I've wanted to write a reply but been short of time.
I'm using generateSitemap.php every night via a cron job but it isn't easy.
First off I never got any of the php maintainance scripts to run reliably on
shared hosts via PHPShell (I was on LunarPages) but that might have been
just me. I'm now on a VPS so I have shell access and things that I had
pulled my hair out on shared hosting are now simple: using SVN to update,
runing update.php etc.
You might have some luck running scripts via cron. I did manage this a
couple of times on lunarpages but setting this up is a real pain, waiting a
minute for them to run and then checking email for feedback. Painful.
As for generateSitemap.php it has some bugs. I've patched mine, submitted my
patch to bugzilla but don't really know what more to do. In particular I
found it not working properly for giving the full path to sub-sitemaps in
the index sitemap. I fixed this with an extra command line option letting me
specify a full path. There's probably some server variable and regexp I
could have deduced but I only have one life and I'd rather just slap
something in a command line.
The bug I've worked around is here:
You can also find my version of the file attached to that post.
The current version also gives every link the same priority. It would be
nice if it didn't as this makes google upset. I haven't had the energy or
spare time to look into this but it would propbably be easy-ish to set the
priority based on number of views.
I hope someo of that helps.
GD
On Thu, Sep 11, 2008 at 9:46 PM, Andrew Miller <andrew(a)richmondwiki.org>wrote;wrote:
That's what I was afraid of. Thanks for the input,
that helps me narrow
down
the list of potential problems.
I will keep investigating sitemap generation tools and report anything
worthwhile.
Andrew
On Thu, Sep 11, 2008 at 1:55 PM, Garth Patil <garthpatil(a)gmail.com> wrote:
I use the generateSitemap.php maintenance script
in MW 1.11.2, and it
takes ~30 seconds for 130,000 pages. However, this is on a dedicated
server, running from the command line.
/Garth
On Wed, Sep 10, 2008 at 8:50 PM, Andrew Miller <andrew(a)richmondwiki.org>
wrote:
Hi,
Anybody have experience with the generateSitemap.php maintenance script
in
> MW 1.12 for building sitemap.xml files? How long does it normally take
to
> run? I've tried running it from PHPShell
(I host on Network Solutions
so
no
> command line) but it times out after a few minutes.
>
> Unfortunately all of the extensions for generating sitemaps seem to be
> outdated and no longer supported since this maintenance script was
added.
> Any other solutions?
>
> Thanks in advance,
> Andrew