Hi Adam,

You know, I hadn't thought about trying this, but I may give it a shot. 
Thanks for the idea!

Brian

On Tuesday, June 17, 2014 1:02:58 PM UTC-7, Adam wrote:
>
> Hi Brian,
>
> Have you tried using the bzr-diff-revid  plugin for Bazaar?
>
> https://launchpad.net/bzr-diff-revid
>
> From the description on the LP page:
>
> "bzr-diff-revid is a plugin for the Bazaar VCS. It tweaks the output of 
> "bzr diff": When possible, it includes the revid of the file version in the 
> diff header, instead of the timestamp. This works with Review Board's bzr 
> support to make diff loading reliable"
>
>
> When using bzr with ReviewBoard I found this plugin avoids the very issues 
> you describe.
>
> Adam
> On 17 Jun 2014, at 22:41, Brian <phxg...@gmail.com <javascript:>> wrote:
>
> My hope at this point is that there was something wrong with the 
> repository. We've used this combination of tools for a couple of years with 
> no issue, even after a couple of server upgrades/reinstalls. 
>
> On Tuesday, June 17, 2014 12:12:47 PM UTC-7, Christian Hammond wrote:
>>
>> Hi Brian,
>>
>> So, I guess Bazaar requires a local timestamp when doing a 'date:’ query. 
>> Review Board attempts to convert the date in the diff to a local date that 
>> would match it. It sounds like something’s going terribly wrong there. I 
>> toyed with it briefly, but couldn’t reproduce the problem you’re hitting.
>>
>> Christian
>>
>> -- 
>> Christian Hammond - chri...@beanbaginc.com
>> Review Board - http://www.reviewboard.org
>> Beanbag, Inc. - http://www.beanbaginc.com
>>
>> On June 17, 2014 at 12:07:13 PM, Brian (phxg...@gmail.com) wrote:
>>
>> Hi Christian,
>>
>> So far I haven't found a pattern in the timestamps that are causing 
>> issues. I have upgraded one of our servers to use RB 2.0.2 and I also 
>> recreated the Bazaar repository that we were using. There are only two of 
>> us working on this project and very little history, so we didn't really 
>> lose anything. I'm hoping that our problem either goes away, or I will be 
>> able to find something when I check a diff after each commit we do.
>>
>> I'll keep you up to date on what I can find out.
>>
>> Thanks for looking into this!
>> Brian
>>
>> On Tuesday, June 17, 2014 3:40:28 AM UTC-7, Christian Hammond wrote: 
>>>
>>>  Hi Brian,
>>>  
>>>  Spent a little bit of time this evening trying to debug this, but no 
>>> real luck.
>>>  
>>>  Can you see if there’s anything in common with the timestamps where 
>>> things fail? Do they wrap around to a new day, new month, anything?
>>>  
>>>  Christian
>>>  
>>>   -- 
>>>  Christian Hammond - chri...@beanbaginc.com
>>>  Review Board - http://www.reviewboard.org 
>>> Beanbag, Inc. - http://www.beanbaginc.com
>>>  
>>> On June 12, 2014 at 3:44:13 PM, Brian (phxg...@gmail.com) wrote:
>>>
>>>  I can tell you what we know so far. When we have this issue, I trace 
>>> the "bad revision" back to what I can find in the Bazaar log for the file. 
>>> We are UTC-7, and the timestamp is usually 7 hours off from a match. This 
>>> may be the case for all files/revisions, I'm not sure yet. Most (if not 
>>> all) of our development machines are set to UTC, but maybe there is one 
>>> somewhere that is not and is affecting this. Typically we can edit the diff 
>>> to match the time in the bzr log and Review Board handles the diff with no 
>>> problem.
>>>
>>> In the last week, we have see an instance or two where this workaround 
>>> has not been successful.
>>>
>>> Separate, but maybe related, we have also had a couple of instances 
>>> where 1 of 24, or 2 of 6, or 4 of 4 diffs in a file did not apply cleanly. 
>>> We're not sure what happened with that exactly, but the timestamps in the 
>>> diff and log were off by 7 hours in those files too.
>>>
>>> I'm actually waiting for my IT person to make a copy of our server 
>>> template right now so I can debug this. I'm planning on getting things set 
>>> up to (hopefully) recreate the issue, then will try new versions of Review 
>>> Board to see if that helps any. If not, I'll look into creating a clean 
>>> Bazaar repository to see if that helps any.
>>>
>>> I'm also open to any other ideas you may have!
>>>
>>> Thanks,
>>> Brian
>>>
>>> On Thursday, June 12, 2014 3:26:31 PM UTC-7, Christian Hammond wrote: 
>>>>
>>>>  Hi Brian,
>>>>  
>>>>  Can you go into more details on how the timestamps affect things? I’d 
>>>> love to find a way to fix/workaround this in RBTools and/or Review Board.
>>>>  
>>>>  Christian
>>>>  
>>>>   -- 
>>>>  Christian Hammond - chri...@beanbaginc.com
>>>>  Review Board - http://www.reviewboard.org 
>>>> Beanbag, Inc. - http://www.beanbaginc.com
>>>>  
>>>> On June 12, 2014 at 3:23:28 PM, Brian (phxg...@gmail.com) wrote:
>>>>
>>>>  Hi Rahul,
>>>>
>>>> The way we have worked around it so far is to manually edit the diff. 
>>>> We find the timestamp in the bzr log that it "should" be (UTC in our 
>>>> case), 
>>>> then edit the diff to match. This had been ok for a while but it is not 
>>>> working more and more often now. I'm actually about to start debugging it 
>>>> a 
>>>> little more, I'm not entirely convince it's a Bazaar issue.
>>>>
>>>> I'll post more if I can figure out what is going on.
>>>>
>>>> Brian
>>>>
>>>> On Thursday, June 12, 2014 4:47:01 AM UTC-7, Rahul Sisondia wrote: 
>>>>>
>>>>> Hi Brian,  
>>>>>
>>>>> I am also facing same issue. So what is the solution or workaround for 
>>>>> this problem ? 
>>>>>
>>>>> Thanks 
>>>>> Rahul 
>>>>>
>>>>> On Thursday, April 17, 2014 2:28:29 AM UTC+5:30, Brian wrote: 
>>>>>>
>>>>>> Following up on this... we found that the developer creating the diff 
>>>>>> had a different time zone on his machine than the server. So it was 
>>>>>> actually a Bazaar issue, not being able to match the times in the 
>>>>>> checkouts.
>>>>>>
>>>>>> On Friday, February 28, 2014 12:46:16 PM UTC-7, Brian wrote: 
>>>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> We are using RB 1.7.12 and twice in the last two days we have gotten 
>>>>>>> errors when uploading a diff of a Bazaar branch saying the requested 
>>>>>>> revision does not exist:
>>>>>>>
>>>>>>> Requested revision: 'date:2014-02-11 22:54:13' does not exist in 
>>>>>>> branch: <<path>>
>>>>>>> Requested revision: 'date:2014-02-11 23:00:10' does not exist in 
>>>>>>> branch: <<path>>
>>>>>>>
>>>>>>> Removing the diff section that this part of the diff refers to 
>>>>>>> allows the diff to be uploaded, but this is obviously a non-optimal 
>>>>>>> solution.
>>>>>>>
>>>>>>> We have been using RB for a couple of years and have never seen this 
>>>>>>> until the last couple of days.
>>>>>>>
>>>>>>> Any idea what might be happening?
>>>>>>>
>>>>>>> Thanks,
>>>>>>> Brian
>>>>>>>
>>>>>>    --
>>>> Get the Review Board Power Pack at 
>>>> http://www.reviewboard.org/powerpack/
>>>> ---
>>>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>>>> ---
>>>> Happy user? Let us know at http://www.reviewboard.org/users/
>>>> ---
>>>> You received this message because you are subscribed to the Google 
>>>> Groups "reviewboard" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>> an email to reviewboard...@googlegroups.com.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>  
>>>>   --
>>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>>> ---
>>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>>> ---
>>> Happy user? Let us know at http://www.reviewboard.org/users/
>>> ---
>>> You received this message because you are subscribed to the Google 
>>> Groups "reviewboard" group.
>>> To unsubscribe from this group and stop receiving emails from it, send 
>>> an email to reviewboard...@googlegroups.com.
>>> For more options, visit https://groups.google.com/d/optout.
>>>  
>>>   --
>> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
>> ---
>> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
>> ---
>> Happy user? Let us know at http://www.reviewboard.org/users/
>> ---
>> You received this message because you are subscribed to the Google Groups 
>> "reviewboard" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to reviewboard...@googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>>
>>  -- 
> Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
> ---
> Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
> ---
> Happy user? Let us know at http://www.reviewboard.org/users/
> --- 
> You received this message because you are subscribed to the Google Groups 
> "reviewboard" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to reviewboard...@googlegroups.com <javascript:>.
> For more options, visit https://groups.google.com/d/optout.
>
>

-- 
Get the Review Board Power Pack at http://www.reviewboard.org/powerpack/
---
Sign up for Review Board hosting at RBCommons: https://rbcommons.com/
---
Happy user? Let us know at http://www.reviewboard.org/users/
--- 
You received this message because you are subscribed to the Google Groups 
"reviewboard" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to reviewboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to