www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - The VIII PPPR (Pending Peeves Progress Review)

reply Stewart Gordon <smjg_1998 yahoo.com> writes:
The time has come once again.

http://www.prowiki.org/wiki4d/wiki.cgi?PendingPeeves


Three months ago, when 0.148 was current, Bugzilla for D was set up.

http://d.puremagic.com/bugzilla/

(Though for some reason, the list of DMD versions there starts at 
0.145.)  Since then, it has amassed just under 200 bug reports.  And it 
seems to have woken up a few old bugs, for example

http://d.puremagic.com/bugzilla/show_bug.cgi?id=65
http://d.puremagic.com/bugzilla/show_bug.cgi?id=66

So two separate issues of covariance not working with interfaces have 
been fixed.  Except that there's a case that still fails, namely when 
there are two return types to override.

http://d.puremagic.com/bugzilla/show_bug.cgi?id=179

Maybe things will be better when even more old bugs are filed in 
Bugzilla.  Who knows?

But the world is still at a loss for why the fix (written nearly a year 
ago now) for this bug still hasn't been committed:

http://d.puremagic.com/bugzilla/show_bug.cgi?id=64


An old peeve is that std.date is lacking functions, or rather 
documentation of them, for constructing and interrogating d_time 
objects.  Now a few of them are documented, but there are still just as 
many that aren't.  And there's been a debate over the quirk whereby some 
parameters are specified as d_time when semantically they're not.  At 
the moment, DDoc seems to be documenting them all as long anyway....


The latest effort at translating the Windows API headers seems to have 
been a success so far.  We've had at least five people working on it - 
anybody else is more than welcome to join in!

http://www.digitalmars.com/drn-bin/wwwnews?digitalmars.D.announce/3194
http://www.prowiki.org/wiki4d/wiki.cgi?WindowsAPI


And an old specification issue is still there, namely the old 
ill-definedness of how opCmp methods must be defined for AAs and sorting 
to work.

http://www.digitalmars.com/drn-bin/wwwnews?digitalmars.D/21572

I was just wondering if it's time we cleared away some of the done stuff 
from the pending peeves page.  I've noticed that somebody's taken the 
liberty to move such stuff to the bottom of each section.  Maybe we 
could delete the stuff that was marked done before somebody decided to 
start putting in at which version number it was done.  What do you 
people think?

Stewart.
Jun 07 2006
parent Don Clugston <dac nospam.com.au> writes:
Stewart Gordon wrote:
 The time has come once again.
 
 http://www.prowiki.org/wiki4d/wiki.cgi?PendingPeeves
 
 
 Three months ago, when 0.148 was current, Bugzilla for D was set up.
 
 http://d.puremagic.com/bugzilla/
 
 (Though for some reason, the list of DMD versions there starts at 
 0.145.)  Since then, it has amassed just under 200 bug reports.  And it 
 seems to have woken up a few old bugs, for example
 
 http://d.puremagic.com/bugzilla/show_bug.cgi?id=65
 http://d.puremagic.com/bugzilla/show_bug.cgi?id=66
 
 So two separate issues of covariance not working with interfaces have 
 been fixed.  Except that there's a case that still fails, namely when 
 there are two return types to override.
 
 http://d.puremagic.com/bugzilla/show_bug.cgi?id=179
 
 Maybe things will be better when even more old bugs are filed in 
 Bugzilla.  Who knows?
 
 But the world is still at a loss for why the fix (written nearly a year 
 ago now) for this bug still hasn't been committed:
 
 http://d.puremagic.com/bugzilla/show_bug.cgi?id=64
 
 
 An old peeve is that std.date is lacking functions, or rather 
 documentation of them, for constructing and interrogating d_time 
 objects.  Now a few of them are documented, but there are still just as 
 many that aren't.  And there's been a debate over the quirk whereby some 
 parameters are specified as d_time when semantically they're not.  At 
 the moment, DDoc seems to be documenting them all as long anyway....
 
 
 The latest effort at translating the Windows API headers seems to have 
 been a success so far.  We've had at least five people working on it - 
 anybody else is more than welcome to join in!
 
 http://www.digitalmars.com/drn-bin/wwwnews?digitalmars.D.announce/3194
 http://www.prowiki.org/wiki4d/wiki.cgi?WindowsAPI
 
 
 And an old specification issue is still there, namely the old 
 ill-definedness of how opCmp methods must be defined for AAs and sorting 
 to work.
 
 http://www.digitalmars.com/drn-bin/wwwnews?digitalmars.D/21572
 
 I was just wondering if it's time we cleared away some of the done stuff 
 from the pending peeves page.  I've noticed that somebody's taken the 
 liberty to move such stuff to the bottom of each section.  Maybe we 
 could delete the stuff that was marked done before somebody decided to 
 start putting in at which version number it was done.  What do you 
 people think?
I think that bugzilla does an adequate job of recording the done stuff, now, so you can safely clear it out. The pending peeves are most useful for the things that aren't exactly bugs, such as the ones you've listed above.
Jun 07 2006