On Wed, 20 Feb 2013 21:48:25 +0400
Dmitry Olshansky <dmitry.olsh@gmail.com> wrote:

> [...]
> 

Cool! I'm almost expecting to see someone reply here with "What a coincidence! I just had the opposite thing happen to me: I was riding a train in Russia, booted my laptop to Arch, started some D coding on a Git project all while some other D user kept watching my screen!" :)

On Wednesday, 20 February 2013 at 21:26:58 UTC, 1100110 wrote:
> On 02/20/2013 12:42 PM, bearophile wrote:
>> Rob T:
>>
>>> and do a ton of boring stuff like stabilize D2/Phobos before moving on
>>> to D3, install better processes for documentation and developing the
>>> language specifications, and of course continue to improve the release
>>> process (it still needs a real beta and stable release), etc.
>>
>> "Stabilize" is the wrong word to use. Implementing the 64 compiler is
>> good, implementing shared libraries is good, porting D runtime/Phobos to
>> RISC CPUs is good, replacing the GC is good, improving the floating
>> point management by DMD is good, and so on and on.
>>
>> But in my opinion what's more needed now is instead to try to complete
>> as much as possible the design and implementation of the
>> missing/broken/incomplete parts of the core language (like finishing
>> const/immutable design, finishing the implementation of pure,
>> redesigning properties, fixing @trusted, doing what's possible with
>> shared, doing what's possible to finish the inference of tags like pure
>> in templated functions, finishing the design of packages, finishing the
>> implementation of the module system, finishing the design of operator
>> overloading, and so on. The complete list of broken/unfinished parts
>> scares me).
>>
>> Bye,
>> bearophile
>
> Yes. Shared Libraries Please.
>
> Thank you.

+1

It's the most important issue for me. Particular feature bugs can be worked around, lack of shared library support is insurmountable for many applications.

On Thursday, 21 February 2013 at 00:48:50 UTC, John Colvin wrote:
[...]
>>
>> Yes. Shared Libraries Please.
>>
>> Thank you.
>
> +1
>
> It's the most important issue for me. Particular feature bugs can be worked around, lack of shared library support is insurmountable for many applications.

Yes, it's a major barrier towards adoption in a production environment. Walter said it's now a priority, so I cross the fingers and hope.

--rt

On Thu, Feb 21, 2013 at 01:55:00AM +0100, Rob T wrote:
> On Thursday, 21 February 2013 at 00:48:50 UTC, John Colvin wrote: [...]
> >>
> >>Yes. Shared Libraries Please.
> >>
> >>Thank you.
> >
> >+1
> >
> >It's the most important issue for me. Particular feature bugs can be worked around, lack of shared library support is insurmountable for many applications.
> 
> Yes, it's a major barrier towards adoption in a production environment. Walter said it's now a priority, so I cross the fingers and hope.
[...]

Better yet, maybe more of us should start delving into DMD's code... Currently only a small number of people really understand the compiler, which has a rather low bus factor. We need to increase that (not only to improve the bus factor, but also so that more people are aware of compiler issues and hopefully how to fix them).


T

-- 
Error: Keyboard not attached. Press F1 to continue. -- Yoon Ha Lee, CONLANG

On 02/20/2013 06:55 PM, H. S. Teoh wrote:
> On Thu, Feb 21, 2013 at 01:55:00AM +0100, Rob T wrote:
>> On Thursday, 21 February 2013 at 00:48:50 UTC, John Colvin wrote:
>> [...]
>>>>
>>>> Yes. Shared Libraries Please.
>>>>
>>>> Thank you.
>>>
>>> +1
>>>
>>> It's the most important issue for me. Particular feature bugs can
>>> be worked around, lack of shared library support is insurmountable
>>> for many applications.

I was playing with LuaD (https://github.com/JakobOvrum/LuaD) and was disappointed to learn that the only way to call D functions from Lua is for the lua code to be 'inside' a D module. (string l = "lua goes here"; or similar)

There are a few tricks you can perform, but it placed a very real limit on its usefulness.  Which is sad because other than that LuaD is beautiful.

>>
>> Yes, it's a major barrier towards adoption in a production
>> environment. Walter said it's now a priority, so I cross the fingers
>> and hope.
> [...]
>
> Better yet, maybe more of us should start delving into DMD's code...
> Currently only a small number of people really understand the compiler,
> which has a rather low bus factor. We need to increase that (not only to
> improve the bus factor, but also so that more people are aware of
> compiler issues and hopefully how to fix them).
>
>
> T
>

I only know enough C/C++ to know that I don't know enough...
If I could, I would totally go tearing through the source right now.

But some of it seems ambiguous to me.  I'm afraid I'd do more harm than good.

On 2013-02-21 01:55, Rob T wrote:

> Yes, it's a major barrier towards adoption in a production environment.
> Walter said it's now a priority, so I cross the fingers and hope.

Several pull request from Martin Nowak have already been merged in that last couple of days which implement parts of what's needed.

-- 
/Jacob Carlborg

On 2/20/2013 4:55 PM, Rob T wrote:
> On Thursday, 21 February 2013 at 00:48:50 UTC, John Colvin wrote: [...]
>>>
>>> Yes. Shared Libraries Please.
>>>
>>> Thank you.
>>
>> +1
>>
>> It's the most important issue for me. Particular feature bugs can be worked around, lack of shared library support is insurmountable for many applications.
> 
> Yes, it's a major barrier towards adoption in a production environment. Walter said it's now a priority, so I cross the fingers and hope.
> 
> --rt

It's _a_ barrier for adoption in _some_ production environments.  So are many other features for many other environments.  It's really easy to over color the priorities based on your own use case.  The community is large with lots of needs.  Juggling priorities is a very grey process.

Hope is about the least likely to produce results (other than maybe disappointment).  I highly encourage those with needs to actively help address them.  It doesn't take a lot of people to add a lot of value.

My 2 cents,
Brad

On 2/21/2013 12:07 AM, Brad Roberts wrote:
> Hope is about the least likely to produce results (other than maybe disappointment).  I highly encourage those with
> needs to actively help address them.  It doesn't take a lot of people to add a lot of value.

I agree. Even fixing minor stuff is surprisingly helpful.

On Thursday, 21 February 2013 at 00:57:46 UTC, H. S. Teoh wrote:
> On Thu, Feb 21, 2013 at 01:55:00AM +0100, Rob T wrote:
>> On Thursday, 21 February 2013 at 00:48:50 UTC, John Colvin wrote:
>> [...]
>> >>
>> >>Yes. Shared Libraries Please.
>> >>
>> >>Thank you.
>> >
>> >+1
>> >
>> >It's the most important issue for me. Particular feature bugs can
>> >be worked around, lack of shared library support is insurmountable
>> >for many applications.
>> 
>> Yes, it's a major barrier towards adoption in a production
>> environment. Walter said it's now a priority, so I cross the fingers
>> and hope.
> [...]
>
> Better yet, maybe more of us should start delving into DMD's code...
> Currently only a small number of people really understand the compiler,
> which has a rather low bus factor. We need to increase that (not only to
> improve the bus factor, but also so that more people are aware of
> compiler issues and hopefully how to fix them).
>
>
> T
I just want to say, as I have only worked in D for ~6 months the usage of Optilink is a real pain.
Couple of days ago I took a strong look at removing 64bit only code from Coff support and allowing it to compile 32bit.
I failed yes.
But I definitely will take another look into this.
Hopefully learn wayy more of how it works!

21-Feb-2013 12:48, Rikki Cattermole пишет:
> On Thursday, 21 February 2013 at 00:57:46 UTC, H. S. Teoh wrote:
>> On Thu, Feb 21, 2013 at 01:55:00AM +0100, Rob T wrote:
>>> On Thursday, 21 February 2013 at 00:48:50 UTC, John Colvin wrote:
>>> [...]
>>> >>
>>> >>Yes. Shared Libraries Please.
>>> >>
>>> >>Thank you.
>>> >
>>> >+1
>>> >
>>> >It's the most important issue for me. Particular feature bugs >can
>>> >be worked around, lack of shared library support is >insurmountable
>>> >for many applications.
>>>
>>> Yes, it's a major barrier towards adoption in a production
>>> environment. Walter said it's now a priority, so I cross the fingers
>>> and hope.
>> [...]
>>
>> Better yet, maybe more of us should start delving into DMD's code...
>> Currently only a small number of people really understand the compiler,
>> which has a rather low bus factor. We need to increase that (not only to
>> improve the bus factor, but also so that more people are aware of
>> compiler issues and hopefully how to fix them).
>>
>>
>> T
> I just want to say, as I have only worked in D for ~6 months the usage
> of Optilink is a real pain.

Checkout this alternative linker:
ftp://ftp.styx.cabel.net/pub/UniLink/

The only issue is that it converts debug info to some other format compatible with Borland's tools (TDS?).

> Couple of days ago I took a strong look at removing 64bit only code from
> Coff support and allowing it to compile 32bit.
> I failed yes.
> But I definitely will take another look into this.
> Hopefully learn wayy more of how it works!


-- 
Dmitry Olshansky