www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Where the schedule of D development?

reply "Boris Wang" <nano.kago hotmail.com> writes:
Walter, can you make a schedule for development of D?

The silence make me so dysphoric. 
Apr 26 2006
parent reply Walter Bright <newshound digitalmars.com> writes:
Boris Wang wrote:
 Walter, can you make a schedule for development of D?
 
 The silence make me so dysphoric. 

There isn't a schedule. We just move forward day by day, trying to do the most important things first.
Apr 26 2006
parent reply antonio <antonio abrevia.net> writes:
Walter Bright escribió:
 Boris Wang wrote:
 Walter, can you make a schedule for development of D?

 The silence make me so dysphoric. 

There isn't a schedule. We just move forward day by day, trying to do the most important things first.

1.- Lets go: move forward month by month or months by months... the need is the key, not the time to solve them (Momo's concept): Debugger, please. 2.- Time to unify the framework (Ares, Mango, ...) 3.- Time to be prof : lets plann the future. Just... follow the Mono's road map... I'm sure than a professional work will give productive results... investments... and so on. Sorry. I have a really poor english. Antonio
Apr 26 2006
next sibling parent nick <nick.atamas gmail.com> writes:
Seconded.

antonio wrote:
 Walter Bright escribió:
 Boris Wang wrote:
 Walter, can you make a schedule for development of D?

 The silence make me so dysphoric. 

There isn't a schedule. We just move forward day by day, trying to do the most important things first.

1.- Lets go: move forward month by month or months by months... the need is the key, not the time to solve them (Momo's concept): Debugger, please. 2.- Time to unify the framework (Ares, Mango, ...) 3.- Time to be prof : lets plann the future. Just... follow the Mono's road map... I'm sure than a professional work will give productive results... investments... and so on. Sorry. I have a really poor english. Antonio

Apr 26 2006
prev sibling next sibling parent reply "Lynn Allan" <l_d_allan adelphia.net> writes:
 Just... follow the Mono's road map...

Hmmmmm ... their Feb, 2006 webpage mentions .Net framework 1.1 being the latest ... that isn't a confidence builder. OS/2 tried (and failed) to keep up with with Win 3.1x compatibility ... while Win9x and WinNt "cleaned its clock." Some observations: * The battle for the desktop has been over since the DotCom bust. Look at RedHat's long term stock chart. Windows is "out in front ... and pulling away" (on the desktop). I write this as a developer with a crushed os/2 career still relatively fresh in my mind, so please don't consider me a Redmond toady. * A prediction: Steven J. will grow tired of computer outflows draining iPod inflows. Who needs another Board of Directors mutiny? Does portability matter? * Very few people have made any money betting against Bill Gates. (The W.B. being perhaps a notable exception, although that is probably a long story <g>) A Modest Proposal: I speculate that one or more of the Very Bright People who participate on this list (certainly not including me) would happily work for Digital Mars for significantly less than their current hourly rate. I further suspect there are Bright wannabe people on this list who would work for Digital Mars for $1.00 per hour. With the right smokin' material, I would even further speculate that "passing the hat" to pay those people might be semi-viable. D is certainly not a committee effort (A Good Thing), but my impression is that it is mature enough that an obligarchy would be appropriate rather than the current monarchy. I doubt even Nicolas Wirth could invent his third language today, and get all the libraries to be competitive in a reasonable time-frame ... (much less the nearly all-important gui debugger. Who besides Bright people can get productive with a complicated language without a world class debugger? Certainly not this plodder ... who is less than the Bright-est bulb in the box) An uninformed opinion: Boost is "out in front" Is it "pulling away?" Is the window of opportunity half open or half closed?
Apr 28 2006
next sibling parent reply Lucas Goss <lgoss007 gmail.com> writes:
Lynn Allan wrote:
 Just... follow the Mono's road map...

Hmmmmm ... their Feb, 2006 webpage mentions .Net framework 1.1 being the latest ... that isn't a confidence builder.

The .Net 1.1 framework was implemented in Mono 1.0 and released on June 30th, 2004. Their newest branch, Mono 1.2, is in the final stages of release and has Generic types support and "execution system and core class libraries (C# 2.0)." It also includes technology previews for XML 2.0, ASP.NET 2.0, ADO.NET 2.0, and more. The main thing it is behind on is the GUI, but the System.Windows.Forms 1.1 is added in the 1.2 release. In MonoDevelop there is also a GUI designer which works very well from what I've tested. I have to use .NET for work, and while I don't really like most things MS does (lock out competition and more), the .NET library is very nice to work with and mono is keeping up nicely. That being said, the Mono roadmap was one of the clearest development paths that I've seen, listing every function/class/documentation/etc. (http://mono.ximian.com/class-status/mono-HEAD-vs-fx-2/index.html) that needed to be implemented and listed the status of each. Not to mention the site is very clean and the documentation is very easy to read (I haven't found Java or D to be as nice). I'd say Mono has a fabulous roadmap that is very inspiring. I could only hope that the D community could come up with something half as good. Lucas
Apr 28 2006
next sibling parent reply Lars Ivar Igesund <larsivar igesund.net> writes:
Lucas Goss wrote:


 That being said, the Mono roadmap was one of the clearest development
 paths that I've seen, listing every function/class/documentation/etc.
 (http://mono.ximian.com/class-status/mono-HEAD-vs-fx-2/index.html) that
 needed to be implemented and listed the status of each. Not to mention
 the site is very clean and the documentation is very easy to read (I
 haven't found Java or D to be as nice). I'd say Mono has a fabulous
 roadmap that is very inspiring. I could only hope that the D community
 could come up with something half as good.
 
 Lucas

It is easy to make a detailed roadmap when you just reimplement an existing API. -- Lars Ivar Igesund blog at http://larsivi.net DSource & #D: larsivi
Apr 28 2006
next sibling parent Lucas Goss <lgoss007 gmail.com> writes:
Lars Ivar Igesund wrote:
 It is easy to make a detailed roadmap when you just reimplement an existing
 API.

Ah, I probably should have left out the API link (for functions/etc.), my mistake. Yes an API would make that easier. They still have a roadmap for all of their projects though, which includes much more than just the API from .NET. It's just easier for a community to go the same direction if they have a target to aim at. Lucas
Apr 28 2006
prev sibling parent Antonio <antonio abrevia.net> writes:
Lars Ivar Igesund escribió:
 Lucas Goss wrote:
 
 
 That being said, the Mono roadmap was one of the clearest development
 paths that I've seen, listing every function/class/documentation/etc.
 (http://mono.ximian.com/class-status/mono-HEAD-vs-fx-2/index.html) that
 needed to be implemented and listed the status of each. Not to mention
 the site is very clean and the documentation is very easy to read (I
 haven't found Java or D to be as nice). I'd say Mono has a fabulous
 roadmap that is very inspiring. I could only hope that the D community
 could come up with something half as good.

 Lucas

It is easy to make a detailed roadmap when you just reimplement an existing API. -- Lars Ivar Igesund blog at http://larsivi.net DSource & #D: larsivi

infraestructures are a very good framework.. (not only an API: is the corel key of this platforms).
Apr 28 2006
prev sibling parent reply Walter Bright <newshound digitalmars.com> writes:
Lucas Goss wrote:
 That being said, the Mono roadmap was one of the clearest development 
 paths that I've seen, listing every function/class/documentation/etc.
 (http://mono.ximian.com/class-status/mono-HEAD-vs-fx-2/index.html) that 
 needed to be implemented and listed the status of each. Not to mention 
 the site is very clean and the documentation is very easy to read (I 
 haven't found Java or D to be as nice). I'd say Mono has a fabulous 
 roadmap that is very inspiring. I could only hope that the D community 
 could come up with something half as good.

Mono does have the advantage that they're not inventing or designing anything, they have a complete specification and complete documentation already (from Microsoft), so it's just filling in between the lines.
Apr 28 2006
parent reply Gabe McArthur <Gabe_member pathlink.com> writes:
 I'd say Mono has a fabulous 
 roadmap that is very inspiring. I could only hope that the D community 
 could come up with something half as good.


I agree: http://www.gnu-d.org Let's get started.
Apr 28 2006
next sibling parent reply Alberto Simon <Alberto_member pathlink.com> writes:
In article <e2u6bf$10o6$1 digitaldaemon.com>, Gabe McArthur says...
 I'd say Mono has a fabulous 
 roadmap that is very inspiring. I could only hope that the D community 
 could come up with something half as good.


I agree: http://www.gnu-d.org Let's get started.

It seems like you are determined to make this happen. But to prevent this from fading away this needs leaders, someone or a few person to organize this thing (the page is a nice idea BTW) into something concrete. In order to make it happen this needs a roadmap and goals and everything. So, where are the volunteers? Regards, Alberto Simon
Apr 28 2006
parent =?iso-8859-1?q?Knud_S=F8rensen?= <12tkvvb02 sneakemail.com> writes:
On Sat, 29 Apr 2006 05:35:26 +0000, Alberto Simon wrote:

 In article <e2u6bf$10o6$1 digitaldaemon.com>, Gabe McArthur says...
 I'd say Mono has a fabulous 
 roadmap that is very inspiring. I could only hope that the D community 
 could come up with something half as good.


I agree: http://www.gnu-d.org Let's get started.

It seems like you are determined to make this happen. But to prevent this from fading away this needs leaders, someone or a few person to organize this thing (the page is a nice idea BTW) into something concrete. In order to make it happen this needs a roadmap and goals and everything. So, where are the volunteers?

I will suggest that we organise this in a bottom up organisation structure. The organisation should be divided into project groups. Then volunteers should join the project groups they like to develop for. Then each group should have a leader and a group member should be able to challenge the group leader for the leader post. When the leader is challenges the whole group vote for which of the two should be the next leader. In this way the group can fast replace a inactive group leader such that the project can continue with maximal momentum. Then in similar way should the group leaders chose a leader for the whole gnu-d project.
Apr 29 2006
prev sibling parent Walter Bright <newshound digitalmars.com> writes:
Gabe McArthur wrote:
 I'd say Mono has a fabulous 
 roadmap that is very inspiring. I could only hope that the D community 
 could come up with something half as good.


I agree: http://www.gnu-d.org Let's get started.

I added your site to www.digitalmars.com/d/dlinks.html. Also, could you please somewhere on the page have the phrase "D programming language", and include it on the template for all the site pages? This will make it findable via google. Thanks!
Apr 28 2006
prev sibling parent Antonio <antonio abrevia.net> writes:
Lynn Allan escribió:
 Just... follow the Mono's road map...

Hmmmmm ... their Feb, 2006 webpage mentions .Net framework 1.1 being the latest ... that isn't a confidence builder. OS/2 tried (and failed) to keep up with with Win 3.1x compatibility ... while Win9x and WinNt "cleaned its clock."

to propouse a "common path"...
 
 Some observations:
 * The battle for the desktop has been over since the DotCom bust. Look
 at RedHat's long term stock chart. Windows is "out in front ... and
 pulling
 away" (on the desktop). I write this as a developer with a crushed
 os/2 career still relatively fresh in my mind, so please don't
 consider me a Redmond toady.
 
 * A prediction: Steven J. will grow tired of computer outflows
 draining iPod inflows. Who needs another Board of Directors mutiny?
 Does portability matter?
 
 * Very few people have made any money betting against Bill Gates. (The
 W.B. being perhaps a notable exception, although that is probably a
 long story <g>)
 
 A Modest Proposal:
 I speculate that one or more of the Very Bright People who participate
 on this list (certainly not including me) would happily work for
 Digital Mars for significantly less than their current hourly rate.

 
 I further suspect there are Bright wannabe people on this list who
 would work for Digital Mars for $1.00 per hour. With the right smokin'
 material, I would even further speculate that "passing the hat" to pay
 those people might be semi-viable.
 

 D is certainly not a committee effort (A Good Thing), but my
 impression is that it is mature enough that an obligarchy would be
 appropriate rather than the current monarchy. I doubt even Nicolas
 Wirth could invent his third language today, and get all the libraries
 to be competitive in a reasonable time-frame ...
 
 (much less the nearly all-important gui debugger. Who besides Bright
 people can get productive with a complicated language without a world
 class debugger? Certainly not this plodder ... who is less than the
 Bright-est bulb in the box)
 
 An uninformed opinion: Boost is "out in front"   Is it "pulling away?"
 Is the window of opportunity half open or half closed?
 

I think than development market has a gap not really covered by any other languaje: * A lot of peopple need's to work with a primary languaje (without virtual machine)... C & C++ (Pascal?) are the key: When you need to work in a very low level and not really big project... C is the best one. If you need a best structured possibility: C++ is a good option (with a hard learn curve... you need only a good project director, very clean development rules and a good quality verification fase for each development fase). When you need a fast structured hight level development... then Java/C# is a good option (You need the same C++ project development rules, but you have a good isolation/independent facilities and a good learn curve) * D is just between C++ & Java/C#: Not Virtual machine, hight abstraction level, good learn curve and, here you are the key, D flavour and Name: a C# alternative focused as a C++/C evolution. D only needs a good infraestructure to cover the gap: Best library support. A complete debugger (future ide is optional) Documentation. Your proposal is clean and acceptable: a "proto" company with a good team producing a core profesional product. Money could be provided by us (the community) and the objective could be provided by the own company. My propossal: 1.- Create the core team with the best ones. 2.- The team has to prepare the future D roadmap. 3.- With the roadmap, people can decide if D future covers their needs or if D must be abandoned. 4.- Let's go with the company. The market oportunity is waiting for "D"...
Apr 28 2006
prev sibling parent reply ElfQT <ElfQT_member pathlink.com> writes:
..."Debugger, please."

Not even a debugger, only debugging info from the compiler. I've stated earlier how I miss good debugging support. (I don't even want to start on why unit testing, while it is great and powerful, isn't enough.) Even I've tried to support a D debugging/debugger project, until it came clear that the executables doesn't contain needed debug info (only source numbers and some basic stuff). And since the compiler is in Walter's hand, and backend is closed, there is no other way to include debug info, only if he works on it. I've never worked with compiler technology, and only scratched the surface with debugging information in debug executables, I wonder what it takes to be able to read/watch (all) D variables. After that, to incorporate debugging into any tool/IDE seem a lesser problem to me. Unfortunately, I've never heard Walter on this topic, is this task a hard one, or where the problem lies (I've heard codeview isn't a nice format, and the new MS "pdb" format is kind of MS internal), or is there a way to contribute to this..., is it planned after all ... only different deflective opinions form the D community. (And frankly, this is the one technical thing that keeps me from working with D (again), and only dare to raise the question again, because I see similar opinions.)
Apr 28 2006
next sibling parent Roald Ribe <rr.nospam nospam.teikom.no> writes:
ElfQT wrote:
 ..."Debugger, please."

Not even a debugger, only debugging info from the compiler.

Useless request unless specifying format I think. DWARF is the only way to go for being able to use free/open debuggers.
 I've stated earlier how I miss good debugging support. (I don't even want to
 start on why unit testing, while it is great and powerful, isn't enough.)
 Even I've tried to support a D debugging/debugger project, until it came clear
 that the executables doesn't contain needed debug info (only source numbers and
 some basic stuff).
 And since the compiler is in Walter's hand, and backend is closed, there is no
 other way to include debug info, only if he works on it.

In DMD, not in GDC. I would guess (have not used D) that gdc produces DWARF, I read a while ago that the D symbol name mangling caused trouble for gdb. Don't know what the status is on that.
 I've never worked with compiler technology, and only scratched the surface with
 debugging information in debug executables, I wonder what it takes to be able
to
 read/watch (all) D variables.

Producing debug info for the symbols.
 After that, to incorporate debugging into any tool/IDE seem a lesser problem to
 me.
 
 Unfortunately, I've never heard Walter on this topic, is this task a hard one,
 or where the problem lies (I've heard codeview isn't a nice format, and the new
 MS "pdb" format is kind of MS internal), or is there a way to contribute to
 this..., is it planned after all ...
 only different deflective opinions form the D community.

pdb is not fully documented/known, and therefore useless outside MS tools. codeview is known, but old an limited. It may be that Walter is "saving" debugger support for a paid version of DMD. Unless he comes out and tells otherwise, I assume at this point that if you want debugger support in a free package, gdc is the way to go. Roald
Apr 28 2006
prev sibling parent reply Walter Bright <newshound digitalmars.com> writes:
ElfQT wrote:
 Unfortunately, I've never heard Walter on this topic, is this task a hard one,
 or where the problem lies (I've heard codeview isn't a nice format, and the new
 MS "pdb" format is kind of MS internal), or is there a way to contribute to
 this..., is it planned after all ...
 only different deflective opinions form the D community.

The Windows dmd compiler does emit symbolic debug info in Codeview 4 format. Unfortunately, Microsoft appears to have dropped support for it in its current debuggers. So, I have included starting with 0.155 the old windbg debugger, which does support it. On Linux, only symbols and line number info is generated, mainly because I haven't figured out the dwarf debug symbol format yet.
Apr 28 2006
next sibling parent Roald Ribe <rr.nospam nospam.teikom.no> writes:
Walter Bright wrote:
 ElfQT wrote:
 Unfortunately, I've never heard Walter on this topic, is this task a 
 hard one,
 or where the problem lies (I've heard codeview isn't a nice format, 
 and the new
 MS "pdb" format is kind of MS internal), or is there a way to 
 contribute to
 this..., is it planned after all ...
 only different deflective opinions form the D community.

The Windows dmd compiler does emit symbolic debug info in Codeview 4 format. Unfortunately, Microsoft appears to have dropped support for it in its current debuggers.

The Open Watcom debugger (wd & wdw) supports CV4 NB09 (NB09 produced with cvpack). I have no idea if it will work with D. Road
Apr 28 2006
prev sibling parent "news.digitalmars.com" <dethjunk yahoo.com> writes:
 The Windows dmd compiler does emit symbolic debug info in Codeview 4 
 format. Unfortunately, Microsoft appears to have dropped support for it in 
 its current debuggers. So, I have included starting with 0.155 the old 
 windbg debugger, which does support it.

Thanks for the clarification. I will reflect on the topic with a new post soon. ElfQT
May 04 2006