To: "Terry Leeper" <tleeper@microsoft.com>
Subject: Re: MASM references
From: Jonathan Kirwan <jkirwan@easystreet.com>
Date: Tue, 08 Jan 2002 18:52:19 -0800

On Tue, 8 Jan 2002 15:28:17 -0800, you wrote:

>Hi Jonathan,
>
>MSDN subscribers already have access to C 1.52.  

I'm well aware of that.  The question I'm raising for you to consider
is whether or not C 1.52C might become available without having to pay
that kind of money for it (and so much more, of course.)


>You are correct.  There are many factors to such a proposal.  One thing
>that we often face here is cost of support.  Even when we declare that
>something is "not supported", we know internally that if we allow
>something on our website or other means of distribution that we would
>need to support it or risk a ton of wrath.  This often goes against the
>grain of what we're trying to accomplish.

Indeed.

I see the same effects you do, as well.  In projects, I might say,
"It'll be _at least_ 5 weeks to get that job done.  But almost
certainly much more, because of these two research subjects
(unknowns.)"

And what will be heard by others will only be just the "5 weeks" part
of it.  Four weeks later, people will be asking me if I'm going to
make the 5-week deadline!

Saying, "not supported" can be like a slow death!  I do understand
some of the worry.

But surely Microsoft has already set up the machinery to deal with
problems like this.  Your products are shipped with many systems via
Dell, Compaq, and a host of other manufacturers.  Explicitly, support
is supposed to come from these businesses.  But I'm sure Microsoft
already gets calls anyway and has a policy on dealing with such
questions.

Is it possible that Microsoft might consider arranging things so that
someone else is in the middle on this?  Perhaps a small distribution
fee might be charged by an outside corporation for a CD, for example,
which included these products.  And they could maintain a web page, as
well.  But other than that, no support.

Well, I don't mean to pretend to understand how best to do this.  I'm
sure there are more creative people on this kind of subject than I am.
Could you talk about the possibility of designing something to meet
Microsoft's needs with others there?


>As far as the need for 1.52 goes, I guess I don't understand why you
>would prefer this over the current compiler.  If it is cost, then I
>would refer you to our Standard Edition, which runs for $99 (you also
>get MASM with that and its documentation, BTW).  The only real caveat is
>that the compiler does not optimize for retail builds, meaning it is not
>meant for real production.  However, this does serve well for academic
>and teaching purposes.  It also supports inline assembly.

Because you've focused on only one facet of what I've mentioned
before, you've misunderstood the fuller scope of my interest.  1.52 is
a 16-bit compiler, as you are more than a little bit aware.  In
embedded systems, it is often a requirement that the system be
operated without the presence of Windows or any other 32-bit O/S.  (In
fact, one system I've worked on recently is still using an 8088.)

I'm not just interested in teaching, though that is certainly a big
part of my interest.  I'm also an embedded programmer, as well.  In
that light, I still use QB 4.5, PDS 7.1, VB-DOS 1.0, VC 1.52C, ML, and
other products as well.  Finally, I also enjoy teaching a lot of very
young kids who cannot even handle the $99 you mentioned.  I'd very
much like to be able to tell them where they could get something of
high quality to learn on.

Because I believe that Microsoft no longer derives profit from the
sale of these products specifically, I'm hoping Microsoft might see
its way clear to making this a reality.

I don't expect it to be necessarily easy, because as you suggest,
"there are issues" to be considered.  I'm just asking for some
thoughtful consideration.  If the results are "No," then so be it.
But at least if some thought was applied, there are some good reasons
for it.

Jon