A service that crawls a given URL and reports spelling or grammar mistakes would be relatively easy. The 'lynx' and 'spell' utilities inherent in most Linux distributions would suffice for the basics.
I think the service would be better marketed as a last line of defense against linguistic embarrasment. If the 'service' found a typo, punctuation (that would be a little harder) or grammar error on a given URL it could simply e-mail a designated contact, show the error in context and suggestions to correct it.
This would be useful only in situations where ajax based "as you type" correction (or more traditional regex post-posting) is not possible, or if a site blindly imported harvested content from other sources without human checking for such errors.
Given the ease of which this could be developed, I say .. go for it.
I think the service would be better marketed as a last line of defense against linguistic embarrasment. If the 'service' found a typo, punctuation (that would be a little harder) or grammar error on a given URL it could simply e-mail a designated contact, show the error in context and suggestions to correct it.
This would be useful only in situations where ajax based "as you type" correction (or more traditional regex post-posting) is not possible, or if a site blindly imported harvested content from other sources without human checking for such errors.
Given the ease of which this could be developed, I say .. go for it.