Tango D2 Development

Posted: 12/18/11 17:26:31

I did not understand. Is it's unofficial attempted to port Tango to D2? Or it's Tango's community project?

Posted: 02/20/12 03:35:15

Just an update on the above effort... in principle all modules that are reasonable to be ported (e.g. not runtime bits) have been ported, and everything seems to work fine (ported a few D1 programs and they seem to work without a hitch). Effort was made to keep the public interface the same, although some changes had to be made in order to avoid excessive GC allocations and unsafety.

My question is... how interested are the Tango developers in taking steps for this to be something more official than it already is? Tango development seems to have been very slow the past few years, and it certainly doesn't seem because Tango is perfect the way it is... additionally, it doesn't seem like D1 development will be tenable for new D users or remain tenable for current D1 users. From what I understand dmd will drop support for D1 by the end of this year, and Tango doesn't build with the most recent LDC either (you have to drop down quite a few revisions).

Now, it is certainly true that there is, in principle, less of a "need" for Tango in D2 since Phobos2 is a great improvement over Phobos1. Still, I think that Tango philosophy of minimal resource use and modularity is not exemplified in the least by either Phobos, and thus there is still a niche for Tango users old and new in D2. Despite what Phobos coding style often implies, it is still possible to code without template bloat or GC overuse. While I had to add quite a few .dups while porting Tango in many cases it was due to the desire to keep the interface the same... with minor changes in the interface it is likely that many of those .dups can be made to disappear.

So... I think that TangoD2 is a Good Thing (TM). I would love to hear the input of the current D1 users Tango and its developers about how to make the porting effort most useful to everyone.