a lot of blog posts that they link into the Antora 
>> generated docs.
> 
> Yes this happens with increasing frequency and it is problematic
> 
>> Many link to the “next” version of the docs which has two problems:
>> 
>> - as release announcements,  they should link to the doc version that’s been 
>> released.
>> - there’s no reason to suppose any particular page in ‘next’ is going to 
>> stay there.
> 
> We should agree never to link to `next` or `latest`, but to concrete
> versions. A validation rule to that effect can be implemented to
> enforce this.

https://github.com/apache/camel-website/issues/779 
<https://github.com/apache/camel-website/issues/779> deals with fixing existing 
blog posts.
https://github.com/apache/camel-website/issues/780 
<https://github.com/apache/camel-website/issues/780> suggests adding a rule to 
keep us out of trouble in the future.
> 
>> From this point of view we should change all the release announcement links 
>> to point to the docs in the released version. That’s an improvement, but it 
>> will cause another problem:
>> 
>> - At the moment our plan is to drop the docs for no-longer-supported 
>> versions when they aren’t used by supported versions of other subprojects.
>> 
>> So, eventually all the links to docs in blog posts will break.
>> 
>> I think we should:
>> 
>> - change all blog post links to point to explicit appropriate doc versions.  
>> I’m somewhat afraid about how big a job this will be, but the current links 
>> to “next” are clearly wrong in all cases.
>> 
>> - Decide what to do about old versions.  I can think of three possibilities:
>> 
>> — Keep all the old docs versions in the docs, perhaps marking them 
>> “unsupported”.  There might be ways to do this without building those 
>> portions of the website every time but this would take some time and thought 
>> and wouldn’t be my highest priority for some time.
>> — Delink broken links in the blog, marking them something “no longer 
>> documented” or “obsolete” or something.
>> — Remove the blog posts about no-longer-documented versions.
>> 
>> I really don’t know what to do here, so I’m hoping discussion will result in 
>> a “best path forward”.
> 
> I'd like to be in a place where we could keep the old versions of the
> documentation, given that we've split off the website source and
> publish repositories. Perhaps there is a way forward by keeping more
> content in the publish repositories, even if it is not built/linked
> from the site. As a form of archived documentation. For that we would
> need to keep everything within the "/_" (CSS, images...) and when
> publishing instead of deleting everything[1] we delete only the built
> versions.

I also would really like to be able to keep all the old doc versions somehow. 
Around some details, either I don’t quite understand you or slightly disagree 
on what is needed or will work.  I opened 
https://github.com/apache/camel-website/issues/781 
<https://github.com/apache/camel-website/issues/781> to try to explain my 
thinking on what we could do.

> 
> The drawback of this approach is that it might overwhelm whatever is
> done outside of our purview to publish the site, but we can try and
> see if it really does and engage INFRA to explore options.

It seems implausible to me that even a few GB of infrequently accessed static 
content would be a problem, and I think we should start by solving the problems 
of how to avoid generating the entire site on every build.

thanks!
David Jencks
> 
> zoran
> [1] 
> https://github.com/apache/camel-website/blob/b978a3e6af3fcc97808515bc01a004ddd28426a6/Jenkinsfile#L99
>  
> <https://github.com/apache/camel-website/blob/b978a3e6af3fcc97808515bc01a004ddd28426a6/Jenkinsfile#L99>
> -- 
> Zoran Regvart

Reply via email to