Hi,

For development involving multiple people in the same time, you cannot rely on 
hardcoded package versions in configurations, because there is no merge 
support, so you will never know when you are missing a version.

The only model that works reliably now is to depend on a baseline for 
development. That means that when you develop, you integrate relentlessly the 
latest package versions.

When you release you can use the same baseline and hardcode the package 
versions.

Cheers,
Doru


On 31 Aug 2011, at 23:02, Alexandre Bergel wrote:

> Dale and I worked on the MetacelloBrowser. I believe the UI needs to be 
> polished. Probably having something based on Glamour would cool to have.
> I worked on Versionner because I just need a commit button.
> 
> Cheers,
> Alexandre
> 
> 
> On 31 Aug 2011, at 17:57, Igor Stasenko wrote:
> 
>> is it just me, or there are more people who know nothing about it?
>> 
>> i know how to do a parallel programs,
>> but i have no means to know what happens in parallel to my little universe 
>> unless i can read it from some well known source :)
>> 
>> On 31 August 2011 23:46, Alexandre Bergel <[email protected]> wrote:
>> :-)
>> I use Versionner for simple case and the MetacelloBrowser for more 
>> complicated stuff.
>> Check at the screenshot. You see, there is a commit button :-) It saves all 
>> the dependent packages, create a new version, ask you for the comment, and 
>> save the config. I rarely use the Monticello browser actually. 
>> 
>> <Screen Shot 2011-08-31 at 17.24.53.png>
>> 
>> Cheers,
>> Alexandre
>> 
>> 
>> On 31 Aug 2011, at 17:23, Igor Stasenko wrote:
>> 
>>> On 31 August 2011 23:19, Alexandre Bergel <[email protected]> wrote:
>>>>> For development configuration is useless and just time sink.
>>>> 
>>>> I strongly disagree. Metacello was sold to please users. Actually, I use 
>>>> it for me, the developper. I create a new version at each commit. It works 
>>>> wonderfully.
>>>> 
>>> 
>>> sorry, but if during development you need to do 2 commits of 20
>>> packages , calculate how much time it would take for you to
>>> manuallly update a configuration for each update.
>>> it is time sink.
>>> 
>>> if you do that once per day, its okay.. but for every commit . no. its
>>> is absurd.
>>> 
>>>> Alexandre
>>>> 
>>>> 
>>>>> But for release it has completely different purpose: it helps us to
>>>>> identify what exactly and what version of it will be put in final
>>>>> version. And gives a simple and concrete answer how to reproduce it,
>>>>> in case of need.
>>>>> Since we're are  talking about release, i don't see how we could do
>>>>> that without configurations.
>>>>> 
>>>>> Yes, sure it could be done manually to manifest all packages and their
>>>>> version(s) included into release, but it still means that
>>>>> someone has to do it one way or another.
>>>>> 
>>>>> 
>>>>> On 31 August 2011 23:03, Stéphane Ducasse <[email protected]> 
>>>>> wrote:
>>>>>> Marcus
>>>>>> 
>>>>>> We only do what is possible :)
>>>>>> 
>>>>>>       - for the configurations of let us take what exist. We cannot 
>>>>>> control everything. I do not have the time and knowledge for that.
>>>>>>       Lukas decided that configurations are evil. Ok perfect now we fix 
>>>>>> them when we know what should be loaded.
>>>>>> 
>>>>>>       - Then for example to integrate the broken highlight I need the 
>>>>>> code first. I still do not know which slice to load in 1.3.
>>>>>>       Two days ago I spent time on it loading the wrong slice. So I 
>>>>>> decided that I will not stress.
>>>>>> 
>>>>>> So we will fix that either in 1.3 or in 1.4 as time allows it.
>>>>>> And more more important: no stress we are payed enough to stress and do 
>>>>>> stuff that hurts us.
>>>>>> 
>>>>>> Stef
>>>>>> 
>>>>>>> On Aug 31, 2011, at 8:22 PM, Lukas Renggli wrote:
>>>>>>> 
>>>>>>>> The image contains outdated versions of RB and OB that miss some
>>>>>>>> critical fixes.
>>>>>>> 
>>>>>>> How can I know?
>>>>>>> 
>>>>>>> It is is build using the latest MetaCello config... if the config needs 
>>>>>>> to be updated,
>>>>>>> it needs to be updated. It does not happen by itself.
>>>>>>> 
>>>>>>>> Also there some proposed fixes don't seem to be
>>>>>>>> included <http://code.google.com/p/pharo/issues/list?q=milestone%3D1.3>
>>>>>>>> yet?
>>>>>>>> 
>>>>>>> No... I have not yet found the energy to do that.
>>>>>>> 
>>>>>>>      Marcus
>>>>>>> 
>>>>>>> --
>>>>>>> Marcus Denker -- http://marcusdenker.de
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> --
>>>>> Best regards,
>>>>> Igor Stasenko AKA sig.
>>>>> 
>>>> 
>>>> --
>>>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
>>>> Alexandre Bergel  http://www.bergel.eu
>>>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>> 
>>> 
>>> 
>>> -- 
>>> Best regards,
>>> Igor Stasenko AKA sig.
>>> 
>> 
>> -- 
>> 
>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
>> Alexandre Bergel  http://www.bergel.eu
>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> -- 
>> Best regards,
>> Igor Stasenko AKA sig.
> 
> -- 
> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:
> Alexandre Bergel  http://www.bergel.eu
> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.
> 
> 
> 
> 
> 
> 

--
www.tudorgirba.com

"Value is always contextual."




Reply via email to