On Wednesday, 25 May 2016 at 12:19:33 UTC, Steven Schveighoffer wrote:
On 5/25/16 4:22 AM, Atila Neves wrote:
On Tuesday, 24 May 2016 at 01:29:53 UTC, Steven Schveighoffer wrote:
Blog post on making Voldemort types without the disk-space issues:

Should we horcruxify Phobos?

As far as low-hanging fruit, I think it may be worthwhile, since this is really an implementation detail. The fixing of true Voldemort types is likely not happening for a few releases, as the principles don't seem to agree on the best mechanism.


I also agree that fixing phobos this way is a good low hanging fruit. Having symbol name and binary size explosion when using basic phobos functionality could act as deterrent for early adoption of the D language.

Reply via email to