A potential client of mine has a DotNet website, which uses some type of traffic tracking that's built into DotNet. DotNet creates an encrypted string and puts it in a hidden field of a form. The string is VERY long (24,000 - 26,000+ characters)

looking at cached versions of the homepage, google gets about 3/4 of the way through the string, then just gives up without getting to the end, and therefore never caches the rest of the page.

several months ago the string was 24k characters, and google made it to the end. But now it's 26k and google gives up.

I initially thought google was just getting to it's page size limit and calling it quits, but they only cached 77 kilobytes. Back when the string was 24k characters, Google's cached page was 137 kilobytes.

MSN and Yahoo are having no such problems, and are getting through the entire page.

Thoughts?

As a rule of thumb Google will cache 100k and then give up. I suspect this will ebb and flow based on how much storage they have.

I think the rest is just speculation. I speculate that Google might read though X amount of jibberish (ie, Microsoft's CMS and dotNet state information hidden input forms) and then decide to give up. You might then also speculate that Google will read through 2X amount of quality text (ie, from the FT or BBC) before reachings its limit.

I would put the hidden element as far down the page as possible and as much keyword rich content up towards the top of the page as possible. Needless to say, Google will not press "submit" buttons so will not be able to input the hidden data to navigate the site.

agreed, my recommendation is going to be to move this field to the bottom of the page, and possibly remove it altogether if they opt to go with a different tracking solution

yea I'm used to google stopping around 100k (although to be honest I've seen them go past that number pretty often). but in this case they stopped at 77k. when they took the whole page, they went all the way to 137 or so. interesting.