Dev website migration - Was: Migration to BitBucket

[Klaim - Joël Lamotte]

On Tue, Oct 2, 2012 at 10:07 AM, Klaim - Joël Lamotte <mjk...@gmail.com> wrote:

I started the migration today.

The repository is already available there: https://bitbucket.org/artofsequence/aos

There is still some work to do:

 - import the issues tickets

 - import the wiki content (though maybe I'll use something else than bitbucket for the wiki, not sure yet)

 - write/complete the readme

Joel Lamotte

So, after having played a bit with bitbucket, I had doubt it would be flexible enough to organize things how I want, in particular on the organization side.

I asked Atlassian for an open-source license of Jira and Confluence that you can currently see at these adresses:

 - http://tracker.artofsequence.org

 - http://docs.artofsequence.org

I didn't finished to update it yet but I think I will not.

My experience so far with these tools is that it is a bit overkill, in particular Confluence.
I don't really like the way they organized all the options to manage Jira and Confluence, which sometime make me loose several minutes just looking for an option that appear when you get on a page this way but not the normal way.

Yesterday, as said in another discussion, I started considering GitHub to host the dev website, mainly because it's both as simple as bitbucket.com and more flexible to manage documentation, multiple projects and global organization.

After some tests ( http://artofsequence.github.com/ ) it appear that I have there all the features I wanted. 

The only thing is that it forces the project to switch to git. It's not a big problem, and it could make the project more visible. I just have to convert current history and also learn more about git as I have only basic knowledge about it (I did several pull requests on github before).

So, to sum up, I'm strongly thinking about moving the project to Git and GitHub.com.

I will be experimenting with this today, converting the project to git/github, see how it goes.
Hopefully it will be faster to move documentation from google code to github than to Confluence.

Joel Lamotte

[Joonhwan Lee]

FYI,  my github id is "joonhwan" https://github.com/joonhwan

2013/2/15 Klaim - Joël Lamotte <mjk...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "Art Of Sequence Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aos-dev+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
이준환
www.joonhwan.org

[Klaim - Joël Lamotte]

On Fri, Feb 15, 2013 at 11:00 PM, Joonhwan Lee <joonhw...@gmail.com> wrote:

FYI,  my github id is "joonhwan" https://github.com/joonhwan

Very cool! :D
Happy to see a dev is interested in helping (as soon as I get things back in stable state)

Joel Lamotte 

[Joonhwan Lee]

ueah i will fork that repo as soon as you start it a kinda officially!!!! :-)

2013년 2월 16일 토요일에 Klaim - Joël Lamotte님이 작성:

--

You received this message because you are subscribed to the Google Groups "Art Of Sequence Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aos-dev+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
이준환
www.joonhwan.org

[Klaim - Joël Lamotte]

I finished yesterday to migrate AND split the current code into several repos that are on github: https://github.com/artofsequence

I decided to not report the history from the hg repository because there is no point right now to do this.

For now I splitted the repos so if you want to work on or use only one project you can. However this makes a bit more
complex to dev if you want to work on all projects. I don't know if github allow to clone a super-repository (like aos-all repository)
so I need to do some experimentations. So it is not yet sure I will keep this organization, beware.

At the moment, if you clone (recursively) aos-all, and try to use CMake to generate a project file, it should work exactly

as if you did it with the original mercurial repository. I didn't change the code yet, I'm still working on making things smooth

and figuring out how to setup the development documentation that is not specific to one project.

The goal would be for a developer to only get the projects he needs to work on or use, not all projects at once.

I'll keep you up to date on this.

Joel

[Klaim - Joël Lamotte]

I finished the migration of issues and documentation (which is still imcomplete, but one thing at a time...).

I posted a news with some details there: http://artofsequence.org/?p=405

[Joonhwan Lee]

Good job. fetching is coming right up.

2013년 3월 1일 금요일 오후 11시 50분 11초 UTC+9, Klaim - Joël Lamotte 님의 말:

[Klaim - Joël Lamotte]

Trying to work with git submodules in the aos-all super repository makes me cry a bit. ;__;
Maybe I'm not used to git specificities yet but I'm often getting very weird stuffs.

I think making sure that AOS Designer repository works well alone would be better.

It depends on UtilCPP (which I will remove soon) and AOSLCPP only. 

It might be simpler to just get both AOSD and AOSLCPP sources and start working on them
instead of using the super repository. 

AOSL repo can be worked alone, once a version is finished it's easy to just spread the new version to each

other projects repositories. 

I might also merge the online demo and AOS Web Player repositories to simplify working with both.

I'll experiment with these ideas and make changes accordingly.

Also, I plan to move the wiki documentations into the repositories. 

GitHub can interpret text files (markdown or whatever) which are inside the repos. It will help keeping

the docs in sync with the repository state.

So basically, still much things to figure to simplify dev on these projects.

Joel Lamotte

[Joonhwan Lee]

Just to simply note.

I tried to build AOS designer in my old mbp, and figured that I needed other projects. so heres what I've done last weekend

-  I cloned root repository

- git submodule init

- git submodule update

-  install xcerce

-  install xsd

-  ran cmake
-  and finally ran make.

The build process was failed due to xsd's icompatibility to llvm clang(it was unable to compile xsd generated code well)

I built Qt and whole other 3rdparty /usr/local/lib s with clang not gcc. and by sweet heart second daughter wanted me to play with her and was unable to go on.

What can I do for handling this build issue?

Joon.

PS. 

I am not familiar with xcerce and xsd which seems to me that it makes xml handling more c++ friendly. 
I just thought if we could use json for saving format and Qt, we could access xml's tree data with just string representation of its node like "aos.scene.1.object.size = 10" using QtScript subsytem and no dependency on tools like xsd.

I was feeling like to implement it myself just to study the QtScript subsystem.

2013/3/12 Klaim - Joël Lamotte <mjk...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "Art Of Sequence Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aos-dev+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
이준환
www.joonhwan.org

[Klaim - Joël Lamotte]

On Wed, Mar 13, 2013 at 2:26 AM, Joonhwan Lee <joonhw...@gmail.com> wrote:

I tried to build AOS designer in my old mbp, and figured that I needed other projects. so heres what I've done last weekend

-  I cloned root repository

- git submodule init

- git submodule update

-  install xcerce

-  install xsd

-  ran cmake
-  and finally ran make.

I'm surprised you had to install xerces manually, it should be a xsd dependency.

 

The build process was failed due to xsd's icompatibility to llvm clang(it was unable to compile xsd generated code well)

I built Qt and whole other 3rdparty /usr/local/lib s with clang not gcc. and by sweet heart second daughter wanted me to play with her and was unable to go on.

What can I do for handling this build issue?

I didn't know (never tried) clang could not compile xsd. I'll see with xsd devs, thanks for the info.

 

PS. 

I am not familiar with xcerce and xsd which seems to me that it makes xml handling more c++ friendly. 

To be clear: xsd generate C++ code from xsd (which then is dependent on the xerces XML library compiled for static linking).

Which mean a big part of AOSLCPP C++ code is generated using xsd/tree (the generator application).

 

I just thought if we could use json for saving format and Qt, we could access xml's tree data with just string representation of its node like "aos.scene.1.object.size = 10" using QtScript subsytem and no dependency on tools like xsd.

I was feeling like to implement it myself just to study the QtScript subsystem.

Qt is a dependency that should be necessary only in the "view" part of AOS Designer. I can't impose it to the other devs for other tools depending on AOSLCPP.

So the "core" part of AOS Designer have to be independent from Qt (it is not totally right now but will be before the end of this month).

It's ok to use QtScript or whatever Qt in the "view" part but not in anything else. However it's not clear to me how it would be useful to use QtScript for this? 

To quickly program the interface?

I'm using xsd mainly because it's currently the simplest way to check that the C++ code is still compatible or not with the changes done on the xsd file of AOSL, 

at compile time, not at runtime. In my experience it have been of great help to drive changes in tools when the format is changed.

For now I'll see with XSD guys what's the problem with clang and act accordingly. I don't think that's normal at all.

In the mean time, is it possible for you to check with gcc too? It's not urgent though.

Joel Lamotte

[Klaim - Joël Lamotte]

Is this the issue you are seeing compiling with clang? http://codesynthesis.com/pipermail/xsd-users/2013-February/003838.html

If it is, they say in the next version it's fixed and they provide a patch: http://codesynthesis.com/pipermail/xsd-users/2013-February/003839.html

Does it help?

Joel Lamotte

[Klaim - Joël Lamotte]

If it's not the problem, please post a full error log in another dedicated discussion thread, 

so that we can try to fix it with xsd people.

Joel Lamotte

[Joonhwan Lee]

Exact that problem. Though the same here's my build log follows. For the time being, I should work on aos only under windows.

In file included from /Users/jhlee/prj/oss/mine/aos-all/library/aoslcpp/source/aosl/activation.cpp:23:

In file included from /Users/jhlee/prj/oss/mine/aos-all/library/aoslcpp/include/aosl/activation.hpp:37:

In file included from /Users/jhlee/prj/oss/mine/aos-all/library/aoslcpp/include/aosl/activation_forward.hpp:44:

In file included from /Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/parsing.hxx:10:

In file included from /Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/parsing/date-time.txx:7:

In file included from /Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/zc-istream.hxx:215:

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/zc-istream.txx:35:7: error: use of undeclared identifier

      'setg'

      setg (b, b, e);

      ^

      this->

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/zc-istream.txx:17:7: note: in instantiation of member function

      'xsd::cxx::zc_streambuf<char>::init' requested here

      init ();

      ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/zc-istream.txx:64:11: note: in instantiation of member

      function 'xsd::cxx::zc_streambuf<char>::zc_streambuf' requested here

        : buf_ (str)

          ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/zc-istream.txx:81:11: note: in instantiation of member

      function 'xsd::cxx::zc_istream_base<char>::zc_istream_base' requested here

        : zc_istream_base<C> (str),

          ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/parsing/date-time.txx:211:25: note: in instantiation of

      member function 'xsd::cxx::zc_istream<char>::zc_istream' requested here

          zc_istream<C> is (year_fragment);

                        ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/parsing/date-time.txx:163:9: note: in instantiation of

      member function 'xsd::cxx::tree::gyear<char, xsd::cxx::tree::simple_type<xsd::cxx::tree::_type> >::parse'

      requested here

        parse (text_content<C> (e));

        ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/type-factory-map.txx:434:41: note: in instantiation of

      member function 'xsd::cxx::tree::gyear<char, xsd::cxx::tree::simple_type<xsd::cxx::tree::_type> >::gyear'

      requested here

        return std::auto_ptr<type> (new T (e, f, c));

                                        ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/type-factory-map.txx:170:11: note: in instantiation of

      function template specialization 'xsd::cxx::tree::factory_impl<xsd::cxx::tree::gyear<char,

      xsd::cxx::tree::simple_type<xsd::cxx::tree::_type> > >' requested here

          &factory_impl<gyear>,

          ^

/Users/jhlee/sdk/xsd-3.3.0-i686-macosx/libxsd/xsd/cxx/tree/type-factory-map.txx:414:21: note: in instantiation of

      member function 'xsd::cxx::tree::type_factory_map<char>::type_factory_map' requested here

          map = new type_factory_map<C>;

                    ^

/Users/jhlee/prj/oss/mine/aos-all/library/aoslcpp/source/aosl/activation.cpp:37:3: note: in instantiation of member

      function 'xsd::cxx::tree::type_factory_plate<0, char>::type_factory_plate' requested here

  type_factory_plate_init;

  ^

/usr/bin/../lib/c++/v1/streambuf:172:10: note: must qualify identifier to find this declaration in dependent base

      class

    void setg(char_type* __gbeg, char_type* __gnext, char_type* __gend);

         

....

2013/3/13 Klaim - Joël Lamotte <mjk...@gmail.com>

If it's not the problem, please post a full error log in another dedicated discussion thread, 

so that we can try to fix it with xsd people.

Joel Lamotte

--

You received this message because you are subscribed to the Google Groups "Art Of Sequence Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aos-dev+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
이준환
www.joonhwan.org