April 28, 2005
The link on the D links page yields a 404.  Where's Andy's site nowadays?

Also, is there a tutorial for this?

John Demme

On Thu, 2005-04-28 at 21:31 +1200, brad beveridge wrote:
> > What should be written (just brainstorming, here..) is a way to "tunnel" D method calls into C++ method calls, possibly via generated C functions in the short term, maybe more directly/transparently at some point in the future.
> > 
> > Kevin
> > 
> 
> D has been integrated with SWIG ... from D links.
> "Andy Friesen has taken SWIG  and modified it to generate code for D."
> 
> SWIG is basically a C/C++ parser that generates code that allows many other languages to interface to C/C++.  It produces very good proxy classes and function wrappers from what I have seen.
> 
> Unfortuantly it isn't rolled into the official SWIG website.  As with many things in D, the tools are 90% there, but are not official and have no maintainer.
> 
> Brad

April 28, 2005
John Demme wrote:
> The link on the D links page yields a 404.  Where's Andy's site
> nowadays?
> 
> Also, is there a tutorial for this?
> 
> John Demme
> 
> On Thu, 2005-04-28 at 21:31 +1200, brad beveridge wrote:
> 
>>>What should be written (just brainstorming, here..) is a way to "tunnel" D
>>>method calls into C++ method calls, possibly via generated C functions in the
>>>short term, maybe more directly/transparently at some point in the future.
>>>
>>>Kevin
>>>
>>
>>D has been integrated with SWIG ... from D links.
>>"Andy Friesen has taken SWIG  and modified it to generate code for D."
>>
>>SWIG is basically a C/C++ parser that generates code that allows many other languages to interface to C/C++.  It produces very good proxy classes and function wrappers from what I have seen.
>>
>>Unfortuantly it isn't rolled into the official SWIG website.  As with many things in D, the tools are 90% there, but are not official and have no maintainer.
>>
>>Brad
> 
> 
The SWIG site has tutorials for other languages, which should apply I think.  If you can't find the download I think I may have it on my machine at home.

Brad
April 28, 2005
Looks like there's a copy from August 1st, 2004 on DSource.

On Fri, 2005-04-29 at 09:17 +1200, Brad Beveridge wrote:
> John Demme wrote:
> > The link on the D links page yields a 404.  Where's Andy's site nowadays?
> > 
> > Also, is there a tutorial for this?
> > 
> > John Demme
> > 
> > On Thu, 2005-04-28 at 21:31 +1200, brad beveridge wrote:
> > 
> >>>What should be written (just brainstorming, here..) is a way to "tunnel" D method calls into C++ method calls, possibly via generated C functions in the short term, maybe more directly/transparently at some point in the future.
> >>>
> >>>Kevin
> >>>
> >>
> >>D has been integrated with SWIG ... from D links.
> >>"Andy Friesen has taken SWIG  and modified it to generate code for D."
> >>
> >>SWIG is basically a C/C++ parser that generates code that allows many other languages to interface to C/C++.  It produces very good proxy classes and function wrappers from what I have seen.
> >>
> >>Unfortuantly it isn't rolled into the official SWIG website.  As with many things in D, the tools are 90% there, but are not official and have no maintainer.
> >>
> >>Brad
> > 
> > 
> The SWIG site has tutorials for other languages, which should apply I think.  If you can't find the download I think I may have it on my machine at home.
> 
> Brad

April 28, 2005
In article <1114718207.20340.4.camel@localhost.localdomain>, John Demme says...
>
>...
>
>   Where's Andy's site nowadays?
>

Andy's site is now here:
http://aegisknight.org/~andy/d/

*** caution!!...it's slow to load up ***

David L.

-------------------------------------------------------------------
"Dare to reach for the Stars...Dare to Dream, Build, and Achieve!"
-------------------------------------------------------------------

MKoD: http://spottedtiger.tripod.com/D_Language/D_Main_XP.html
April 28, 2005
Simple solution to get D adopted--get Linus Torvalds to convert Linux to D.  Or perhaps some brave soul can fork it and convert the basic Linux toolchain and kernel to D.  I was recently thinking of how to automatedly convert C to a better language, and D looks as good as anything I was coming up with (although, I would like a syntax to expose member objects as if they were superclasses for a pretense of multiple inheritance).  Some kind of quick-and-dirty converter that just works, followed by a code cleanup at leisure.  Perhaps some linty tool that marks up the trouble spots for manual patching prior to compilation would help, too.

Maybe somebody can fix up a D environment in the Eclipse IDE without too much hassle?  C/C++ in Eclipse isn't here yet, but D would be much easier, probably not much harder than Java.  My $0.02...


April 28, 2005
Ben Hinkle wrote:
> "Vladimir" <kv11111@mail.ru> wrote in message news:d4qvgs$114p$1@digitaldaemon.com...
> 
>>Matthew wrote:
>>
>>
>>>"Georg Wrede" <georg.wrede@nospam.org> wrote in message
>>>news:42702E24.90904@nospam.org...
>>>
>>>>Benjamin Herr wrote:
>>>>
>>>>>He also thinks it is inconsistent to have (object)
>>>>>reference types and value types with no difference in syntax.
>>>>
>>>>Hmm. This has stuck in my eye before.
>>>>
>>>>Somehow, though, I haven't been able to formulate any specific
>>>>complaint about it. It's just a gut feeling: "this is not good".
>>>
>>>I have much the same feeling. My gut's usually right - and I believe
>>>it is in this case - but it can be somewhat ungenerous with its
>>>explanations. ;)
>>
>>I has the same feeling too.
>>
>>-- 
>>         Vladimir
> 
> MATLAB uses "." independent of value/reference semantics and it works fine. The difference between value/reference only shows up in assignment behavior (which is the definition of value/reference anyway). But then MATLAB also doesn't have pointers. 

IMHO, getting rid of -> and just using . was a good idea, but that is a different question.

I, like the others (it's nice to know I'm not alone) don't like the current syntax.  I've stated before my beefs with it:

1) Makes it hard to write a template that handles both struct pointers and object references correctly.
2) Leads to a very common newbie error: segfault because he used an uninitialized class reference.  (If we had used C++ syntax, then Java programmers would have had a learning curve...but their error would be caught at build time, not as a runtime error, which, IMHO, is far superior)
3) Made it so that class objects can't be on the stack.  Let's be honest, here, the whole reason that we have 'auto' is a kludge to get around the syntax problem.
4) Invited confusion over === and ==


Don't get me wrong - I still use D all the time.  Except for work, I don't even touch C and C++ anymore.  But if I ever write D++, this will be one thing that I will change. :)
April 28, 2005
Michael Hainze wrote:
> Simple solution to get D adopted--get Linus Torvalds to convert Linux to D.  Or
> perhaps some brave soul can fork it and convert the basic Linux toolchain and
> kernel to D.

Though I think you are jesting, it would be nice if D had a high-profile "killer app" that would make people sit up and take D seriously as a language.

As an aside - how hard would it be to make a toy OS kernel in D.  Or, if not a toy OS an embedded environment?

Brad
April 29, 2005
In article <d4r4vt$18hb$1@digitaldaemon.com>, Walter says...
>
>
>"J C Calvarese" <jcc7@cox.net> wrote in message news:d4n2nl$1hr$1@digitaldaemon.com...
>> I think you're right. There's nothing necessarily /wrong/ the Digital Mars website, but a little CSS and a few simple graphical elements can go a long way. Obviously jazzy web design isn't Walter's favorite hobby, but there are many D programmers who are skilled at designing flashy web pages.
>
>Is a flashy web page really an advantage? I see a lot of flashy web pages, and they don't impress me that much. The DM pages are for programmers, who (I assume) want to see a straightforward page that loads fast and is not gimmicky. Maybe I'm all wrong about this.

Walter, technically the D programming language pages are doing everything right.


I for one feel that the introduction page should make better use of whitespace to lead the eye down the page to the informative links (comparison, newsgroup, download, etc).

The left-hand nav, while being a great index for the manual, is actually a distraction as its available up-front.  It should probably be left behind a "reference" or "docs" link from a stripped-down main menu.  The menu itself very reference oriented, leaving the higher-traffic links at the bottom (download, tools, community, etc).

Also, as you say on the first page "This is the reference document".  Perhaps its time for an additional, more simple page?

Last, and certainly not least, a site's "look & feel" can say volumes (subconsiously) about how well the site, and the advertised product, are assembled.  It makes a statement about quality and commitment that is formulated much earlier (practically eons in terms of mental effort) than when the user even thinks about clicking "download".

I'll also add that the current rendition of the site, while improved, isn't much different than it was years ago.  In truth, I *left* that site after that first visit, feeling lost, only to revisit months later.  D's purpose simply wasn't conveyed to me easily, and the stability of the product was punctuated with a large question-mark.  IMO, the DNG has been more helpful in conveying D's promise and mission than the site (however actions *do* speak louder than words, so please take this with a mountain of salt). :(


>
>I do agree that I am not going to win any awards for my english composition, and if anyone has any suggestions on rewrites for any part of the text, I'm all ears!
>

If you don't mind, I'll withold the rest of my critiques, as I'm still formulating them.  Plus, I believe that I can probably do the topic more justice by trying a few things offline.  If anything comes of this exercise, I'll share my findings here.

- EricAnderton at yahoo
April 29, 2005
>> MATLAB uses "." independent of value/reference semantics and it works fine. The difference between value/reference only shows up in assignment behavior (which is the definition of value/reference anyway). But then MATLAB also doesn't have pointers.
>
> IMHO, getting rid of -> and just using . was a good idea, but that is a different question.
>
> I, like the others (it's nice to know I'm not alone) don't like the current syntax.  I've stated before my beefs with it:
>
> 1) Makes it hard to write a template that handles both struct pointers and object references correctly.

I'm not sure what you are referring to. Can you give a pointer to a post or examples?

> 2) Leads to a very common newbie error: segfault because he used an uninitialized class reference.  (If we had used C++ syntax, then Java programmers would have had a learning curve...but their error would be caught at build time, not as a runtime error, which, IMHO, is far superior)

IMO better segv than slicing objects passed by value. Any non-trivial class hierarchy is destined to be manipulated by reference.

> 3) Made it so that class objects can't be on the stack.  Let's be honest, here, the whole reason that we have 'auto' is a kludge to get around the syntax problem.

I'm with Walter (and Java/C#) that class objects on the stack should be rare. If there are issues with structs that require lots of auto classes then structs need attention. Actually auto classes don't have anything to do with being on the stack - they just force the dtor to be called on scope exit. IMO I haven't seen a D class that should be auto (I've seen good uses auto variables but declaring a class as auto forces all uses to be auto).

> 4) Invited confusion over === and ==

Agreed - though the current design isn't unreasonable in the big picture.


April 29, 2005
In article <d4rptj$1vuc$1@digitaldaemon.com>, Brad Beveridge says...
>
>Michael Hainze wrote:
>> Simple solution to get D adopted--get Linus Torvalds to convert Linux to D.  Or perhaps some brave soul can fork it and convert the basic Linux toolchain and kernel to D.
>
>Though I think you are jesting, it would be nice if D had a high-profile "killer app" that would make people sit up and take D seriously as a language.
>
>As an aside - how hard would it be to make a toy OS kernel in D.  Or, if not a toy OS an embedded environment?

Didn't Sun already try this for Java?  Seems to me that maybe they were onto something with pushing their language/platform in such a manner. :)

Outside that, the "Killer app" territory is likely to be covered by a mature project built on top of what we already have.  Right now, we're still building the toolchain (build and the like) and compatibility/platform libs.

Personally, I'm doing my damndest to get some good stuff out there.  I doubt I'm in "killer app" territory with DSP and it's XML stack, but those will pay dividends later.

FYI, I have a modified version of the DMDFE that spits out XML instead of .obj files.  This was actually pretty easy to do, thanks to Ben's pioneering. I plan to use it as part of a documentation toolchain, which will likely be XSLT driven.

IMO, what's needed is something that makes people go "wow" first, and "made with what?" second. ;)

- EricAnderton at yahoo