[ic] Bypassing the whole [more] system

Grant interchange-users@icdevgroup.org
Sat Sep 28 15:00:01 2002


I was considering making my own system to handle my product category search
results, and I was hoping to get your opinions on the idea.

Currently, my category pages are made up of 1-3 pages of displayed items.
The searches for the displayed items are done in-page so I have 19 different
category pages, each page named with its category.  What this results in is
when a link to a product category is clicked, the URL changes to
www.mystore.com/cgi-bin/cat/category.html.  The URL is thus very clean and
when IC 5.0 is released the [mv_tmp_session] feature will allow the URL to
be perfect for search engine indexing as it will contain no query strings
(when requested by a spider).  Also, the "category" text in the URL will
make the page rank higher for a search term such as "category".  However,
the next 1-2 pages of each category display (Page 1 | Page 2 | Page 3),
which are linked to via [more] tags, have very messy URLs with something
like this:

www.mystore.com/cgi-bin/cat/scan/MM=571aedb8936b4d36ae73662be78b10b9:9:17:9?
mv_session_id=phJTbBEsm&mv_pc=5&mv_more_ip=1&mv_nextpage=page

I was considering setting up maybe 10 pages per category (they could be
almost identical, and maybe identical once I really think about it)
something like this: category_p1.html, category_p2.html, etc.  I could then
structure the pages myself to only show nine (my match limit) items per page
and generate the appropriate "Page 1 | Page 2 | Page 3" links depending on
how many items are returned from the in-page category search.  This would
mean that ALL of my category pages are search-engine friendly for the two
reasons mentioned above, AND all of my products on their flypages which are
only linked to in the pages beyond Page 1 of the category pages will be
indexed when they otherwise would not.

What do you guys think?

- Grant