12

When accessing SO with a bookmark that has a list of tags separated by or, for instance:

http://stackoverflow.com/questions/tagged/html%20or%20javascript?tab=newest&pagesize=50

When logging in from this URL, the browser ends up redirected to http://stackoverflow.com/users/login?returnurl=%2fquestions%2ftagged%2fhtml+or+javascript#log-in

Before logging in

After logging in, the URL the browser gets redirected to is http://stackoverflow.com/questions/tagged/htmlorjavascript, which obviously isn't a very popular tag.

after logging in This is Firefox 13.0.1, by the way.


UPDATE

Further testing reveals that whitespace or %20 get encoded to + when pressing the login button, and that + gets ignored when decoding the url to redirect to after the login.

But, if the initial URL contains + instead of whitespace or %20, the url when logging in contains the tags separated by %2b. They get decoded correctly to + when redirecting after the login so it works properly if the tags in the URL are separated by +.

Such a pity Firefox decodes + and %2B to whitespace when used in a bookmark.

2
  • 7
    Just you wait! Some day, the htmlorjavascript tag will be the most popular on SO, and you'll be sorry you mocked it!
    – animuson StaffMod
    Commented Jun 22, 2012 at 8:48
  • 1
    Just encountered same issue - after login the URL I was returned to had %20 and * characters removed. However, this is with Chromium, not Firefox. Commented Jan 14, 2013 at 22:41

1 Answer 1

4

This was an issue of overly aggressive URL sanitization. Fixed in the next build.

4
  • 1
    Not fixed according to this new report and I just reproduced it myself. Commented May 28, 2013 at 10:14
  • @ShaWizDowArd It seems fixed at least for URL's not including + signs. Works fine for html%20or%20javascript and html+or+javascript as of now. Didn't try using + and other characters though. Commented May 28, 2013 at 10:28
  • @XaviLópez I see. Still, the fix is not complete as there are plenty of tags containing the + character. Commented May 28, 2013 at 11:37
  • @ShaWizDowArd looks like this was an issue of the ' ' character not being encoded and decoded the same way, and in order for tags to be parsed correctly, spaces need to be preserved and not converted to a +. Should be fixed in the next build. Commented Jun 11, 2013 at 18:26

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .