Search Engine on start.fedoraproject.org

[Toshio Kuratomi]

On Sat, Jul 24, 2010 at 06:07:05PM -0800, Jeff Spaleta wrote:
> On Sat, Jul 24, 2010 at 5:26 PM, Toshio Kuratomi <a.badger at gmail.com> wrote:
> > I think that you're weighing the possibility of running it ourselves too
> > much compared to the convenience of outsourcing.  If it truly is compelling
> > to outsource the service, that very likely means that it will be hard for us
> > to deploy and maintain it in house.  And if that's the case, then we don't
> > really save much by mandating simple API compatibility.
> >
> 
> Explain to me again why we don't already an internal web search
> solution inside our infrastructure?
> 
There's probably a few different potential answers to this but I think the
answer is that no one who is consistently working on infrastructure 20 or
more hours a week has decided that we should implement one of the available
solutions and pushed for it to go into production.  There's been a lot of
review of search engines done and documented here:

https://fedoraproject.org/wiki/Infrastructure/Search

It appears that the front runners at the moment are:

* dataparksearch
* mnogosearch
* xapian

Having tested the xapian and nutch test instances, I think that the front
runners can probably be used to provide search for Fedora.  I'm not sure if
the people who are working on this off and on want to keep testing out
alternatives or actually pull the trigger, though.

-Toshio
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
Url : http://lists.fedoraproject.org/pipermail/advisory-board/attachments/20100725/c30f7105/attachment.bin