From: Vladimir Prus (ghost_at_[hidden])
Date: 2004-06-30 02:49:01
Aleksey Gurtovoy wrote:
> Vladimir Prus writes:
> > Aleksey Gurtovoy wrote:
> >> The XLS-based regression reports have been re-worked to improve their
> >> usability and informativeness. In particular, the detailed report page
> >> is now spit into seperate pages, one per library, mainly to reduce the
> >> load time. The whole thing can be seen at
> >> http://tinyurl.com/3a45y.
> >> Comments on a further directions for improvement, as well as reports
> >> of bugs and broken links are welcome and appreciated. The broken logo
> >> and test source links will be fixed in the next run.
> > Some comments:
> > 1. The "fail" entries in user summary used to be white before, and
> > now are green. I think that the previous behaviour was better, now
> > fail entries are hard to find.
> Hmm, these used to be yellow, actually. They turned white accidently,
> and I should have checked for the original color before fixing it. Put
> on the TODO list, thanks.
> On a side note, if you provide a note that explains each such failure,
> it will both make them more visible *and* actually give user a clue
> how serious the problem is.
I believe there are notes for all failures, except for unusable msvc-stlport.
See, for example:
The notes for individual failures do not show up in the summary page; I though
it was intended this way.
> > 3. The linux regression results looks very dump now -- no colors, not
> > table decorations, and so on. I'm not sure if this is related to your
> > changes
> It is. In my annoucement I totally forgot to mention that both the XSL
> reports upload procedure and the report links need to be adjusted to
> the new directory structure. [Sorry, Martin!]
> > -- but looks like missing CSS somewhere. This happens both in
> > Konqueror and Mozilla.
> It's worse than just a missing CSS -- if you look at the report dates,
> you'll notice that these are left-overs from the old format. The new pages
> are there, though -- you just have to access them directly:
Thanks, I'll look at those....
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk