Last update October 5, 2009

Pending Peeves

These are issues that people feel need to address, hopefully in time for D 1.0. Be realistic, do not make this a wish/feature request list! (see Feature Request List to suggest a new feature)

Maybe this could form the basis for an official 'to do for 1.0' list.

It is important to read all the posts linked to in order to fully understand the issues and how they should be resolved, and not cause duplications.

Bugs in DMD:

[partly done] Deprecation problems ( D:25715 et seq., d.D.bugs:1202, Bugzilla:551 (fixed) et seq.)
Debugger output ( D:28413)
[partly done] Abstract attribute problems ( d.D.bugs:1939, d.D.bugs:1940, Bugzilla:204 (fixed))
Const struct member is a non-constant expression ( d.D.bugs:3290, Bugzilla:609)
[spec changed, DMD 0.162, but still a mess] Compiler accepts ';' as statement ( d.D.bugs:4828, Bugzilla:327, Bugzilla:576 (fixed))
Various forward reference problems ( Bugzilla:340)

Things that it isn't clear if are bugs (either the compiler or the spec needs fixing):

[partly done, DMD 0.168] Initialising non-static and associative arrays ( D:26695)
Current D grammar not context free? ( D:26476, d.D.bugs:1955 et seq.)
in/inout/out for arrays needs clear defining ( d.D:5468)
Abstract method cannot have contracts ( d.D.bugs:1939)
Semantics of \u ( d.D:11391 et seq.)
[done?, DMD 0.167] Typedef arithmetic ( d.D.bugs:3843)
False matches in associative arrays if opCmp is inconsistent with opEquals ( d.D.bugs:2643, d.D.bugs:4556 et seq., d.D:27376, Bugzilla:612 (fixed))
[regression, fixed 0.176] Compiler accepts incomplete initialisation of an array ( d.D.bugs:4300, DocComments/Arrays#Contradiction, Bugzilla:508)
[done?, DMD 0.162] Covariance doesn't work with interfaces ( d.D.bugs:1726, d.D.bugs:2056 et seq., d.D.bugs:3287, Bugzilla:179)

Holes in the spec:

Issues that need addressing if copying GC is to be allowed (probably is, considering GC-related restrictions on pointer operations) ( D:26273)
Object.opCmp predefined wants at least a rationale, if not a change ( D:26144, d.D:10558, d.D:21572)
When are function calls without trailing parentheses allowed? ( d.D:35742)

Things missing from Phobos:

[done?] std.file - copy function ( D:27490, d.D.bugs:3040)
[partly done] std.date - functions to get/set/construct by components ( D:27490, but see also d.D:5405)
[partly done] a decent I/O system ( D:25096)
Complete translation of the WindowsAPI headers ( d.D:15248 et seq., d.D.announce 3194 et seq., Bugzilla:317)
[partly done] Some essential conversions between character encodings ( d.D:28264 et seq.)

Language features waiting to be implemented or at least removed from the spec:

[done, DMD1.048 and DMD2.033] Inheritance of in/out contracts ( d.D:7333, Bugzilla:302)
[done] Other features related to in/out contracts ( d.D:7199, d.D:7227)

Miscellaneous showstoppers (things that will either break a lot of code, or make the spec look dreadful):

[done, D2 only] Remove printf() from object.d, so it's no longer implicitly defined ( d.D.bugs:5838).
[partly done, DMD 0.149] Provide access to the IEEE rounding modes, precision and exception status ("blah blah blah") in float.html.
[done?, DMD 0.163] Object.opCmp - about time we came to a decision ( d.D:21572)

Uncertainty about what will be in D 1.0:

Improvements to IFTI support ( d.D:38780)
When is it time for a 1.0 feature freeze? ( d.D:41553)

FolderDiscussions