My little poll received 32 votes. Here's how they divided up:

D1    = 13
D2    = 11
D1/D2 =  8

Okay, so this may not be a very scientific poll, the sample is small and
non-random. In particular I wonder if D1 users may be more motivated to
respond because they feel marginalised and want to stand up and be
counted, whereas pampered D2 users don't give a damn and just sit back !
However, it does seem to reveal that D1 development accounts for about
half of all Delphi devlopment at the moment.
--
Kevin Urben

--
Bob Cousins, Software Engineer.
http://www.demon.co.uk/sirius-{*word*104}netics/

I started selling a retail DOS program in 1992. It is a very specialized
application, but it is directed towards a real general user market (many
customers are very novice computer users). Until maybe a year ago, I was
convinced I was better off sticking to DOS rather than migrating to
Windows. I had never lost a sale because it was a DOS program, but I know
I would have lost sales if it required Windows.

Typical support call:

"What kind of computer do you have?"

   "It's an IBM compatable."

"Yes, but is it a Pentium, 486, 386."

   "I don't know."

"How long ago did you buy the computer"

    "Oh, a long time ago. Maybe 5 years ago."

"Hmm... Do you know how much RAM memory you have?"

    "There should be plenty of memory. It has a 20 MB hard disk."  :-)

And it turns out it is a 5 year old 286. It wasn't that long ago that
someone specified they needed low density 5 1/4" disks for their XT.

My point is that it is going to be years until the need for 16-bit Windows
applications goes away. This type of user doesn't want to go out and spend
the money to buy a new Windows 95 computer, and doesn't have the technical
expertise to upgrade to Windows 95 even if his current computer is big
enough to handle it.

My hesitation in choosing Delphi 1.0 as a tool for 16-bit Windows
development was having read that Borland feels they will be able to drop
support within a year or so, which could leave me{*word*154}. I am less
concerned about upgrades to the current Delphi 1.0 than SUPPORT. What if
some weird interaction problem pops up with running Delphi 1.0 on Windows
'97 and Borland doesn't support the product anymore?

There are certainly features which could be added to Delphi 1.0. Migration
of some of the new 2.0 controls back to 1.0 would be very nice. Support
for 16-bit OCX / ActiveX (like VB4).

To me, a hot topic at the moment is the ability to produce 16-bit
applications with the look and feel of Windows 95. I am just starting with
Delphi 1.0, and it strikes me 1.0 really already successfully does this to
a large extent (and that is why I chose the product).

HINT TO BORLAND: I am currently using Delphi 1.0 from a Delphi 1.0 box. I
haven't discovered what the advantage of upgrading to the Delphi 1.0 in a
Delphi 2.0 box might be, so I probably won't upgrade (it could be there
are bug fixes or something in the latest 1.0 in the 2.0 box, but I haven't
been able to locate anything that explains this on the Borland Web site,
etc.)

However, if Delphi 1.0 was upgraded with a few new features like more
controls and 16-bit OCX, that would be a significant reason for someone
like me to go out and purchase the upgrade.

Any new features which would help a 16-bit application attain a Windows 95
look and feel would be of great interest.

rdbbolo...@aol.com

(1)  The Delphi2 compiler has some really nice implementation-features that
would be equally useful... and equally apropos... on a 16-bit platform.  I
would like to see a 16-bit implementation of it.

(2)  Otherwise, I think that both compilers should be a permanent fixture in
the product line and that both of them should be distributed =together= in at
least the Developer edition of the product.

(3)  As much as practicable, the 16- and 32-bit versions of the language
should be identical, with clearly defined support for conditional-compilation
that would allow a single source-file to be used with both compilers.  (This
is, of course, here now.  It just needs to be identified as a requirement.)

(4)  The 16-bit world is not going away.  Neither is the 32-bit.  What we
developers need is a single tool that allows us to readily support both camps
without buying the whole product twice.  (And for that matter, you don't need
to be printing two sets of manuals either.)  In other words, what you've done
with the BC++ product is a fairly good strategy.

(5)  A reasonable, justifiable price is all we ask.  To me, the feature set
and the value-for-price-paid is ultimately more important than the price
itself.  I know I'm asking for a fairly large unit of software here and I'm
prepared to pay a fair price for it.  (I'm in to around $450 or so... other
opinions may vary sharply!)

/mr/

I agree with all you said. Reading your statement, I infer that D1 and
D2 can NOT be used in a parallel way, i.e. programming in one of
those, and compile in both?

Chris

*******************************************
   Christoph Juengling, Kassel, Germany
   Systemanalyse - Beratung - Software
   http://members.aol.com/ChrisJueng
------------------------------------------
You may take a look at the homepage of the
German Access User Group on

   http://members.aol.com/ChrisJueng/daug
*******************************************