On Saturday, 6 June 2015 at 03:41:13 UTC, Joakim wrote:
On Friday, 5 June 2015 at 20:23:17 UTC, David Nadlinger wrote:
On Monday, 1 June 2015 at 18:14:40 UTC, Andrei Alexandrescu
wrote:
As much as I want it, ddmd seems to not be happening for
2.068 because of, simply put, insufficient resources.
Why is that?
A first test release does not seem to be further away than,
say, full rangeification of Phobos.
Answered earlier in the thread:
http://forum.dlang.org/post/[email protected]
That's not a full answer. I worked with Daniel to get LDC to
successfully compile DDMD the Saturday after DConf, which is part
of the reason why we can confidently make the 20% claim in the
first place (i.e., be sure that is not a C++ vs D issue).
Still, I'm confident that getting a LDC release ready would be
less work than, say, properly refactoring all of Phobos to avoid
allocations by using ranges.
Sorry if I appear a bit grumpy, but even though recently a number
of people have been clamoring for more focus on high-impact,
strategically important work, not a single one of them has showed
up at the doorsteps of GDC/LDC with any patches so far. This
strikes me as rather schizophrenic and dishonest, especially
given that the same people are quick to mention the importance of
those compilers in other contexts. Either that, or they seem to
maintain the conception that DMD is somehow a viable option for
performance-critical code. In the latter case, I don't have much
hope for D in the long term, given that this would imply that
decisions are made involving an alarming level of delusional
double-think.
- David
