Kyle Furlong wrote:
> What are the obstacles to writing an open source D complier?

I would like to answer with a question. What was the obstacles before Walter to someone else to invent D? Where were the open-source guys?

It seems to
> me that having D be a one man show is not the most efficient way of bringing D into the mainstream, seeing as there is a healthy community of programmers here who all could lend a hand fixing bugs and adding features.

The short anwser:

"a healthy community of programmers" can't do what a "one man army" can.

The long answer:

Have you read the forum? Haven't you noticed that everyone here want different things from D. How this could possibly lead to a good language? If Walter respects all this whims D would become a half-scripting-slower-than-java-python-like-syntax-crap! I believe many people are missing the point. Of course everyone can write compile. For one it will be easier for other it will be harder. Hell! Even I can do it. But this is really not the point (in my opinion, of course). Just to have a new so-called language. To have a piece of art like D it takes someone wiser. Whit a lot of experience (not only in programming?). It's like.. you know. When you are kid. Your parents tell you something and you think to yourself "They are wrong! I'm right!".. Then you become (most of us) a little more mature and you see things from a different perspective. Then you think "What a fool I was...". In my modest opinion in this community most of us (I include myself too) are the kids. Walter is the adult (and few other guys too). Yes, it is hard for most people to accept they are wrong... But I think sometimes it is wiser to sit back and wait. Just accept that Walter is right. In most cases time will prove it... D couldn't be what it is without him. I believe that making D open source will ruin it. Of course many people have good suggestions. Even things that Walter missed. And many times Walter accepts these suggestions. I think we have to respect his work the way it is. If someone don't like it - well go and make E. What is the problem with that? If one man can do it then the healthy guys can do it twice as fast and maybe even better... Also what's the problem with GDC?

>
> "a healthy community of programmers" can't do what a "one man army" can.
>

> 
> If one man can do it then the healthy guys can do it twice as fast
> and maybe even better...

I understand that Walter is a very experienced, knowledgeable, intelligent individual and that he has the best in mind for the D language. I also understand that some members of the community are lacking in some of these qualities or have them to a lesser degree. I respect Walter for what he has done so far, its amazing really!

All that said, he *is* just one man, with the time considerations that a man has. How much better if, under the guidance of Walter and with his approval, knowledgeable and experienced members of the community were allowed to help work out bugs and push dmd towards a 1.0 that has been coming for 6 years, and has never arrived.

I get the feeling that the programming world is leaving D behind because the efforts of just one man, however heroic they may be, just cant keep up.

"Walter Bright" <newshound@digitalmars.com> wrote in message news:dfrat8$182t$2@digitaldaemon.com...
> Any language that is being used is a work in progress. I can point to any
> number of languages with mature, unchanging compilers and specifications,
> they'll be in the "dead language" bin <g>. Heck, I just put out another
> drop
> of the C++ compiler.

I'm not sure how that relates to my post, but I agree with you.

> A much more interesting question is "is the compiler/language in a state where I can effectively use it"?

Or perhaps "are the compiler/language/docs/library/community/tools in a state where I can be effective"? The expectations that programmers have of a "new language" nowadays include many things that have nothing to do with the compiler or language per se.

In article <dfrkpk$1hpn$1@digitaldaemon.com>, Kyle Furlong says...
>
>Walter Bright wrote:
>> Any language that is being used is a work in progress. I can point to any number of languages with mature, unchanging compilers and specifications, they'll be in the "dead language" bin <g>. Heck, I just put out another drop of the C++ compiler.
>> 
>> A much more interesting question is "is the compiler/language in a state where I can effectively use it"?
>> 
>> 
>
>I think the point/discussion I was trying to get at was that there is a large community of skilled, enthusiastic programmers here who I'm sure would like to help the d cause in a more intimate way.

*raises hand*

You said it.  I'm working flat-out on shoring up what I feel are massive gaps in D's overall capabilities.  But there's so much more to be done than just runtime linking, or enabling D for the web.

> If the dmd
>compiler needs to stay closed, fine, but use the community.

I agree with this.  If we look outside what additional stuff we want D to have, then the following are things that need to be completed within the existing (de-facto) specification.

- The D ABI (http://www.digitalmars.com/d/abi.html) is incomplete, which is a huge obstacle for implementing a runtime code-emit interface or even a whole compiler.  Anyone who has ever hacked on D binaries should seriously give this document a once-over.  Addendums and extensions should be placed on the Wiki somewhere.

- Rooting out bugs.  Presently, a good number of the bugs on the bugs newsgroup are likely to be faults in the frontend code.  If the community took more time to find out *where* in DMD things fail (and even attempt to patch it), Walter would have a faster turnaround for fixing things; this would push us to V1.0 faster.

- Wiki.  IMO, the Wiki is underutilized.  The community could use things like an almanac (listing of common and uncommon D idioms), and more in-depth proposals for any and all suggested improvements to D (the newsgroup works for collaboration, but is a lousy archive for docs).  Walter should be looking through the wiki comments to append to the existing documentation, but at the very least, he has linked the offical docs to the comments sections.

I'm sure there's more, but I can't think of anything else right now.  For example, I would love to see a self-hosted D compiler.  But until we get the basics out of the way, we're quite a way off from that goal.

- EricAnderton at yahoo

In article <dfrkpk$1hpn$1@digitaldaemon.com>, Kyle Furlong says...
>
>I think the point/discussion I was trying to get at was that there is a large community of skilled, enthusiastic programmers here who I'm sure would like to help the d cause in a more intimate way. If the dmd compiler needs to stay closed, fine, but use the community.

The compiler is one area where I think too many hands could ruin the soup. Small compiler development teams are not uncommon in any case.  EDG has a very small group of developers and they develop the reference implementation of C++. Beyond that, just go ahead and do whatever you think should be done.  Submitting Phobos changes can feel like a bit of a crapshoot but most make it in eventually.  And there's always Ares ;-)


Sean

In article <dfrsok$1p33$1@digitaldaemon.com>, =?ISO-8859-1?Q?Jari-Matti_M=E4kel=E4?= says...
>
>IMO we need to know, if Walter is going to dump Phobos in the near future. We desperately need a standard library that is well designed.

If Phobos ends up as the standard library for 1.0, will you use it despite the fact that you don't like it?  Personally, I don't intend to, though I may have to reconsider if I ever decide to write an open source application for D.  At the moment, I use Ares exclusively, even in its sparse state.  But then I don't have time for much outside of Ares anyway :)


Sean

In article <dfs3vh$1vnu$1@digitaldaemon.com>, pragma says...
>
>- Rooting out bugs.  Presently, a good number of the bugs on the bugs newsgroup are likely to be faults in the frontend code.  If the community took more time to find out *where* in DMD things fail (and even attempt to patch it), Walter would have a faster turnaround for fixing things; this would push us to V1.0 faster.

I've often considered this.  I don't suppose the DMD source would compile and run as shipped?


Sean

Kyle Furlong wrote:
> How much better if, under the guidance of Walter and with his approval, knowledgeable and experienced members of the community were allowed to help work out bugs and push dmd towards a 1.0 that has been coming for 6 years, and has never arrived.

Maybe... I don't know...

> 
> I get the feeling that the programming world is leaving D behind because the efforts of just one man, however heroic they may be, just cant keep up.

I don't feel left behind. I feel like I am leaving them behind. The programming world is going down driven by microsoft and sun with languages like java and .net bullshit. The programming is becoming drag-and-drop and I am happy that D is left behind this nonsense. D is actualy reamaning a programming language instead of slow scripting lamers friendly drag and drop gui designer. And after few years when all newborn programmers become a .net or java programmers we, the D programmers, will rule the world because we will be one of the not many left that can do a fast software and low level software like drivers, oses and stuff ;) (I am joking but in same time serious)

Kyle Furlong wrote:
> What are the obstacles to writing an open source D complier?


There _is_ an open source D compiler. Go check out GDC http://home.earthlink.net/~dvdfrdmn/d/.


> It seems to me that having D be a one man show is not the most efficient way of bringing D into the mainstream, seeing as there is a healthy community of programmers here who all could lend a hand fixing bugs and adding features.


I am affraid you are overestimating us, or underestimating the art of compiler construction / language design.

D is as much a one man show as it is because, given the above open source D compiler, so few people have the capacity or ability to fix bugs or add features.

IMHO, anyone who sees deficiencies in Walter's handling of D should ask himself: why don't I just go ahead and improve D?

And, the changelog shows that Walter is accepting patches, so just go ahead :-)


Bastiaan.

In article <dfs9jb$257v$1@digitaldaemon.com>, bobef says...
>
>> I get the feeling that the programming world is leaving D behind because the efforts of just one man, however heroic they may be, just cant keep up.
>
>I don't feel left behind. I feel like I am leaving them behind. The programming world is going down driven by microsoft and sun with languages like java and .net bullshit. The programming is becoming drag-and-drop and I am happy that D is left behind this nonsense.

It was the same thing during the dotcom boom.  But the popular trend has little to do with what's actually being done in a lot of programming houses.  And there are many applications that need to employ services that these environments don't provide.  Though I grant that both Java and .NET are improving rapidly in some realms that C++ has been a bit stagnant (concurrency, for example).  Personally, I  think that D is in a somewhat unique position, as it has constructs suiting both a systems language like C and from a higher-level language like Java.  For me, the synergy is ideal.


Sean