#2452 refs are not updated after a change in pull-request
Closed: Fixed Opened by lslebodn.

There was a request to change few things in pull-request
https://pagure.io/SSSD/sssd/pull-request/3456#comment-30562

However, if I fetch changes from pagure in the same way as I do with github
then I get old patch.

How to reproduce:

sh$ git clone https://pagure.io/SSSD/sssd.git
Cloning into 'sssd'...
remote: Counting objects: 83609, done.
remote: Compressing objects: 100% (16311/16311), done.
remote: Total 83609 (delta 69173), reused 80795 (delta 66977)
Receiving objects: 100% (83609/83609), 26.23 MiB | 872.00 KiB/s, done.
Resolving deltas: 100% (69173/69173), done.
sh$ cd sssd/
sh$ git fetch origin pull/3456/head:pr3456_typo
remote: Counting objects: 29, done.
remote: Compressing objects: 100% (14/14), done.
remote: Total 15 (delta 13), reused 3 (delta 1)
Unpacking objects: 100% (15/15), done.
From https://pagure.io/SSSD/sssd
 * [new ref]             refs/pull/3456/head -> pr3456_typo
sh$ git log -1 -p pr3456_typo | grep "responses sent by the"

Actual output:

+    'pam_response_filter' : _('Filter PAM responses sent by the pam_sss'),

Expected output:
//empty output because latest version does not contain it.
https://pagure.io/SSSD/sssd/pull-request/3456#request_diff


Metadata Update from @pingou:
- Issue tagged with: bug

Ok I can reproduce on the pagure project, will try to reproduce on locally as well

And I can reproduce locally but I'm running into issues with force push to update the ref :(

Metadata Update from @pingou:
- Issue assigned to pingou

Figured a way around it, up fore review at: https://pagure.io/pagure/pull-request/2454

@pingou FYI. I am not sure whether it is related but comment which is in description https://pagure.io/SSSD/sssd/pull-request/3456#comment-30562 does not exist anymore. And old pull-request is gone as well. There is a new pull-request with different content but the same ID

Could you explain it?

The comments can be deleted by the author or the admins but not notifications and if the user force-pushed to their branch the content could have changed but then we would have notifications of rebases/updates. So this sounds weird.

Do you have traces of something in the git repo for requests ? (like finding two requests w/ the same ID but different uuid?)

The comments can be deleted by the author or the admins but not notifications and if the user force-pushed to their branch the content could have changed but then we would have notifications of rebases/updates. So this sounds weird.

I got notification of new pull-request

Date: Thu, 27 Jul 2017 09:00:30 +0000 (UTC)                                                               
From: Yuri Chornoivan <pagure@pagure.io>                                                                  
To: lslebodn@redhat.com                                                                                   
Subject: [SSSD/sssd] PR #3456: Fix minor typos                                                            
yurchor opened a new pull-request against the project: `sssd` that you are following:                     
``                                                                                                        
Fix minor typos                                                                                           
``                                                                                                        
To reply, visit the link below or just reply to this email                                                
https://pagure.io/SSSD/sssd/pull-request/3456

Do you have traces of something in the git repo for requests ? (like finding two requests w/ the same ID but different uuid?)

Yes there are two files with the same ID

sh-4.4$ git clone ssh://git@pagure.io/requests/SSSD/sssd.git
Cloning into 'sssd'...
remote: Counting objects: 563, done.
remote: Compressing objects: 100% (563/563), done.
remote: Total 563 (delta 360), reused 142 (delta 0)
Receiving objects: 100% (563/563), 86.07 KiB | 0 bytes/s, done.
Resolving deltas: 100% (360/360), done.
sh-4.4$ cd sssd/
sh-4.4$ 
sh-4.4$ grep -RniI 3456 *
ac353e349f844eee85425f0fdc4f8832:82:    "id": 3456,
bca734b6b77e41398bbc045fd1cb2e8b:166:    "id": 3456,

Do you need to see them?

The notification seems in line with the PR I'm seeing at: https://pagure.io/SSSD/sssd/pull-request/3456 same creation time, same author, same title.

I'd be interested in the content of these two files but we can discuss about them on IRC if you prefer.

@yurchor when trying to debug the second issue mentioned in this ticket (comments disappearing) I am seeing some information that make me wondering if you did something special.

Did you ?

From what I see, it looks like your fork got deleted and re-created, is that what happened?

@yurchor when trying to debug the second issue mentioned in this ticket (comments disappearing) I am seeing some information that make me wondering if you did something special.
Did you ?
From what I see, it looks like your fork got deleted and re-created, is that what happened?

Yes, that's right.

I'm not a git guru. Editing the commit it's mostly a top action that I know. When the things went unpredictably (issue 1) I have decided to scratch all over and make a clean start. So I had deleted the old fork and started a new one. Then committed the new patch as it was recommended by lslebodn.

Commit cd4dad65 fixes this issue

Metadata