using id's vs names in the url

[Albert Ting]

Gervase Markham writes:
> From: Gervase Markham <gerv at mozilla.org>
> To: developers at bugzilla.org
> Subject: Re: using id's vs names in the url
> Date: Wed, 08 Dec 2004 22:55:22 +0000
> 
> Vlad Dascalu wrote:
> > Albert Ting wrote:
> > 
> >> Is it consider bad form to use id's in the URL?  That is, instead of
> >>
> >>    http://bugzilla/editproducts.cgi?action=edit&product=foobar
> >>
> >> we do
> >>    http://bugzilla/editproducts.cgi?action=edit&product_id=30
> >>  
> >>
> > I don't think that it's bad form, nope. If anything, it improves 
> > performance, shortens the URL on the average case, and eliminates 
> > encoding issues. It makes the URL harder to read on another hand, but 
> > URLs wheren't made for that I guess :)
> 
> They may not have been, but they have ended up being, and we need to 
> take account of that. I've found it handy in the past to be able to read 
> a query URL and determine exactly what the query is.
> 
> I also often (with the help of autocomplete) type
> http://bugzilla.mozilla.org/enter_bug.cgi?product=Bugzilla
> (or some other product) into my URL bar. So we shouldn't break that.

Aw shucks, so could one make products unique only within a classification?
Most of the indexing is based on just the product name, it would seem a lot
of work to change that.

Actually, one could still support both styles but for situations where it
has to be specific, use the product id.