|Forums, PR and sigs|
are rewrites worth bothering with?
| 10:39 am on Jul 9, 2004 (gmt 0)|
We run a few forums based on Snitz, they're well indexed by Google and have been around a few years and all posts have PR 0. Not too bothered about that but I was wondering the advantages / disadvatages of changing from the dynamic topic.asp?TOPIC_ID=1234567 to static URLs.
I mean, we really don't want to attract forum spammers but rewarding long term posters with links in their sigs would only seem fair - comments welcome.
| 1:42 pm on Jul 9, 2004 (gmt 0)|
I've wondered about the PR0 shown in the toolbar dynamic pages - to me, this seems to be a false indicator as I've had "PR0" pages rank well for moderately competitive keywords. I always assumed that the PR0 was a toolbar issue of some kind, particularly since there's a technique for checking PR on "PR0" dynamic pages that seems to produce a fairly logical result.
Nevertheless, static-looking URLs are nice and might improve your site's performance in other SEs. Even Google seems to like static URLs somewhat better.
Snitz's URLs don't look too messy for dynamic pages, but you could probably employ something like ISAPI Rewrite to get rid of the query string.
| 10:31 pm on Jul 9, 2004 (gmt 0)|
The Google toolbar seems not to handle URLs of dynamic pages too well, even when you display them as static. I wondered why I did well with posts that have a PR0, until I noticed the varying PR of the index page.
When "http://www.example.com/forum.cgi" sits alone in the address bar, the page has a PR of 5; when "http://www.example.com/forum.cgi/" appears, the PR is zero.
My working explanation is that the generated pages take the PR of the straight cgi page, but that the toolbar software gets confused if you ask it a complicated question. Thus, in my case, the posts all have a PR of 5, but the toolbar can't work this out.
| 4:09 pm on Jul 10, 2004 (gmt 0)|
My experience is that with a static looking URL Google and other bots tend to go deeper into content result being more pages are indexed.
PR distribution seems to spread more evenly also with static urls, but as mentioned is the an anomoly with the toolbar getting confuse or less pr passed.
| 4:28 pm on Jul 17, 2004 (gmt 0)|
|I mean, we really don't want to attract forum spammers but rewarding long term posters with links in their sigs would only seem fair - comments welcome. |
Nothing to report as yet - but due to Matt Cutts recent comments that "sig link are increasing becoming worth nothing"... I'm testing this to a near absolute.
In theory, for his statement to be absolutely true I would believe Google may have programmed discounted the sig file rather than simply looking for the unnatural external linking pattern created by multiple posts.
Another possibility is that multiple links from same site are "worth less"... e.g the first link counts full value (as whatever it was worth before), the second - less, third less even still and so on - possibility in a reverse logrithmic scale since Google PageRank is of similar logrithmic design.
The second possibility seems more reasonable - since in the same statement Matts Cutts suggested the "site wide links" (as per PageRank brokers) have a similar effect.
Nonetheless, (and not to be totally off topic) the reason I bring this up > if either is true the consideration for "rewarding long term posters" is likely under-productive.
| 7:48 pm on Jul 17, 2004 (gmt 0)|
> when "http://www.example.com/forum.cgi/" appears, the PR is zero.
"http://www.example.com/forum.cgi" denotes a file, called "forum.cgi".
"http://www.example.com/forum.cgi/" denotes a directory, called "forum.cgi/", which *may* resolve to "http://www.example.com/forum.cgi/index.html", "http://www.example.com/forum.cgi/index.php", or nothing, depending on how you have configured your server.
So, it's no wonder that those two URLs have different PR.
| 10:16 pm on Jul 17, 2004 (gmt 0)|
|no wonder that those two URLs have different PR |
They don't have different PRs, but the toolbar reports different PRs - presumably for the reasons you state.
My point is that, to check the PR, you have to be careful what URL is showing in the address bar. You have to check using a URL that is unambiguously a page, otherwise you get an inaccurate result.