Puntuado con 4 de 5 estrellas

One of the most useful addons, since lots of sites worth watching don't have RSS. However, needs a few improvements:

1) Handling of errors - perhaps an option to ignore a subset. redirects usually should be ignored; 404 usually counts as a change... but seriously - treating 101 ("network unreachable") as "page changed"?

2) Ignoring reading/writing of cookies, per page. Sometimes allowing e.g. auto-login or theme choice is a good thing, sometimes not. E.g.: I only want to check a forum for new threads, which can be done anonymously; but if logged in, "new posts here" variable will be reset by SiteDelta's visit, so doing it on a schedule is a bad idea. Same goes for any other "changes after your last visit" pages. It would be even better to interact with cache, but that's overcomplicated.

3) Check the pages BEFORE other addons, such as e.g. Greasemonkey scripts. So that e.g. Webcomic Reader or other custom layout script toggled on/off won't count as a change.

4) Allow watch for elements other than html body or visible divisions - sometimes changes in html head or script constants are also need monitoring, and sometimes something that can't be pointed with a mouse needs ignoring.

5) Controlled minimum delay after the browser starts. No point to slow down FF startup when it's used to look up a help file or something.

Esta valoración es de una versión anterior del complemento ( 

1) You are right, redirects are an interesting topic (follow but compare to original page? watch the destination page? mark as changed?). "101 Network Unreachable" sounds like an interesting interpretation of the HTTP protocol (101 should be "Switching Protocols").

2) Optionally ignoring cookies might be an interesting feature, however most pages will look differently without cookies so you'll always get change notifications when switching from background scan to manual visit and back again.

3) The scan has intentionally been delayed until after such modifications to make sure the page looks similar to when it is visited manually; otherwise the first automatic scan after a manual visit will cause a notification.

4) SiteDelta monitors the DOM text content of the loaded page; for your use cases (scripts, body, ...) it might be better to use a solution that compares the raw HTML code.

5) This sounds like a good idea, I'll add it to my list of feature requests.