Garbage Collection vs. the DSA

Garbage Collection vs. the DSA

[W. Wesley Groleau (Wes)]

--
R. Dewar says:
> ... clear evidence (real $$ coming in) of ... interest in the [DSA]

J. Anthony gasps:
> What makes you think any interest ... if the DSA HAD NOT BEEN INCLUDED?

Forget the dollars, you crass materialist!  :-)

Although both are optional, the DSA IS specified and GC isn't--because
a significant contingent of designers and/or reviewers believed DSA was
important, while GC could not mobilize enough support to get in.  (The
same could be said for programmer-controlled GC, i.e., Finalization, vs.
transparent language-controlled GC.)

---------------------------------------------------------------------------
W. Wesley Groleau (Wes)                                Office: 219-429-4923
Hughes Defense Communications (MS 10-40)                 Home: 219-471-7206
Fort Wayne,  IN   46808                  (Unix): wwgrol@pseserv3.fw.hac.com
---------------------------------------------------------------------------

Re: Garbage Collection vs. the DSA

[Jon S Anthony]

In article <9610211437.AA06861@most> "W. Wesley Groleau (Wes)" <wwgrol@PSESERV3.FW.HAC.COM> writes:

> > What makes you think any interest ... if the DSA HAD NOT BEEN INCLUDED?
> 
> Forget the dollars, you crass materialist!  :-)

The point is, if the DSA had not been in the RM, ACT would not have a
DSA and so they would not have those $$ coming from attacted _new_ Ada
users.


> Although both are optional, the DSA IS specified and GC isn't--because
> a significant contingent of designers and/or reviewers believed DSA was
> important, while GC could not mobilize enough support to get in.

That's the "too in-bred team" problem already mentioned.

>  (The same could be said for programmer-controlled GC, i.e.,
> Finalization, vs.  transparent language-controlled GC.)

Finalization is not GC.  Finalization is an othogonal issue, which
happens to be hacked for MM in certain circumstances some programmers.

/Jon
-- 
Jon Anthony
Organon Motives, Inc.
Belmont, MA 02178
617.484.3383
jsa@organon.com

Re: Garbage Collection vs. the DSA

[Jon S Anthony]

In article <dewar.846048897@merv> dewar@merv.cs.nyu.edu (Robert Dewar) writes:

> The point is, if the DSA had not been in the RM, ACT would not have a
> DSA and so they would not have those $$ coming from attacted _new_ Ada
> users.
> 
>   I have no idea what you are talking about, we implement many things, some
>   in the RM (like DSA) others not (like all the DEC extended import export
>   pragmas), and the choice is based on customer demand!

This is not very hard to understand.  It is simply stating that the
chicken/egg dilemma for DSA does not exist and so you can't use DSA
attracting new Ada users as any sort of argument suggesting that that
is why DSA was done.  DSA was done before this - as you well know.  In
fact, aspects of the GPL DSA implementation were done before we
started real work on the IDL=>Ada mapping.  It was even suggested and
considered for a time that the mapping be based on, and thus require,
the DSA.  That position was rejected for a number of reasons.


> That's the "too in-bred team" problem already mentioned.
> 
>   How do you know? This seems to be an argument that goes like this.

Well, you don't know for sure.  That is why it is in scare quotes.


>    I am right, I know I am.
>    Community X seems to disagree with me!
>    Therefore community X must be too inbred.
> 
>   There is another conclusion from these premises!

Given that your argument isn't even valid I don't think this is what
you wanged to say.  What you really want to say is simply that the so
called first premise is false.  Shrug.  Maybe.  Maybe not.  Could be
either way.

/Jon
-- 
Jon Anthony
Organon Motives, Inc.
Belmont, MA 02178
617.484.3383
jsa@organon.com

Re: Garbage Collection vs. the DSA

[Robert Dewar]

The point is, if the DSA had not been in the RM, ACT would not have a
DSA and so they would not have those $$ coming from attacted _new_ Ada
users.

  I have no idea what you are talking about, we implement many things, some
  in the RM (like DSA) others not (like all the DEC extended import export
  pragmas), and the choice is based on customer demand!

That's the "too in-bred team" problem already mentioned.

  How do you know? This seems to be an argument that goes like this.

   I am right, I know I am.
   Community X seems to disagree with me!
   Therefore community X must be too inbred.

  There is another conclusion from these premises!