Hi Ineiev and Savannah hackers,
> [ine...@gnu.org - Tue Sep 11 00:44:50 2012]:
> > Savannah hackers; can you confirm that the new.py call is done
> *after*
> > the cvs commit has successfully completed?
>
> I think there may be another possibility: "can you confirm that the
> new.py
> call is *(
> [ward - Mon Sep 17 14:01:37 2012]:
>
>
> Looks like the cvs checkout is not always completing before the new.py
> script times out. Sigh.
Here's the most recent example:
Sep 16 05:08:49 wildebeest savannah-cvs-commit: Request received...
Sep 16 05:08:49 wildebeest savannah-cvs-commit: type: w
On 09/14/2012 11:34 PM, Michael J. Flickinger via RT wrote:
> I can verify this trigger runs after something has been committed. The
> error you are encountering is caused by the script on the webserver.
Do you think the error may cause those delay in updating the pages?
On 09/14/2012 11:34 PM, Michael J. Flickinger via RT wrote:
I can verify this trigger runs after something has been committed. The
error you are encountering is caused by the script on the webserver.
Do you think the error may cause those delay in updating the pages?
On 9/14/12 1:35 AM, Pavel Kharitonov via RT wrote:
> On 09/11/2012 08:46 AM, Ineiev wrote:
>>
>> I think there may be another possibility: "can you confirm that the new.py
>> call is *(successfully) done* after the cvs commit has successfully
>> completed?"
>
> For instance, this is what CVS has j
On 9/14/12 1:35 AM, Pavel Kharitonov via RT wrote:
On 09/11/2012 08:46 AM, Ineiev wrote:
I think there may be another possibility: "can you confirm that the new.py
call is *(successfully) done* after the cvs commit has successfully
completed?"
For instance, this is what CVS has just said to
On 09/11/2012 08:46 AM, Ineiev wrote:
>
> I think there may be another possibility: "can you confirm that the new.py
> call is *(successfully) done* after the cvs commit has successfully
> completed?"
For instance, this is what CVS has just said to me on
'cvs ci -m"Update" gnu/po/thegnuproject.
On 09/11/2012 08:46 AM, Ineiev wrote:
I think there may be another possibility: "can you confirm that the new.py
call is *(successfully) done* after the cvs commit has successfully
completed?"
For instance, this is what CVS has just said to me on
'cvs ci -m"Update" gnu/po/thegnuproject.ru.po
On 09/10/2012 05:48 PM, Ward Vandewege via RT wrote:
This points to either some sort of race condition, or a bug in the order
in which the cvs commit and the new.py call are done on the savannah side.
Savannah hackers; can you confirm that the new.py call is done *after*
the cvs commit has succ
