Seaside 3.0.4?

> On 02/14/2011 10:43 AM, Avi Shefi wrote:
>>
>> Hi,
>> Where are we standing w.r.t. Seaside 3.0.4?
>> Is it ready?
>> Cheers,
>> Avi.
>
> Avi,
>
> Today, I will start working through a slug of Gemstone-specific bugs -- so
> far I know that the recent changes in WAUrlEncoder have thrown a monkey
> wrench into some GemStone-specific changes and I figure that this is
> responsible for a bunch of the test failures, but I haven't had a chance to
> look in detail at the failures  ...

> 2011/2/14 Dale Henrichs<[hidden email]>:
>> On 02/14/2011 10:43 AM, Avi Shefi wrote:
>>>
>>> Hi,
>>> Where are we standing w.r.t. Seaside 3.0.4?
>>> Is it ready?
>>> Cheers,
>>> Avi.
>>
>> Avi,
>>
>> Today, I will start working through a slug of Gemstone-specific bugs -- so
>> far I know that the recent changes in WAUrlEncoder have thrown a monkey
>> wrench into some GemStone-specific changes and I figure that this is
>> responsible for a bunch of the test failures, but I haven't had a chance to
>> look in detail at the failures  ...
>
> [1] has been a huge win for us, if there existed something similar for
> GemStone that would totally rock and help us find issues earlier.
> Should a shell script be not good enough it's quite easy to do a
> Hudson plugin in the language that shall not be named, look in the
> Seaside svn repo for an example. We also do custom Hudson plugin
> development ;-)
>
>> The packages have been recorded for the metacello configurations, so you can
>> resume checkins...I don't know if the hudson image was saved or not...
>
> I didn't save it yet, I ran into problems with the network connection.
> If it's fixed this evening then I can download it and do the rest.
>
>   [1] http://hudson.lukas-renggli.ch/view/Seaside%203.0/job/Seaside%203.0/
>
> Cheers
> Philippe
> _______________________________________________
> seaside-dev mailing list
> [hidden email]
> http://lists.squeakfoundation.org/mailman/listinfo/seaside-dev

> On 02/14/2011 11:03 PM, Philippe Marschall wrote:
>> 2011/2/14 Dale Henrichs<[hidden email]>:
>>> On 02/14/2011 10:43 AM, Avi Shefi wrote:
>>>>
>>>> Hi,
>>>> Where are we standing w.r.t. Seaside 3.0.4?
>>>> Is it ready?
>>>> Cheers,
>>>> Avi.
>>>
>>> Avi,
>>>
>>> Today, I will start working through a slug of Gemstone-specific bugs -- so
>>> far I know that the recent changes in WAUrlEncoder have thrown a monkey
>>> wrench into some GemStone-specific changes and I figure that this is
>>> responsible for a bunch of the test failures, but I haven't had a chance to
>>> look in detail at the failures  ...
>>
>> [1] has been a huge win for us, if there existed something similar for
>> GemStone that would totally rock and help us find issues earlier.
>> Should a shell script be not good enough it's quite easy to do a
>> Hudson plugin in the language that shall not be named, look in the
>> Seaside svn repo for an example. We also do custom Hudson plugin
>> development ;-)
>>
>>> The packages have been recorded for the metacello configurations, so you can
>>> resume checkins...I don't know if the hudson image was saved or not...
>>
>> I didn't save it yet, I ran into problems with the network connection.
>> If it's fixed this evening then I can download it and do the rest.
>>
>>  [1] http://hudson.lukas-renggli.ch/view/Seaside%203.0/job/Seaside%203.0/
>>
>> Cheers
>> Philippe
>> _______________________________________________
>> seaside-dev mailing list
>> [hidden email]
>> http://lists.squeakfoundation.org/mailman/listinfo/seaside-dev
>
> Philippe,
>
> I wish I had the time .. Norbert has integrated a GemStone build into the Hudson server already, but there are additional factors.  90% of the time, there is a port/merge required to integrate the Pharo changes into the GemStone buiild structure, so the automation that is required to do that goes beyond just loading files ... and in the case of WAUrlEncoder it required a developer to untangle the initial test failures and discover the true underlying problems ...
>
> The bulk of the issues that I've run into during this episode could have been flushed out in the pharo-based hudson server with the addition of more build combinations ...

>
> On 15.02.2011, at 18:33, Dale Henrichs wrote:
>
>> On 02/14/2011 11:03 PM, Philippe Marschall wrote:
>>> 2011/2/14 Dale Henrichs<[hidden email]>:
>>>> On 02/14/2011 10:43 AM, Avi Shefi wrote:
>>>>>
>>>>> Hi,
>>>>> Where are we standing w.r.t. Seaside 3.0.4?
>>>>> Is it ready?
>>>>> Cheers,
>>>>> Avi.
>>>>
>>>> Avi,
>>>>
>>>> Today, I will start working through a slug of Gemstone-specific bugs -- so
>>>> far I know that the recent changes in WAUrlEncoder have thrown a monkey
>>>> wrench into some GemStone-specific changes and I figure that this is
>>>> responsible for a bunch of the test failures, but I haven't had a chance to
>>>> look in detail at the failures  ...
>>>
>>> [1] has been a huge win for us, if there existed something similar for
>>> GemStone that would totally rock and help us find issues earlier.
>>> Should a shell script be not good enough it's quite easy to do a
>>> Hudson plugin in the language that shall not be named, look in the
>>> Seaside svn repo for an example. We also do custom Hudson plugin
>>> development ;-)
>>>
>>>> The packages have been recorded for the metacello configurations, so you can
>>>> resume checkins...I don't know if the hudson image was saved or not...
>>>
>>> I didn't save it yet, I ran into problems with the network connection.
>>> If it's fixed this evening then I can download it and do the rest.
>>>
>>>   [1] http://hudson.lukas-renggli.ch/view/Seaside%203.0/job/Seaside%203.0/
>>>
>>> Cheers
>>> Philippe
>>> _______________________________________________
>>> seaside-dev mailing list
>>> [hidden email]
>>> http://lists.squeakfoundation.org/mailman/listinfo/seaside-dev
>>
>> Philippe,
>>
>> I wish I had the time .. Norbert has integrated a GemStone build into the Hudson server already, but there are additional factors.  90% of the time, there is a port/merge required to integrate the Pharo changes into the GemStone buiild structure, so the automation that is required to do that goes beyond just loading files ... and in the case of WAUrlEncoder it required a developer to untangle the initial test failures and discover the true underlying problems ...
>>
>> The bulk of the issues that I've run into during this episode could have been flushed out in the pharo-based hudson server with the addition of more build combinations ...
>
> Just for the record. I have a basic working solution for gemstone. I have a script that creates a stone skeleton (a directory plus files where a stone can be started). This is invoked to created a stone in a hudson workspace. I started doing something like Lukas' builder for gemstone. I called it gsbuilder by now. It assembles scripts to be run inside the gemstone image. The scripts are feeded to the hudson stone with topaz. The log you get from this is quite ok. The skeleton script can be used to create a fresh stone installation or wrap around an existing extent. So I can use copy artefacts also for gemstone.
> And I tweaked the HudsonBuildTools in a way that it can run at least the test runner (dont know if lint is supposed to work in gemstone. I build everything in pharo and in gemstone so no need to have lint tests twice) in gemstone the same way as in Lukas' builder.
>
> What it doesn't do at the moment is copying the extent for testing. So all tests are loaded into existing extent. But that can be fixed quite easy. For my testing server I rsync the last successful hudson artifacts to start a stone to test further. This involves again some scripts for two different users etc. So you can see it is a bit of work and I would need some time to make distributable. If there is a high demand I try to hurry. If someone likes to step in and help finishing I would be glad.
>
> Norbert
> Norbert
>
>
> _______________________________________________
> seaside-dev mailing list
> [hidden email]
> http://lists.squeakfoundation.org/mailman/listinfo/seaside-dev