|
Boost Users : |
Subject: Re: [Boost-users] Could searching and indexing Boost docs might work better?
From: Dave Abrahams (dave_at_[hidden])
Date: 2010-09-18 09:36:19
On Thu, Sep 16, 2010 at 3:54 AM, Paul A. Bristow
<pbristow_at_[hidden]> wrote:
> But I have now squandered a few minutes trying again to find this
> information using Google (putting myself in the shoes of the OP).
>
> A search for "Running specific test units selected by their name" didn't get
> me to this page (on the first page at least).
>
> Adding site::boost.org helped reduce clutter from other unit test systems.
>
> Even "Running specific test units selected by their name" site:boost.org
> (note quotes mean search for exact text)
>
> didn't get to the current (or nearly) releases by name, and I didn't spot
> this page quickly.
>
> Using boost.org and entering "Running specific test units selected by their
> name" into the search box also didn't produce the latest docs page above,
> only out of date stuff. Limiting search to www.boost.org didn't get 1.44
> docs (only misleading ref to 1.38).
This is because of the way our versioned docs are done. You can
always get to the latest version of a page by substituting "release"
for the version number in the URL, but because that is just a
redirect, it doesn't end up in Google's index. Maybe the "release"
URLs should be the actual pages that are redirected from the latest
version number and we should be telling google not to index any of the
pages that have a version number in the URL.
> This may be why there are so many questions from puzzled users on this list?
>
> Is there anything we can do to make Google work better for us?
I think it's a matter of picking a site structure that's less
esoteric, and more like what most other projects do.
-- Dave Abrahams BoostPro Computing http://www.boostpro.com
Boost-users list run by williamkempf at hotmail.com, kalb at libertysoft.com, bjorn.karlsson at readsoft.com, gregod at cs.rpi.edu, wekempf at cox.net