harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Geir Magnusson Jr." <g...@pobox.com>
Subject Re: [testing] Eclipse Unit Tests run under Cruise Control
Date Wed, 27 Dec 2006 10:43:00 GMT

On Dec 27, 2006, at 12:58 AM, Egor Pasko wrote:

> On the 0x24C day of Apache Harmony Geir Magnusson, Jr. wrote:
>> On Dec 25, 2006, at 12:31 AM, Egor Pasko wrote:
>>
>>> On the 0x247 day of Apache Harmony Nina Rinskaya wrote:
>>>> Hi all,
>>>>
>>>> That's just to let you know that I run Eclipse Unit Tests under
>>>> Cruise
>>>> Control on Linux (SLES 9)/ia32 and windows (xp)/ia32. During  
>>>> last 10
>>>> days Alexey Ignatenko and I worked on enabling EUT on Harmony and
>>>> have
>>>> increased EUT pass rate from ~60% up to 93-95% (not taking into
>>>> account some intermittent failures). Many-many thanks to Alexey
>>>> Ignatenko for all his great work.
>>>>
>>>> Does it make sense to report EUT regressions to the mailing list?
>>>
>>> definitely!
>>>
>>> when your local CC becomes stable enough, it makes sense to forward
>>> it's output to alerts@ automatically. (but do not forget to  
>>> update the
>>> exclude list regularly)
>>
>> Agreed that this will be good to get posted to the list...
>>
>> but the above statement about the exclude lists confuse me.  What do
>> you mean "update the exclude list regularly"?  If this is about local
>> exclude lists for CC, we need a better solution - something general
>> so that everyone can see what is being excluded for the CC 'cloud'.
>
> IMHO, we should commit the EUT infrastructure some day so that
> everyone can do 'ant eut.fetch' and 'ant eut.test', and CC does the
> same.

Yes!  Anything anyone is running as CC should be :)

geir

>
>>>
>>>> Well, first try. Today my local Cruise Control reported significant
>>>> EUT pass rate downgrade with similar failures logs, and Eclipse
>>>> failed
>>>> to start (silently died after "Select Workspace" dialog). It looks
>>>> like it is caused by commit r489307 - just comparing classlib
>>>> rollbacks to r489307 (Eclipse dies, some EUT tests fail) and  
>>>> r489306
>>>> (Eclipse 'Hello world' passes, EUT tests pass as usually). Could
>>>> anybody take a look at this issue please?
>>>
>>> gosh! could anyone, please, rollback the r489307 ASAP? I cannot  
>>> fix it
>>> quickly :(
>>>
>>> -- 
>>> Egor Pasko
>>>
>>
>>
>
> -- 
> Egor Pasko
>


Mime
View raw message