Rename std.string.toStringz?

[Daniel Gibson]

Am 26.06.2011 04:21, schrieb Walter Bright:
> On 6/25/2011 7:02 PM, Jonathan M Davis wrote:
>> The current plan at the moment when it comes to deprecation is that
>> things
>> which are deprecated will generally be labeled as scheduled for
>> deprecation
>> for 6 months, deprecated for 6 months, and then fully removed - though
>> the
>> exact length may vary depending on release dates and what exactly is
>> being
>> deprecated (e.g. stuff that's been around longer is more likely to
>> take longer
>> to remove than stuff that's changed after a single release - which
>> hopefully
>> happens very rarely). So, code shouldn't be breaking out from under
>> people,
>> and they'll have some time to make the necessary changes as best fits
>> what
>> they're doing.
> 
> 
> Using the deprecation mechanism to give people warning of breaking
> changes is good. But *still* we need to restrict breaking things only if
> we've got compelling, and I mean compelling, reasons to do so.
> 
> We simply cannot build a D community on a mass of code that won't
> compile. Dsource is full of that.

Dsource is full of projects that
1. are for D1 (or very early versions of D2), so, as you mentioned, they
won't compile anymore (changing names won't make much difference for them).
2. haven't been maintained for ages.

But what about existing D2 code that does still work (maybe with trivial
fixes)? Is it unmaintained as well or could we assume that the
maintainer would fix any breakage caused by renaming some stuff in Phobos?

So: Is there much existing, *working*, D2 code that isn't maintained
anymore? And could we expect this code to continue working for the next
years without being maintained if we do not change some names in Phobos
(but only do any other needed changes to the language and to Phobos)?

Cheers,
- Daniel