On Wed, 30 Oct 2013 18:39:01 +0000
Frankie Onuonga <frankie.onuonga(a)gmail.com> wrote:
Hi guys,
I hope all is well.
I am writing in regards to something I had written about two months
ago. This in regards to the search engine.
https://fedorahosted.org/fedora-infrastructure/ticket/1055#trac-add-comment
I think we need to seriously fix this thing once and for all.
That would be nice. ;) and in a sustainable way. :)
I am looking at designing a solution for us.
I however would like to start from the basics.
We will need to look at the solutions that are present at the current
time. This should allow us to critic and analyse them.
I will then take this issue into the meeting for voting and whatever
is decided is what I shall implement.
Kindly do advice if you have any pointers on this project.
I however must be honest and say it will take me about 2 months to
complete when i start because I have a day job elsewhere.
This does not mean I will not want assistance. As always assistance is
highly appreciated.
Great.
So, really to go back to the beginning requirements here:
- Crawler that can index our:
MUST: wiki, docs, meetbot logs (These should be in the first cut)
Might be nice: fedorahosted projects, pkgs git repos, packages,
tagger, etc
- DB: postgres
- Friendly and usable web frontend people can search with.
- Nice if it doesn't use much resources to crawl or search and can be
tuned.
IMHO, it doesn't need to be ultra fancy, but it does need to handle
docs, wiki and meetbot well (as those are the things where search right
now is bordering impossible).
just my 2 cents
kevin