Maniphest T68156

"Help -> Report a Bug" throws "400 Bad Request" message
Closed, Resolved

Assigned To
Sergey Sharybin (sergey)
Authored By
rudy serna (Rooster)
Aug 3 2019, 1:03 AM
Tags
  • BF Blender
Subscribers
Dan McGrath (dmcgrath)
Demeter Dzadik (Mets)
Joseph Brandenburg (TheAngerSpecialist)
Ray Molenkamp (LazyDodo)
rudy serna (Rooster)
Sergey Sharybin (sergey)
Stephen Swaney (stiv)

Description

System Information
Operating system: macOS High Sierra
Graphics card:

Blender Version
Broken: 2.80 2019-07-19
Worked: (optional)

tried to submit a bug report using "Help -> Report a Bug" and received an error message "400 Bad Request"

Exact steps for others to reproduce the error - - try submitting a bug request using Help...
Based on the default startup or an attached .blend file (as simple as possible).

Event Timeline

rudy serna (Rooster) created this task.Aug 3 2019, 1:03 AM
Ray Molenkamp (LazyDodo) added a subscriber: Ray Molenkamp (LazyDodo).Aug 3 2019, 1:10 AM

it's ok until the %0A then it http 400's

Ray Molenkamp (LazyDodo) assigned this task to Sergey Sharybin (sergey).Aug 3 2019, 2:00 AM
Ray Molenkamp (LazyDodo) added a subscriber: Sergey Sharybin (sergey).

@Sergey Sharybin (sergey) as the local phab guru mind taking a peek here?

Demeter Dzadik (Mets) added a subscriber: Demeter Dzadik (Mets).Aug 3 2019, 2:06 AM

Duplicate of T68147, which might be a duplicate of T68007.

Joseph Brandenburg (TheAngerSpecialist) added a subscriber: Joseph Brandenburg (TheAngerSpecialist).Aug 3 2019, 2:19 AM

Happens for me, too, and I don't think it's Blender. I have reported bugs from 2.80 and 2.80 RC3 before, and both are giving me the 400 error right now. It's on the maniphest/phabricator or whatever this site is called.

Dan McGrath (dmcgrath) added a subscriber: Dan McGrath (dmcgrath).Aug 3 2019, 4:47 AM

The logs show:

2019-08-03 04:45:26: (response.c.346) invalid character in URI -> 400 /maniphest/task/edit/form/1?description=%2A%2ASystem%20Information%2A%2A%0AOperating%20system%3A%20Linux-4.15.0-55-generic-x86_64-with-Ubuntu-18.04-bionic%2064%20Bits%0AGraphics%20card%3A%20GeForce%20GTX%20960/PCIe/SSE2%20NVIDIA%20Corporation%204.5.0%20NVIDIA%20390.116%0A%0A%0A%2A%2ABlender%20Version%2A%2A%0ABroken%3A%20version%3A%202.80%20%28sub%2075%29%2C%20branch%3A%20heads/v2.80%20%28modified%29%2C%20commit%20date%3A%202019-07-29%2014%3A47%2C%20hash%3A%20%60rBf6cb5f54494e%60%0AWorked%3A%20%28optional%29%0A%0A%0A%2A%2AShort%20description%20of%20error%2A%2A%0A%5BPlease%20fill%20out%20a%20short%20description%20of%20the%20error%20here%5D%0A%0A%2A%2AExact%20steps%20for%20others%20to%20reproduce%20the%20error%2A%2A%0A%5BPlease%20describe%20the%20exact%20steps%20needed%20to%20reproduce%20the%20issue%5D%0A%5BBased%20on%20the%20default%20startup%20or%20an%20attached%20.blend%20file%20%28as%20simple%20as%20possible%29%5D%0A%0A
Dan McGrath (dmcgrath) added a comment.Aug 3 2019, 4:50 AM

An old Blender 2.77 redirects to https://developer.blender.org/maniphest/task/create/?project=2&type=Bug and shows:

Ray Molenkamp (LazyDodo) added a comment.Aug 3 2019, 4:55 AM

Looks like a lighttpd error, perhaps relevant?

http://redmine.lighttpd.net/boards/2/topics/8610

Dan McGrath (dmcgrath) added a comment.Aug 3 2019, 5:06 AM

I'm wondering if it's:

http://redmine.lighttpd.net/projects/lighttpd/wiki/Server_http-parseoptsDetails

We run 1.4.54, which we just upgraded to, and that page mentions:

  • "method-get-body" => "enable" (since 1.4.54)
    • permit GET sent with request body (default: reject with 400 Bad Request)
Dan McGrath (dmcgrath) added a comment.Aug 3 2019, 5:17 AM

Okay, seems that specifically it was:

  • server.http-parseopts = ( "url-ctrls-reject" => "disable" )

After a reload, the page loads fine now. The question is, why was that done? The original bug report was one guy asking a valid "WHY", and the other guy taking offense, but not really answering the question, so I really don't know about any consequences of this. That said, I don't feel that we can afford to break bug reporting for _every_ Blender release, do you? :)

I'll leave you guys to close the report, but I will mention that the old 2.77 URL is still broken, so perhaps we should keep it open until we figure out what is going on there, as well.

rudy serna (Rooster) added a comment.Aug 5 2019, 7:30 PM

no sorries - CLOSE it

rudy serna (Rooster) added a comment.Aug 6 2019, 1:09 AM

to LazyDodo; for myself, please close this report - thanks

Sergey Sharybin (sergey) changed the task status from Unknown Status to Resolved.Sep 4 2019, 11:44 AM

Seems to be fixed now.
Thanks everyone, closing! :)