On Fri, 2010-09-17 at 13:32 -0700, Karsten Wade wrote:
On Mon, Sep 13, 2010 at 08:00:18PM -0400, Paul Frields wrote:
> Hello Docs team and Publican knowledge-meisters,
> I received this email from a contributor and tested his results. This
> appears to be a valid bug. How can we properly link to the interior of
> a document on the docs.fedoraproject.org
site? All targets seem to
> take one to the top of a document, which is not only unhelpful, but
> breaks the simple principle of least surprise.
We might also want to have SELinux Troubleshooter point to a wiki
page, and we can update that wiki page with links to the latest guide
It's a bit unclear to me where the bug lies. I see:
1. SELinux Troubleshooter is deep linking when maybe it should go to a
2. The SELinux FAQ has a section without an ID tag.
3. The SELinux FAQ has a broken link to itself.
> ---------- Forwarded message ----------
> From: Dan Haskell <geekus.org(a)gmail.com>
> Date: Mon, Sep 13, 2010 at 2:00 PM
> Subject: SE Linux docs
> Clicked on the link in the SELinux Security Alerts screen that says
> "You can generate a local policy module to allow this access - see
> FAQ" and it takes me to:
> Which is the top of the whole document. :(
> So, I find the link there that says "To see how to write a simple
> policy module, check out Local Policy Customizations. " and it takes
> me to:
> Which, again, is the top of the whole document. Sigh.
> Please don't respond with a link to the correct document. Someone
> needs to fix this. People hate SELinux enough already...
This is a tad confusing because there are multiple things going on. We
(relatively) recently reorganized docs.fp.o to support Publican
publishing and managing the site with git. As a result, a lot of links
in old documents were broken. Sometimes new documents simply replaced
f12 with f13, which means even some new documents don't point to the
right place. I believe the site essentially redirects 404's to the top
of the site.
However, this specific document has some rather peculiar behavior, which
makes me wonder just how it was produced. Even the table of contents
doesn't work reliably, something that is not evident on other documents.
In addition, some things on docs.fp.o seem not quite right. We had an
event last night where over 150,000 pages were committed and I'm not
sure we have totally recovered from that event.
But, I do think there is something peculiar about this particular
document, and it might be helpful to rebuild and republish it.
Since you are the author, did you test it before publishing? Does this
behavior exist on a local copy? Is the copy on docs.fp.o actually the
same as the one you committed?