[dba-Tech] Dearly Departed Databases (R.I.P.)

Shamil Salakhetdinov shamil at users.mns.ru
Mon Jul 18 07:36:48 CDT 2005


<<<
> Just think about it. Suppose you could do everything you can in Access,
and
> then compile the result into a stand-alone product that doesn't require
the
> run-time or anything else, and compiles to say 2mb per app.
>>>
Arthur,

I suppose MS Access is so big in its and related file sizes because it has a
lot of features.
Only MSACCESS.EXE (2003) size is 6+MB.
How one can shrink that into 2MB per app?

MS Access is a powertoy.
It was always targeted like that by MS.

My daughter who is 19 years old and who is on her 2nd grade in highschool
does do great things with it and MS Word and MS Excel during her Informatics
courses study - she does in minutes and hours what ten years ago was almost
mpossible to do using months of professional programming - and all looks
like a game...

I did spend a lot of time working with MS Access.
I know it well.
It's getting depreciated for professional development because with .NET one
can do similar things with almost comparable ease(speed) and many other
things, which can't be done in MS Access because of its programming lanuage
limitations.

Yes, using MS Access together with MS Office a lot of office work can be
automated and this is where MS Access really shines...
When you need to do such office automation on more professional level - use
.NET Tools for MS Office.

I did program a way to much on VBA - you know all that my investigations on
the VBA edge - I should have better spend most of this time on something
else - VBA limitations are naturally designed by MS and MS don't have plans
to remove them as far as I can guess....

Last year I did program three separate MS Excel add-ins using VBA(in Excel).
When I tried to combine them into one I did get met with VBA limitations
once again - together they just stopped to work, were GPFing
unexpectedly(because of a lot of code?) etc. - and I did quit this my idea.
Later on, last autumn, I ported all that three add-ins in four days into
VB.NET + .NET Tools for MS Offcie - they worked like a charm, used common
library code etc.etc....

MS Access is getting out for professional development, especially for
off-the-shelf solutions - it's becoming more and more like a
powertool/powertoy to make everyday work by powerusers - and the "army" of
these powerusers is becoming bigger and bigger wih every year...
Professional MS Access programmers have to switch to something more
professional (pun intended:)) ....

Of course I do use MS Access to make some quick tests and as poweruser to
process and keep all kinds of information. And I do plan to continue doing
that way this my internal office work. But when I will decide to make
something off-the-shelf from that internal office automation I will use C#
or C++ - they have no limitations and they are WYDIWYG-programming tools as
opposed to WYDYMFTI-programming VBA-based tools...

What is great with VB/VBA is that they allow very high level of "edit and
continue" programming, which doesn't yet exist in VB.NET and C# (they
promise to have it in VS.NET 2005). I was originally assembler programmer,
then C, Pascal, C++ and then VBA, VB6  - all the "real" (not VB and VBA)
programming languages force you to "think/design first then program" and I
did do design/program a lot in the past on the paper then typed all that on
punch cards or into computer program sources files (there was computer time
limitation here in the past), then debugged all that - and voila' - the
program is ready...

With VB6/VBA this my previous programming habits/style were completely
ruined with incremental rather chaotic and hectic "edit and continue"
programming style. Yes, this is good to create "quick&dirty" office
automation small solutions but they are not that good(bad in fact) to work
on something more serious...

Nowadays as far as I see a new programming style/paradigm arises (exists in
fact as far as I can guess but I have still to meet somebody working this
way) - and this is a combination of "think/design" and "edit&continue"
styles based on all the last 10-15 years experience in OOP&D and the modern
development tools (e.g. VS.NET) - to become a high level programmer/software
architect in this area experience and only real life development/programming
experience is the answer. The more real software you create using this
technique the more you get how right and useful it's - so after 15 years of
VB/VBA "madness" we again return to real programming using real programming
languages and real-life programming experience... That looks like a natural
turn of evolution spiral.... (BTW,  IMO to know C/C++ and even ATL/COM isn't
"all you need" to become a high level programmer - I have seen/see programs
written using these languages and development tools, which I wouldn't call
professionally written. Yes they work and are used in business applications
but the overhead to support them is too high to be true.... )

I mean: VBA is getting out of mainstream development programming languages
because of its natural limitations,  which are becoming so obvious now and
which make one trying to make the living from programming uneasy to
compete...

I still like VBA/VB especially because they are the only(?) tools for quick
prototyping using their unique "edit&continue" feature but this is planned
to be inherited by VS.NET 2005....

Shamil

P.S. WYDIWYG - What You Design Is What You Get - whatever design you create
based on the most modern programming design concepts you will no doubt be
able to implement it using C# and C++.
WYDYMFTI - What You Design You May Fail To Implement because of programming
language limitations, you'll spend a lot of time with very low end
result.... Be prepared, don't get desperate, you can keep trying but you
will fail again...

----- Original Message ----- 
From: "Arthur Fuller" <artful at rogers.com>
To: "'Discussion of Hardware and Software issues'"
<dba-tech at databaseadvisors.com>
Sent: Monday, July 18, 2005 4:18 AM
Subject: RE: [dba-Tech] Dearly Departed Databases (R.I.P.)


> Thanks for reminding me of these babies, Shamil. I too worked a little in
> DataFlex, but never dbVista. I once met and interviewed Robert Carr,
> architect of Framework -- perhaps my fave program ever written, though his
> choice of language syntax was asinine... but that's neither here nor
there.
> Robert impressed me as one of the smartest people I have ever met. After
> Framework he ventured into areas with promise, I guess, but they didn't
pan
> out. The handwriting recognition stuff, and so on.
>
> I think I'll try to reconnect with him and see what he's up to.
>
> Access is like a truly gifted bastard child in the MS family. Nobody wants
> to admit its progeny. Nobody wants to admit that given any 89 problems,
you
> can solve them more quickly in Access than in any other MS development
tool.
> I feel profoundly sorry for the Access development team. They try to
please
> us -- and for their efforts I am extremely grateful -- but they get shat
> upon from above because Access was intended to be a stupid little toy for
> the great unwashed and uneducated. So there are these development folks,
> writing a powerhouse tool, that the Bosses not only don't want to hear
about
> but regard as a threat to the more profitable revenue streams.
>
> A handful of the Access team should quit MS and launch an Access-compiler
> project. This is very parallel to how Clipper came about. Ashton-Tate
> refused to release a compiler, and Brian Russell had the vision to quit
the
> Framework development team and architect Clipper. Rich McConnell played a
> big part but it was Brian's vision that got it from drawing board to
> delivered product. Not to say others didn't help, but it was Brian and
Rich
> that made it happen.
>
> We need some players like that in the Access world. None of us knows
enough
> about the internals to pull it off. It takes members of the development
team
> with the guts to quit and the ambitions to release a killer product.
>
> Just think about it. Suppose you could do everything you can in Access,
and
> then compile the result into a stand-alone product that doesn't require
the
> run-time or anything else, and compiles to say 2mb per app.
>
> Arthur
>
> -----Original Message-----
> From: dba-tech-bounces at databaseadvisors.com
> [mailto:dba-tech-bounces at databaseadvisors.com] On Behalf Of Shamil
> Salakhetdinov
> Sent: July 17, 2005 8:48 AM
> To: Discussion of Hardware and Software issues
> Subject: Re: [dba-Tech] Dearly Departed Databases (R.I.P.)
>
> DataFlex - I did work with it in 1993-1994 - it was great, OOP, 4GL etc. -
> it was widely used epecially in Autralia and Sweden, UK etc. - is getting
> nowehere now?
>
> dbVista (Raima Data Manager) - great too, C/C++ centric, cross-platform(PC
> OSes)  mainly used  in embedded systems now....
>
> Ashton Tate Framework - great tool - anscestor of all nowadays Office
> suits...
>
> MS Access :) - it's getting depreciated now as a development tool(?) - not
a
> mainstream development tool like my collegue working at MS (:) ) says - do
> you agree? :)))
>
> Shamil
>
>
> _______________________________________________
> dba-Tech mailing list
> dba-Tech at databaseadvisors.com
> http://databaseadvisors.com/mailman/listinfo/dba-tech
> Website: http://www.databaseadvisors.com




More information about the dba-Tech mailing list