Label: ♦english

Thanks for asking. I Actually have been thinking about a new release, after having added several very interesting new features and done many improvements, I am eager to release it as soon as possible. But given what remains to be done before the new release, the soonest possible time would be January next year. I will create a post to summarize the recent development and show what can be expected in the next release.

I am afraid the current daostudio is not compatible with the current dao, but this incompatibility can be fixed without too much efforts, and will be fixed soon. For the modules, I think DaoOpenGL should still work if the binding codes are regenerated by the current autobind.dao. DaoQT and DaoFLTK should work too, but I haven't tested for some time, I will have a check. Other modules such DaoSDL, DaoGSL and DaoMathGL are likely to work by updating the binding codes with the current autobind.dao. The rest modules may require some fixings.

Limin, please make a "feature freeze" like 1 or 2 weeks before the release, and then ask us for help writing a lot of scripts thoroughly testing (and showcasing) all the new and old features. For example, I still have a script testing a lot of inheritance stuff on my pc.

I too develop OpenGL - but only in C++ for now. It would be interesting to add a new "OpenGL3" wrapper, which could use the "gl3.h" header you can find in the opengl extension registry website. OpenGL 3 was kind of a "cleanup" of opengl, and thus is now much smaller! Anyway, that woud take some time (which I currently lack) but I just wanted to raise the subject.

Don't forget about the Wikipedia article, Pompei2! I guess it's what Dao strongly needs now.

Yes, I will freeze new feature development for some time before the next new release, I want it to be well tested. But there is still one more feature I would like to add before the next release. It will be a feature to bring back Asynchronous Function Call and Message Passing Interface, but in a much simplified form. I will open a new topic to discuss about it, to see if it worths or doable to added it now. For OpenGL3, I will have a look at it.

Good call for the Wikipedia article, I nearly forgot it :) Will do it at a time close to the release.

We are at the end on janurary. Any news about the packaging of the next version of Dao ? Can we expect a release soon ? (just curious, don't want to bother you)

Maybe a roadline would be useful, where Limin states what features he wants to be done before the next release, then we have at least a little clue :-)
Also, I think that if we want the release to be good, we will need at least a month of time for testing, bugfixing and (very important) documenting all the new features.

Agree, we need thorough bug-fixing and major documentation update. Besides, some of unfinished features (refer to the forum) could be finally dealt with.

I am afraid the next release will not be very soon, I can not give a precise timeline for this, but I am sure it will be in the first quarter this year :)

Roughly speaking, the following thing should be done before the next release:

1. Some minor missing features in the template class support;
2. Some missing things in the synchronous class feature;
3. Reorganization of some built-in libraries such as sys , std and reflect ;
4. Completed improvements to error message reporting;
5. Fix most (if not all) the issues in the issue list in the google code project page;
6. Implement/Fix what are suggested in the discussions on this forum;
7. More complete testing;
8. Updated documentations;

There are also some work to be done for the extending modules, mainly some fixing and update in the binding codes. In the past a few weeks, some modules (DaoCGI, DaoCLoader, DaoGSL, DaoOpenGL, DaoQT, DaoVTK, DaoFLTK, DaoSDL and DaoSQL with MySQL and SQLite backends) have been fixed or updated, what is left is a PostgreSQL backend for DaoSQL, a updated DaoLapack and a few minor modules. Modules for OpenGL3 may also be created for this release.

As it can be seen, there are still many things to be done, but probably except the synchronous class, none of them is going to be tough:)

So ... a few month later, and many commits in GIT (I'm watching it weekly), is there a release soon in preparation ?
regards, Yoann

As I see things, I would guess "not in the next months", because there are a few areas still "under construction".
I would recommend Limin to create a list of features or tasks needed for the next release 1.2 and then stick to that, new features could be added in a 1.3 branch, but not added to 1.2 as it seems 1.2 will never get released if always new ideas come up :p (Same happens to me in my project btw ;))

Isn't Dao already in version 1.2 for nearly 9 months? :) At least DaoVM's "welcome screen" claims that. Anyway, lots of things still need to be completed/updated/fixed (I suppose the TO-DO list hasn't changed much since Limin's last post), thus we indeed shouldn't expect the release soon.

For me, the following things seem to need to be done above all (in the respective order):

• completing planned features (develop our faith!);
• comprehensive testing and rigorous bug-fixing (strengthen our faith, purge ourselves of doubts!);
• full documentation update including description of standard modules (state our faith in writing!);
• new Wikipedia article (propagate our faith, enlist new followers!).

As you can see, it's a time-proven "recipe" :)

Hi, Yoann, thank you for your patience in waiting for the next release. You can follow the development on dao.googlecode.com . The latest codes are committed into the HG repository in this googlecode project. The GIT repository on sourceforge is no longer updated. Googlecode is our preferred hosting platform now.

Regarding the timeline of the next release, please have a look at: http://daovm.net/space/dao/thread/639 .

Though this release has been delayed for quite some time, but not much longer. With the completion of the new parser (I was really uncomfortable with the previous one), near-completion of LLVM-base JIT compiler (DaoJIT) and Clang-based automatic binding tool (ClangDao), I believe it is time to seriously work toward a new release:) I have posted something here: The next release is going to be a big one !

From now on, I will mostly concentrate on finishing those unfinished things and updating documentations. Synchronous class and curry will probably stay as they are now. If support for C++ template and Objective-C in ClangDao turns out to be more complicated than I thought, they may also be left for the future development. Another thing I am not so sure is the Dao template class that is supported not long time ago, this implementation may not handle some situations properly, if it turns out there is a fundamental flaw in the implementation which requires a significant amount of codes to fix it, this feature may just be taken out for simplicity.

I am wondering what version number to use for the next release: 1.2? I am not sure, this does not reflect the amount of new features and improvements that will be included in the next release. Maybe the version number need to jump a little bit:)

It all depends where you come from :)

It seems to me that the open source language community uses 1.1 -> 1.2 to mean some big changes (Python, Ruby) while the closed ones use 1.1 -> 1.2 for small fixes (C#, Java). (Or is that linux vs windows?) So I personally would tend to name it 1.2

I think in the next release the total amount of changes accumulated since Dao 1.1 will be huge (much greater than in the previous release). I doubt languages like Python or Ruby have ever undergone comparable changes during "n.m -> n.(m+1)" development. So I believe if we accomplish all currently required work without leaving anything behind, Dao could earn somewhat higher "promotion" :)

Probably there is no other language has undergone such amount of changes in a single release interval (since version 1.0 of course). I searched through the release histories of some languages, from what I can see, there was nothing comparable. Interestingly, I saw a jump from version 1.1 to version 2.1 in Lua (Lua version history ), I don't know if there were some minor releases in between.

Initially I was thinking to choose the next version number among 3 options: 1.2 , 1.5 or 2.0 . Now 2.0 seems more preferable. Because it will truly reflect the amount of changes that can be expected, and it will have better promotional effects. However, 2.0 will imply a higher standard for feature supports, testing and documentations. If the next release is intended to have a very good impact, this has to be THE WAY!!!

PS, Dao also means THE WAY :)

So now, are we close to the next release ? You said in the other thread, hopefully in august ;-)
I'm really curious to see all the enhancements Dao have gone along the way. Seem's promissing. Without any release, I hope you did'nt loose to much folowers.

Well, the next release doesn't really seem close. Having a look at the project state at Google Code, you can see how much needs to be resolved. New GC, module compatibility issue (BTW, how about the new API functions I proposed?), broken template and synchronous classes, a lot of "hanging" proposals and unfinished functionalities, a swarm of known bugs (no doubts there are many more yet to discover)... Have I forgotten anything? :) Anyway, there is a lot of pending work and no guaranty that no issues will be added in the nearest future.

I also thirst for the next release, but too much haste won't do any good now.

Not quite yet, as Nightwalker has explained why. I do hope I can release it sooner, but without working on it on full time, it is probably not going to happen in another month. I really appreciate very much your patience.

Thank you for reminding that issue, I just replied to it. The new GC is going to be dropped off, as it didn't turn out as I expected, I will stick with the current GC and try to improve it instead. Two other things worth mentioning are DaoJIT and ClangDao (mainly the support for Qt is missing).