dmd platform support - poll

[Nick Sabalausky]

"John Reimer" <terminal.node at gmail.com> wrote in message 
news:28b70f8c104808cb3654b402e6d0 at news.digitalmars.com...
> Hello Yigal,
>
>> John Reimer wrote:
>>
>>> Hello Yigal,
>>>
>>>> John Reimer Wrote:
>>>>
>>>>> Hello Tim,
>>>>>
>>>>>> Yigal Chripun Wrote:
>>>>>>
>>>>>>> personally I don't see a point in JVM/.NET - One of the best
>>>>>>> things about D is that you get the ease of use of Ruby/python/etc
>>>>>>> with the benefits of native compiling like in c/c++. Why throw
>>>>>>> that away and make yet another version of Java/C# ?
>>>>>>>
>>>>>> Supporting .net would give you access to the most modern and
>>>>>> probably best-currently-supported Windows API. It would, if you
>>>>>> counted Mono, add a very nice cross-platform UI framework.
>>>>>> Finally, depending on what version was supported, it might enable
>>>>>> you to write Silverlight apps in D, permitting flash-like apps
>>>>>> that run cross-functionally in a web browser.
>>>>>>
>>>>>> TK
>>>>>>
>>>>> Agreed.
>>>>>
>>>>> Concerning .NET and D technology, I say go for it... especially if
>>>>> someone has the initiative to keep such a port going (afterall,
>>>>> such initiative is really the most important virtue for any hope of
>>>>> success). For myself, I'm kind of learning not to "restrain" D with
>>>>> my personal biases. Sometimes we just can't predict what kind of
>>>>> benefits might be in store for the language, the platform, or other
>>>>> people; such expiditionary moves might not be successful in
>>>>> themselves, but they could be the critical factor that brings D to
>>>>> the limelight in some future endeavor.
>>>>>
>>>>> D may be successful in areas we don't necessarily predict or
>>>>> prefer, and .NET is just one of several interesting possibilities
>>>>> to explore. Therefore, I don't think we should get too
>>>>> tunnel-visioned about "D is better because it's a compiled
>>>>> language". It may be important to keep the vision a little more
>>>>> open to other technologies (like VM's and such) especially as
>>>>> optimizations improve in these areas. Otherwise, D will be at risk
>>>>> of loosing it's general purpose nature... and being permanently
>>>>> fixated as a niche language. Porting to .NET, therefore, becomes a
>>>>> clever way of "proving" D's viability on other technology
>>>>> platfroms.
>>>>>
>>>>> I haven't used C#, but I can bet that D could offer a very
>>>>> competitive and comfortable programming environment such that it
>>>>> would be a welcome alternative even in the .NET world. Microsoft
>>>>> may even come to see the benefits, since D might attract an even
>>>>> more diverse audience to the platform, people who would have
>>>>> otherwise avoided it. You never know. ;)
>>>>>
>>>>> That'd probably be all it would take for me to start experimenting
>>>>> with .NET and Mono.
>>>>>
>>>>> -JJR
>>>>>
>>>> You make a valid point. Attracting new developers to D by supporting
>>>> more platforms is a worthy long-term goal for the D language.
>>>>
>>>> however, I have a 64 bit PC and since Walter is only one person with
>>>> limited time, I'd personally want that Walter spend his time in the
>>>> short term on adding support for 64 bit, not working on a .net port.
>>>> I can live with C# on .net for now and as I said before it's close
>>>> enough to a D.net.
>>>>
>>>> On the native compiled front I really wouldn't want to go back to
>>>> C++ after using D.
>>>>
>>>> so answering Walter's original question:
>>>>
>>>> for me .net port is VERY low priority compared to 64 bit support
>>>> which is a HIGH priority.
>>>>
>>> Yep, I understand your point as valid also when it comes to
>>> prioritizing which platform most needs to be worked on. That's why
>>> the completion of .NET support would probably have to be done by
>>> another... which may be the case already.
>>>
>>> -JJR
>>>
>> also, while I agree with having a D.net to get more exposure for D and
>> attracting new developers, I personally wouldn't use it and I doubt
>> any other existing D developers would use it. There's also another
>> concern about such a port - the libraries. The phobos/Tango split hurt
>> D enough and adding the .net libs (or the JVM ones) to the mix will
>> make this issue even worse.
>>
>
>
> Two things I question in your post:
>
> (1) The assumption that existing D developers won't use it.
>

I'm going to lump .NET and JVM (and the Flash VM) together here and say that 
there are some things (particularly on the internet) that unfortunately, due 
to existing software infrastructure (ex, browser), can't realistically be 
done with natively-compiled code. Whenever I'm unlucky enough to find myself 
in one of these situations, I'd much rather use D than any of the languages 
usually associated with such platforms. And even with .NET in particular, 
while C# does have a lot of good things going for it (including a number of 
things I would like to see added to D), the templates and metaprogramming 
are just abysmal and I personally find those more important than the C# 
niceities that D lacks.

> (2) The assumption that the D port will only use .net libs.
>

I'd view this as more of a concern that the D port would use the .NET libs, 
rather than an assumption. If a D.NET ended up needing to use the .NET libs 
in place of phobos and tango, or if the .NET platform necessitated certain 
changes in the phobos/tango APIs, then that would indeed be a problem that 
would take away much (though not all) of the usefulness of a D.NET.

> I don't think we have enough information yet on either point.  Perhaps you 
> know more than I do. :)
>
> -JJR
>
>