How to organize Moose on SmalltalkHub
Locked
How to organize Moose on SmalltalkHub
 
|
Hi doru
I have some scripts (available in the wonderful chapter on Gofer) to migrate code from ss and ss3 to SmalltalkHub. Now the question is how do we organize moose? I propose to create a team Moosers? and Moose as a project Then what do we put underneath? Fame as subproject? Mondrian Metanool Stef _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
Re: How to organize Moose on SmalltalkHub
|
|
Hi,
I already created a Moose team, and a Moose project inside. I see the forces as follows: - We need to have each engine individually loadable because they are often used in orthogonal projects - There are projects that are developed separately (like Roassal), and thus these are better living outside of the Moose team repository for the moment. - Projects that are either no longer maintained, like Fame, or developed only within the context of Moose, like MooseAlgos, should go into the Moose team repository. Nevertheless, I think each of these projects would better have their own repository inside the Moose team repository. - Moose is the platform, not the actual analyses. Thus, the engines should be grouped in a separate configuration. We can call it MooseEngines. Right now, this would include: Fame, Magritte, Metanool, Glamour, Arki, Merlin, Roassal, (Mondrian for a short time), EyeSee, MooseAlgos, PetitParser, MooseCore, the Core of MooseFinder. - FAMIX should become a family of meta-models and should receive its own configuration and repository with everything related inside: FAMIX-Core, FAMIX-Extensions, FAMIX-Java, FAMIX-Smalltalk, Hismo, Dynamix. In the same repository, I would add also FAMIXAnalysis to include various extensions such as MondrianPaintings or RoassalPaintings or specific browsers and algorithms. - At the moment, we still need to clean DSM and SmallDude. Until they are not cleaned, I would not migrate them as they are. What do you think? Cheers, Doru On Jan 29, 2013, at 1:11 AM, Stéphane Ducasse <[hidden email]> wrote: > Hi doru > > I have some scripts (available in the wonderful chapter on Gofer) to migrate code from ss and ss3 to SmalltalkHub. > Now the question is how do we organize moose? > > I propose to create a team Moosers? and Moose as a project > > Then what do we put underneath? > > Fame as subproject? > Mondrian > Metanool > > > Stef > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- www.tudorgirba.com "Every now and then stop and ask yourself if the war you're fighting is the right one." _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
|
|
> I see the forces as follows:
> - We need to have each engine individually loadable because they are often used in orthogonal projects Yes yes! > - There are projects that are developed separately (like Roassal), and thus these are better living outside of the Moose team repository for the moment. it makes sense. I have created a team ObjectProfile. It will be in it. > - Projects that are either no longer maintained, like Fame, or developed only within the context of Moose, like MooseAlgos, should go into the Moose team repository. Nevertheless, I think each of these projects would better have their own repository inside the Moose team repository. Yes, this is a nice step toward getting a better modularity. > - Moose is the platform, not the actual analyses. Thus, the engines should be grouped in a separate configuration. We can call it MooseEngines. Right now, this would include: Fame, Magritte, Metanool, Glamour, Arki, Merlin, Roassal, (Mondrian for a short time), EyeSee, MooseAlgos, PetitParser, MooseCore, the Core of MooseFinder. Yes. > - FAMIX should become a family of meta-models and should receive its own configuration and repository with everything related inside: FAMIX-Core, FAMIX-Extensions, FAMIX-Java, FAMIX-Smalltalk, Hismo, Dynamix. In the same repository, I would add also FAMIXAnalysis to include various extensions such as MondrianPaintings or RoassalPaintings or specific browsers and algorithms. Why not FAMIXExtra (or FAMIXExtension) instead of FAMIXAnalysis ? > - At the moment, we still need to clean DSM and SmallDude. Until they are not cleaned, I would not migrate them as they are. Make sense Alexandre > > > > On Jan 29, 2013, at 1:11 AM, Stéphane Ducasse <[hidden email]> wrote: > >> Hi doru >> >> I have some scripts (available in the wonderful chapter on Gofer) to migrate code from ss and ss3 to SmalltalkHub. >> Now the question is how do we organize moose? >> >> I propose to create a team Moosers? and Moose as a project >> >> Then what do we put underneath? >> >> Fame as subproject? >> Mondrian >> Metanool >> >> >> Stef >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > -- > www.tudorgirba.com > > "Every now and then stop and ask yourself if the war you're fighting is the right one." > > > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
Re: How to organize Moose on SmalltalkHub
|
|
In reply to this post by Tudor Girba-2
I'm ok on the principles but we should pay attention not to make things complex.
Hence I would not have Moose Engines. I was not talking about configuration but location To me a configuration is orthogonal to the structure of the repository. I was talking more about the structure on Roassal Moose Fame MooseCore FAMIX/ MooseAlgos Metanool Arki Merlin EyeSee FAMIX Glamour Magritte ? do we have our own PetitParser? Each of these packages should have a configurationOf that is loading. Stef > Hi, > > I already created a Moose team, and a Moose project inside. > > I see the forces as follows: > - We need to have each engine individually loadable because they are often used in orthogonal projects > - There are projects that are developed separately (like Roassal), and thus these are better living outside of the Moose team repository for the moment. > - Projects that are either no longer maintained, like Fame, or developed only within the context of Moose, like MooseAlgos, should go into the Moose team repository. Nevertheless, I think each of these projects would better have their own repository inside the Moose team repository. > - Moose is the platform, not the actual analyses. Thus, the engines should be grouped in a separate configuration. We can call it MooseEngines. Right now, this would include: Fame, Magritte, Metanool, Glamour, Arki, Merlin, Roassal, (Mondrian for a short time), EyeSee, MooseAlgos, PetitParser, MooseCore, the Core of MooseFinder. > - FAMIX should become a family of meta-models and should receive its own configuration and repository with everything related inside: FAMIX-Core, FAMIX-Extensions, FAMIX-Java, FAMIX-Smalltalk, Hismo, Dynamix. In the same repository, I would add also FAMIXAnalysis to include various extensions such as MondrianPaintings or RoassalPaintings or specific browsers and algorithms. > - At the moment, we still need to clean DSM and SmallDude. Until they are not cleaned, I would not migrate them as they are. > > What do you think? > > Cheers, > Doru > > > > On Jan 29, 2013, at 1:11 AM, Stéphane Ducasse <[hidden email]> wrote: > >> Hi doru >> >> I have some scripts (available in the wonderful chapter on Gofer) to migrate code from ss and ss3 to SmalltalkHub. >> Now the question is how do we organize moose? >> >> I propose to create a team Moosers? and Moose as a project >> >> Then what do we put underneath? >> >> Fame as subproject? >> Mondrian >> Metanool >> >> >> Stef >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > -- > www.tudorgirba.com > > "Every now and then stop and ask yourself if the war you're fighting is the right one." > > > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
Re: How to organize Moose on SmalltalkHub
|
|
Hi Stef,
On Jan 30, 2013, at 1:52 AM, Stéphane Ducasse <[hidden email]> wrote: > I'm ok on the principles but we should pay attention not to make things complex. Actually, I think they are complex at this moment because of unclear structure. And we can see this in the configuration (see the default: and coreDefault: methods). > Hence I would not have Moose Engines. I was not talking about configuration but location > To me a configuration is orthogonal to the structure of the repository. Sure. I just said that once we move, we will have to reword the configurations anyway, so it is a good opportunity to reorganize. > I was talking more about the structure on > > Roassal > Moose > Fame > MooseCore > FAMIX/ > MooseAlgos > Metanool > Arki > Merlin > EyeSee > FAMIX > Glamour > Magritte ? do we have our own > PetitParser? No. For now, Magritte and PetitParser should stay where they are: the repository is stable, and we have enough to do :). > Each of these packages should have a configurationOf that is loading. Exactly. The reason why I would like to have explicit engines loadable easily is that right now it is difficult to communicate Moose, and most people still understand FAMIX and metrics when we say Moose. Now that I think of it, we can say that Moose are the engines, and the rest are in the MooseSuite. Another reason why we have to make the distinction between engines and analyses explicit, is to help future projects split according to these dimensions. For example, because we did not have the distinction, DSM and SmallDude have bundled together both the generic engine and the actual detection and visualization for FAMIX. This made it almost impossible to load within the context of Moose (that is why we now have default: and coreDefault: in the configuration). This distinction is needed. Cheers, Doru > > Stef > >> Hi, >> >> I already created a Moose team, and a Moose project inside. >> >> I see the forces as follows: >> - We need to have each engine individually loadable because they are often used in orthogonal projects >> - There are projects that are developed separately (like Roassal), and thus these are better living outside of the Moose team repository for the moment. >> - Projects that are either no longer maintained, like Fame, or developed only within the context of Moose, like MooseAlgos, should go into the Moose team repository. Nevertheless, I think each of these projects would better have their own repository inside the Moose team repository. >> - Moose is the platform, not the actual analyses. Thus, the engines should be grouped in a separate configuration. We can call it MooseEngines. Right now, this would include: Fame, Magritte, Metanool, Glamour, Arki, Merlin, Roassal, (Mondrian for a short time), EyeSee, MooseAlgos, PetitParser, MooseCore, the Core of MooseFinder. >> - FAMIX should become a family of meta-models and should receive its own configuration and repository with everything related inside: FAMIX-Core, FAMIX-Extensions, FAMIX-Java, FAMIX-Smalltalk, Hismo, Dynamix. In the same repository, I would add also FAMIXAnalysis to include various extensions such as MondrianPaintings or RoassalPaintings or specific browsers and algorithms. >> - At the moment, we still need to clean DSM and SmallDude. Until they are not cleaned, I would not migrate them as they are. >> >> What do you think? >> >> Cheers, >> Doru >> >> >> >> On Jan 29, 2013, at 1:11 AM, Stéphane Ducasse <[hidden email]> wrote: >> >>> Hi doru >>> >>> I have some scripts (available in the wonderful chapter on Gofer) to migrate code from ss and ss3 to SmalltalkHub. >>> Now the question is how do we organize moose? >>> >>> I propose to create a team Moosers? and Moose as a project >>> >>> Then what do we put underneath? >>> >>> Fame as subproject? >>> Mondrian >>> Metanool >>> >>> >>> Stef >>> _______________________________________________ >>> Moose-dev mailing list >>> [hidden email] >>> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >> >> -- >> www.tudorgirba.com >> >> "Every now and then stop and ask yourself if the war you're fighting is the right one." >> >> >> >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev -- www.tudorgirba.com "From an abstract enough point of view, any two things are similar." _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
Re: How to organize Moose on SmalltalkHub
|
|
>
>> I was talking more about the structure on >> >> Roassal >> Moose >> Fame >> MooseCore >> FAMIX/ >> MooseAlgos >> Metanool >> Arki >> Merlin >> EyeSee >> FAMIX >> Glamour >> Magritte ? do we have our own >> PetitParser? > > No. For now, Magritte and PetitParser should stay where they are: the repository is stable, and we have enough to do :). so this is why I asked why you put then in your list :) Now there is an important question is where do we store changes and extension to PetitParser? Because Jan is doing that or guillaume and the message is unclear to them. I want a system that welcome extension. > >> Each of these packages should have a configurationOf that is loading. > > Exactly. The reason why I would like to have explicit engines loadable easily is that right now it is difficult to communicate Moose, and most people still understand FAMIX and metrics when we say Moose. Now that I think of it, we can say that Moose are the engines, and the rest are in the MooseSuite. > > Another reason why we have to make the distinction between engines and analyses explicit, is to help future projects split according to these dimensions. For example, because we did not have the distinction, DSM and SmallDude have bundled together both the generic engine and the actual detection and visualization for FAMIX. This made it almost impossible to load within the context of Moose (that is why we now have default: and coreDefault: in the configuration). > > This distinction is needed. So make a clear list of project so that I see how I can help. > > Cheers, > Doru > >> >> Stef >> >>> Hi, >>> >>> I already created a Moose team, and a Moose project inside. >>> >>> I see the forces as follows: >>> - We need to have each engine individually loadable because they are often used in orthogonal projects >>> - There are projects that are developed separately (like Roassal), and thus these are better living outside of the Moose team repository for the moment. >>> - Projects that are either no longer maintained, like Fame, or developed only within the context of Moose, like MooseAlgos, should go into the Moose team repository. Nevertheless, I think each of these projects would better have their own repository inside the Moose team repository. >>> - Moose is the platform, not the actual analyses. Thus, the engines should be grouped in a separate configuration. We can call it MooseEngines. Right now, this would include: Fame, Magritte, Metanool, Glamour, Arki, Merlin, Roassal, (Mondrian for a short time), EyeSee, MooseAlgos, PetitParser, MooseCore, the Core of MooseFinder. >>> - FAMIX should become a family of meta-models and should receive its own configuration and repository with everything related inside: FAMIX-Core, FAMIX-Extensions, FAMIX-Java, FAMIX-Smalltalk, Hismo, Dynamix. In the same repository, I would add also FAMIXAnalysis to include various extensions such as MondrianPaintings or RoassalPaintings or specific browsers and algorithms. >>> - At the moment, we still need to clean DSM and SmallDude. Until they are not cleaned, I would not migrate them as they are. >>> >>> What do you think? >>> >>> Cheers, >>> Doru >>> >>> >>> >>> On Jan 29, 2013, at 1:11 AM, Stéphane Ducasse <[hidden email]> wrote: >>> >>>> Hi doru >>>> >>>> I have some scripts (available in the wonderful chapter on Gofer) to migrate code from ss and ss3 to SmalltalkHub. >>>> Now the question is how do we organize moose? >>>> >>>> I propose to create a team Moosers? and Moose as a project >>>> >>>> Then what do we put underneath? >>>> >>>> Fame as subproject? >>>> Mondrian >>>> Metanool >>>> >>>> >>>> Stef >>>> _______________________________________________ >>>> Moose-dev mailing list >>>> [hidden email] >>>> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >>> >>> -- >>> www.tudorgirba.com >>> >>> "Every now and then stop and ask yourself if the war you're fighting is the right one." >>> >>> >>> >>> >>> _______________________________________________ >>> Moose-dev mailing list >>> [hidden email] >>> https://www.iam.unibe.ch/mailman/listinfo/moose-dev >> >> >> _______________________________________________ >> Moose-dev mailing list >> [hidden email] >> https://www.iam.unibe.ch/mailman/listinfo/moose-dev > > -- > www.tudorgirba.com > > "From an abstract enough point of view, any two things are similar." > > > > > _______________________________________________ > Moose-dev mailing list > [hidden email] > https://www.iam.unibe.ch/mailman/listinfo/moose-dev _______________________________________________ Moose-dev mailing list [hidden email] https://www.iam.unibe.ch/mailman/listinfo/moose-dev |
«
Return to Moose
|
1 view|%1 views
| Free forum by Nabble | Edit this page |