Defect #7307
closed
HTTP 500 error on query for empty revision
Added by Etienne Massip almost 14 years ago.
Updated almost 14 years ago.
Description
On revision page, you can change the revision id then click on the 'Ok' button to get to the new revision page.
If you remove the revision id field content and leave it blank then you'll get a HTTP 500 error.
Additional question (UX) : why is the 'Ok' button displayed on the revision page and not on the repository page ?
Tested with a SVN repo.
Files
- File leanspawithacai.jpg added
- Assignee set to Maxim Krušina
- File deleted (
leanspawithacai.jpg)
- Assignee deleted (
Maxim Krušina)
Toshi MARUYAMA wrote:
This is ad hoc patch.
Do you want me to test it ?
Etienne Massip wrote:
Toshi MARUYAMA wrote:
This is ad hoc patch.
Do you want me to test it ?
I tested in Subversion.
I will research other SCMs.
- Assignee set to Toshi MARUYAMA
- Status changed from New to Closed
- Target version set to 1.1.1
- Resolution set to Fixed
Hmm. I'm using 4707 right now, and the issue doesn't seem fixed for me. I click the repository tab, click view all revisions, then in the "Revision: <input box> <ok button>" I just click the ok button without entering anything, and up pops a 500 error.
- Status changed from Closed to Reopened
r4707 will return a 500 status code indeed. Maybe we could simply redirect to the revisions list when no revision parameter is given.
I'm running r4709 and I get the same redmine 500 error I would get if I typed in a revision number which would not exist.
I don't have the Internal Server Error HTTP 500 any more (with svn, at least) so, as for myself, this issue is resolved.
Yes, it's a user friendly 500 error but still, it's a 500 error response (meaning
Internal Server Error) and this is not really appropriate. I think we should:
- response with a 404 when the revision that was entered was not found
- do a redirect to the revision list when no revision parameter is given
Jean-Philippe Lang wrote:
Yes, it's a user friendly 500 error but still, it's a 500 error response (meaning Internal Server Error) and this is not really appropriate. I think we should:
- response with a 404 when the revision that was entered was not found
- do a redirect to the revision list when no revision parameter is given
Sure, both looks like a (far for 404) better behavior.
500 error changed to 404 in r4711.
- Status changed from Reopened to Closed
Also available in: Atom
PDF
Updated by Toshi MARUYAMA 
Updated by 
Updated by 
Updated by 
Updated by