Hi, Eric.  Boy am I happy to meet you.  :) I am now using my off-line
reader so here goes.

ES>LR> I am taking a look at my next move as far as converting my apps.
ES>LR> It seems that for the time being I am stuck with converting it
ES>LR> to Windows due to borland's lack of support for OS2 - for Pascal
ES>LR> guys like me.

ES>When OS/2 v2 was introduced, I adopted it for all in-house machines, and was
ES>able to convince my clients that it was the choice for them. At the time, mo
ES>of my software was written in Borland Pascal, with lots and lots if inline A
ES> Though it all ran just fine under OS/2 in a VDM, I was faced with the probl
ES>of producing native OS/2 software with no Pascal compiler.

   I don't understand.  Why did you feel *compelled* to go to native
   OS/2 apps(with your apps) if they ran well under OS/2's VDM?


ES>In June, 1993, after personally speaking with Zak Urlocker, who was then Pas
ES>Product Manager at Borland, I was convinced that there will *never* be a Pas
ES>compiler for OS/2 from Borland.

   I don't want to hear this, really.  :)  God god, whatever did Zak say
   to you in mid 1993 that convinced you so thoroughly that you decided
   to move to C++ for OS/2?

   Are you saying you converted all your existing DOS based apps to C++
   and OS/2 in one fell swoop?  Gulp.

   My apps are too large to do that. I have too many man-years in their
   creation for me to convert them to any other language in any cost
   effective manner.   It would be ruinously expensive for me to do
   THAT.  I might as well go out of the software development
   business than convert - now.  Maybe later - when I can hire some
   bright, young, energetic bucks.  :)

   But then I believe my DOS based apps is so unique, took so many
   man-years to research and implement, and is virtually without any
   significant competitors *right now*, that I believe my users won't be
   overly concerned for a while - or have much choice - I hope.  :)

   At least this is what I am banking on - for a while anyway - maybe
   another year or so - *after* I start selling it - and making money.

   Users can still run DOS based apps on their OS/2, Windows,
   WindowsNT.  They might not like a DOS apps, but I beleive the
   benefits of my apps will cause them to use my apps, since there is
   no other product even close to mine - yet.

   I estimate that even if a company threw 5 pros at my product and tons
   of money - for one year, they couldn't come out with a better product
   for a year or so.  And if they did, they would find I have a better
   one then theirs - ready to sell while they were developing.

   I believe I am at least three years ahead of any possible
   potential competitors out there.  I designed, planned and
   implemented my plan to remain ahead of any competitors.

   Still, as you point out, if Borland is out of the Pascal compiler
   game for OS/2, then they are going to PO, big time, about 3 million
   Borland Pascal customers worldwide - 40 percent of which Borland
   claims are pros. These are Borland figures - not mine.


 ES>The choice was clear:  Dump Borland, invest
 ES>Watcom and IBM C++ compilers, and spend a year polishing my C++
 ES> abilities to professional level.

    I guess that makes sense to you for your situation; it makes
    little sense to me with a huge conversion problem from Pascal to
    C++, and OS/2.

   Right now there is a big market for both DOS and Windows as far as I
   can determine - much bigger than for OS/2 - and all I, and others
   like me, have to do is essentially *recompile* my current DOS (Pascal
   based) apps using BP 7.0, make a few adjustments for Windows PM,
   etc.,  and my product is ready for the DPMI, 386/486/Pentium
   world and:

   1. 16-bit Windows consummers.

   2. 32-bit Windows consummers (WINS32???) .

   3. 32-bit WindowsNT consummers.

   All of which I believe, and I hope I am not wrong, my customers can
   even run on OS/2.

   This is a LOT less effort than conversion of a large DOS apps *after*
   a huge learning curve for just C++ and OS/2 alone - not to mention
   the huge learning curve for C++ for the other operating systems.

   So I am not sure how you justified going to C++ and OS/2 alone from
   Pascal if you have any apps of any size and mucho man-years in them.

   But I do want to hear your justifications anyway.  :)  I try to keep
   an open mind when it comes to making money.  

   Eventually, I will have to do what you have done, abandon
   Borland and it's Pascal, for C++ & OS/2.  I am just betting I won't
   have to do it for another year or so.  What do you think?

 ES>LR> Has   anyone been where I am and done this - converted their
 ES>LR> procedure Pascal apps to TV and OOPs?

ES>Wouldn't touch it.  I would recommend forgetting the notion of
"con- verting
ES>anything.

   Yes, well we seem to agree on that point - mostly.  At some point I
   will have to convert if you are right and Borland is going to abandon
   it's 3,000,000 Pascal users as far as OS/2 is concerned.

   But then I believe this because I hear, and believe, that OS/2 is
   great.  But I am one of those who heard, years ago, how HOT UNIX,
   and/or, it's brethen would be big time when DOS users could run
   their DOS apps under UNIX-like operating systems.  I am sure glad I
   didn't hold my breath on that prognostication.

   In case you hadn't noticed, I am not a big fan of C++ or OOPS.  I
   have *studied* it.

   It (OOPS) sounded great when it came out.  However, as I have found
   out, it is a lot more complicated than Pascal or 'C' and the
   professional consulting demand is still not there - and if it was
   there, it wouldn't be met.

   Not many pros have caught up - or are even more than familar with C++
   - and many don't want to use C++ from what I hear of my fellow
   consulting brethen.

   The one big feature of C++ is it's ease in extensibility where you
   want to sell code without giving out source.  So how many large
   companies are in this place or want to go to this place.  They have
   the source code, use it internally - so where is the *great need* for
   extensibility.

   I hear all the C++ compilers are slow and memory hogs in both
   compilation and execution states.

   Who cares?  I care and so do many pros.  It is not the first time
   that DRAM prices became what everyone thought was insignificant - and
   found out they were very wrong when their budgets went south.
>>> Continued to next message


--- GEcho 1.00