Heh.

Walter removed my post mentioning that I'd prettyprinted DMDScript, arranged it into folders, and had activex half working?

It'll be on the web elsewhere once I finish getting activex working.  That's all I can say.

I'll be sure to keep you guys informed.

Update:

I've currently got:
- a non-functional ActiveXObject() in DScript;
- a functional ActiveXObject() in D;

I derived the functional ActiveXObject from the one provided in Core32, and ported it over to the MinGW work done by Stewart Gordon (which had a bug with ITypeInfo btw)

Now I have to:
- Put the two together to bind ActiveXObject in DScript to actual COM
- Write the Put/Get/Call routines for the ActiveXObject so they don't require
special methods (ex: ie.set("Visible",toVariant(true)); )

It's now 1pm here, if I'm lucky I'll have it done before bedtime. *crosses fingers*

Sincerely,
Dan

Dan wrote:
> Update:
> 
> I've currently got:
> - a non-functional ActiveXObject() in DScript;
> - a functional ActiveXObject() in D;

Sounds great!

> 
> I derived the functional ActiveXObject from the one provided in Core32, and ported
> it over to the MinGW work done by Stewart Gordon (which had a bug with ITypeInfo btw)

I assume your referring to this project:
http://www.prowiki.org/wiki4d/wiki.cgi?WindowsAPI

What was the error? Did you tell Stewart about it?


> Now I have to:
> - Put the two together to bind ActiveXObject in DScript to actual COM
> - Write the Put/Get/Call routines for the ActiveXObject so they don't require
> special methods (ex: ie.set("Visible",toVariant(true)); )
> 
> It's now 1pm here, if I'm lucky I'll have it done before bedtime.
> *crosses fingers*

-- 
jcc7

jcc,

Yes, that's the project.  No I didn't tell him.

The error was that some of his objbase.d functions expected ITypeInfo**'s, instead of just ITypeInfo*'s.  I just ran into two of them; and found it worked with just ITypeInfo*'s (actually ran)

I have put the two parts together, and now have an ActiveXObject within the
DScript engine.  When I hard code ie.set("Visible",toVariant(true)); etc, it works.

The remaining issue now is to bind the working ActiveXObject to Dscript.  That will involve some heavy modification of loadMembers() etc.

jcc, I just realized that nothing on dsource.org is licensed.  It just claims to be "Open Source Development".  Any idea where I could get licensing information (to cover myself?)

== Quote from Dan (murpsoft@hotmail.com)'s article
> jcc,
>
> Yes, that's the project.  No I didn't tell him.
>
> The error was that some of his objbase.d functions expected ITypeInfo**'s, instead of just ITypeInfo*'s.  I just ran into two of them; and found it worked with just ITypeInfo*'s (actually ran)

Oh, I think I ran across the same thing the other day. It seems to work when I used a cast.


> I have put the two parts together, and now have an ActiveXObject
> within the DScript engine.  When I hard code
> ie.set("Visible",toVariant(true)); etc, it works.

That's great!


> The remaining issue now is to bind the working ActiveXObject to Dscript.  That will involve some heavy modification of loadMembers() etc.
>
> jcc, I just realized that nothing on dsource.org is licensed.  It
> just claims to be "Open Source Development".  Any idea where I could > get
licensing information
> (to cover myself?)

Is "Placed into public domain" good enough for your purposes?

I think the header of each file has the same public domain notice: http://www.dsource.org/projects/bindings/browser/trunk/win32/w32api.d

I think they picked the MinGW headers as a basis because they are public domain and the intention is to have the D port be public domain, too.

jcc7

Dan escribió:
> jcc,
> 
> Yes, that's the project.  No I didn't tell him.
> 
> The error was that some of his objbase.d functions expected ITypeInfo**'s, instead
> of just ITypeInfo*'s.  I just ran into two of them; and found it worked with just
> ITypeInfo*'s (actually ran)
> 
> I have put the two parts together, and now have an ActiveXObject within the
> DScript engine.  When I hard code ie.set("Visible",toVariant(true)); etc, it works.
> 
> The remaining issue now is to bind the working ActiveXObject to Dscript.  That
> will involve some heavy modification of loadMembers() etc.
> 

I think I solved some of those things in Walnut. Check its (old and abandoned) source at DSource.

> jcc, I just realized that nothing on dsource.org is licensed.  It just claims to
> be "Open Source Development".  Any idea where I could get licensing information
> (to cover myself?)


-- 
Carlos Santander Bernal

Okay Carlos, I'll check it out.

Glad to know I don't have to bury the work because of licensing.  : )

I'm gonna be off for the rest of the day.  I'm sick with the flu.

I tried to sleuth down the source for walnut, but it's not visible online to either google or dsource.org/ that I could find.

If someone could direct me to the location, I'd appreciate it.  Thanks, Dan.

Dan wrote:

> I tried to sleuth down the source for walnut, but it's not visible online to either google or dsource.org/ that I could find.
> 
> If someone could direct me to the location, I'd appreciate it.  Thanks, Dan.

You should get to the subversion repository using the below

svn co http://svn.dsource.org/projects/walnut walnut

-- 
Lars Ivar Igesund
blog at http://larsivi.net
DSource, #d.tango & #D: larsivi
Dancing the Tango

Okay, from what I've read, the stuff that was part of walnut was related to Core32
and MinGW (sp?)

I borrowed code from jcc7 and carlos already as public domain.

I created a DScript native object, the ActiveXObject, and attached the com functionality from the borrowed code.  I've since modified toVariant to make it implicit.  It's now done when you set, setByRef or call anything.

I'm now going to try to expose the names of the ActiveXObject set/get/call "properties/methods" to DScript, and then work on being able to call them.