Domaintalk - A Proposed New Project

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Domaintalk - A Proposed New Project

Bob Williams
For some time I have been developing the idea of a extension to Smalltalk, which I call Domaintalk. That extension would have "areas" where the language natural to the problem being solved would be used. For example, when solving a numerical problem then mathematical notation would be used or when accessing a relational database SQL would be used. In other words, each of those area would correspond to the language used by "Domain Experts", thus the name domain.The rational behind this is that when you ask a domain expert how to solve a particular problem they will give you the solution in their natural language and you must then convert that into Smalltalk code. Why not just eliminate the last step an let the compiler for the domain produce a compiled method that can be used by the VM--a domain specific would be required in addition to the compiler.

There are two reasons for my posting this email: (1) I want to measure the interest of this group in such a project, and (2) I want to know if there is a better place to post a much longer description of the project or should I just break the description into smaller segments and post each under a different subject?
Reply | Threaded
Open this post in threaded view
|

Re: Domaintalk - A Proposed New Project

EstebanLM
you have to checkout helvetia, by Lukas Renggli:


cheers, 
Esteban

On 21 Apr 2014, at 16:17, Bob Williams <[hidden email]> wrote:

For some time I have been developing the idea of a extension to Smalltalk, which I call Domaintalk. That extension would have "areas" where the language natural to the problem being solved would be used. For example, when solving a numerical problem then mathematical notation would be used or when accessing a relational database SQL would be used. In other words, each of those area would correspond to the language used by "Domain Experts", thus the name domain.The rational behind this is that when you ask a domain expert how to solve a particular problem they will give you the solution in their natural language and you must then convert that into Smalltalk code. Why not just eliminate the last step an let the compiler for the domain produce a compiled method that can be used by the VM--a domain specific would be required in addition to the compiler.

There are two reasons for my posting this email: (1) I want to measure the interest of this group in such a project, and (2) I want to know if there is a better place to post a much longer description of the project or should I just break the description into smaller segments and post each under a different subject?

Reply | Threaded
Open this post in threaded view
|

Re: Domaintalk - A Proposed New Project

jfabry

I second that, it was exactly what I was thinking !

On Apr 21, 2014, at 11:25 AM, Esteban Lorenzano <[hidden email]> wrote:

> you have to checkout helvetia, by Lukas Renggli:
>
> http://scg.unibe.ch/research/helvetia
>
> cheers,
> Esteban
>
> On 21 Apr 2014, at 16:17, Bob Williams <[hidden email]> wrote:
>
>> For some time I have been developing the idea of a extension to Smalltalk, which I call Domaintalk. That extension would have "areas" where the language natural to the problem being solved would be used. For example, when solving a numerical problem then mathematical notation would be used or when accessing a relational database SQL would be used. In other words, each of those area would correspond to the language used by "Domain Experts", thus the name domain.The rational behind this is that when you ask a domain expert how to solve a particular problem they will give you the solution in their natural language and you must then convert that into Smalltalk code. Why not just eliminate the last step an let the compiler for the domain produce a compiled method that can be used by the VM--a domain specific would be required in addition to the compiler.
>>
>> There are two reasons for my posting this email: (1) I want to measure the interest of this group in such a project, and (2) I want to know if there is a better place to post a much longer description of the project or should I just break the description into smaller segments and post each under a different subject?
>



---> Save our in-boxes! http://emailcharter.org <---

Johan Fabry   -   http://pleiad.cl/~jfabry
PLEIAD lab  -  Computer Science Department (DCC)  -  University of Chile


Reply | Threaded
Open this post in threaded view
|

Re: Domaintalk - A Proposed New Project

kilon.alios
Sounds interesting, maybe you could use Pillar to outline the basic features you are aiming for and detail how you will achieve them. Then you can link it back here for discussion, this way people wont have to go back through their emails to find a description for it. 

I was thinking something similar for Graphics and Music which are areas that interest me, I am interested into reading other people opinions as I am not very familiar with DSLs. 

I was just studying CSOUND the DSL for audio synthesis and music composition and dreaming how cool it would feel to have that used from inside Pharo as an extension to it. 




On Mon, Apr 21, 2014 at 6:01 PM, Johan Fabry <[hidden email]> wrote:

I second that, it was exactly what I was thinking !

On Apr 21, 2014, at 11:25 AM, Esteban Lorenzano <[hidden email]> wrote:

> you have to checkout helvetia, by Lukas Renggli:
>
> http://scg.unibe.ch/research/helvetia
>
> cheers,
> Esteban
>
> On 21 Apr 2014, at 16:17, Bob Williams <[hidden email]> wrote:
>
>> For some time I have been developing the idea of a extension to Smalltalk, which I call Domaintalk. That extension would have "areas" where the language natural to the problem being solved would be used. For example, when solving a numerical problem then mathematical notation would be used or when accessing a relational database SQL would be used. In other words, each of those area would correspond to the language used by "Domain Experts", thus the name domain.The rational behind this is that when you ask a domain expert how to solve a particular problem they will give you the solution in their natural language and you must then convert that into Smalltalk code. Why not just eliminate the last step an let the compiler for the domain produce a compiled method that can be used by the VM--a domain specific would be required in addition to the compiler.
>>
>> There are two reasons for my posting this email: (1) I want to measure the interest of this group in such a project, and (2) I want to know if there is a better place to post a much longer description of the project or should I just break the description into smaller segments and post each under a different subject?
>



---> Save our in-boxes! http://emailcharter.org <---

Johan Fabry   -   http://pleiad.cl/~jfabry
PLEIAD lab  -  Computer Science Department (DCC)  -  University of Chile



Reply | Threaded
Open this post in threaded view
|

Re: Domaintalk - A Proposed New Project

Reza Razavi
In reply to this post by Bob Williams
That's an attractive and quite largely studied topic. If Etoys [1] appears a typical example of such extension "areas", I'd recommend posting also to their mailing list(s), while explaining how your idea compares to Etoys (would help generating more interest). 

[1] http://en.wikipedia.org/wiki/Etoys_(programming_language)
[2] http://www.squeakland.org/discuss/

On April 21, 2014 at 4:17 PM Bob Williams [hidden email] wrote:

For some time I have been developing the idea of a extension to Smalltalk, which I call Domaintalk. That extension would have "areas" where the language natural to the problem being solved would be used. For example, when solving a numerical problem then mathematical notation would be used or when accessing a relational database SQL would be used. In other words, each of those area would correspond to the language used by "Domain Experts", thus the name domain.The rational behind this is that when you ask a domain expert how to solve a particular problem they will give you the solution in their natural language and you must then convert that into Smalltalk code. Why not just eliminate the last step an let the compiler for the domain produce a compiled method that can be used by the VM--a domain specific would be required in addition to the compiler.

There are two reasons for my posting this email: (1) I want to measure the interest of this group in such a project, and (2) I want to know if there is a better place to post a much longer description of the project or should I just break the description into smaller segments and post each under a different subject?