Database Connectivity

Cliff Wells clifford.wells at comcast.net
Mon Oct 11 16:05:50 EDT 2004


On Mon, 2004-10-11 at 11:17 -0700, R Baumann wrote:

<snip stuff I don't disagree with>

> 3.  When Access 2000 was released, MS included a version of MS-SQL Server
> called the MSDE(Microsoft Database Engine) that was a deliberately
> watered-down version of MS-SQL Server 7.  By watered down, I mean that MSDE
> could handle up to 5 concurrent PROCESSES before a governor kicked in to
> deliberately slow things down.  This was to prevent MSDE being used in areas
> where an industrial strength DB Server needs to be used.  A marketing issue?
> You bet.  But, MSDE has been know to work very reliably in small user LANS,
> and workgroups of up to about 25 users with a medium to somewhat heavy
> workload.  And, if the workload is fairly light up into the hundreds.  If
> the workload is light, OR, ESPECIALLY if the developer knows what they're
> doing, the governor won't kick in that often. Processes can be made to be
> efficient.  Unless you're a huge corporation with hundreds or thousands of
> users, this is more than good enough to handle most small companies, and if
> those companies grow larger, the upgrade path to full blown SQL Server is
> just a simple step to attach the tables from MSDE to SQL Server.

But the question remains: why use a "watered down" version of a
database, or pay for an "industrial-strength" version of it when there
are free alternatives that actually surpass both versions?

> <GASP!>  M$ sells stuff for money!  Why not?  I like to eat, you like to
> eat.  Don't like Microsoft?, sounds like a personal problem to me. 

Frankly your entire rant sounds like a personal problem.

>  I don't
> like everything they do either, but to pay my salary, the company I work for
> has to make money, not live on indignant or righteous attitudes.  I use
> whatever works best for the particular customer I have to work with. 

Reasonable attitude.  Despite being a Linux afficianado, I still
recommend Windows for most user's desktops, simply because there isn't
enough payoff to justify the learning curve for most people were they to
switch.  Linux has major advantages over Windows in many areas, but the
desktop currently isn't one of them.

>  If MS
> has a product that does what I need it to, I'll use it.  If Python is a
> better way to do things, I'll use it, if QBasic, if Delphi, if...

You'll use QBasic?  And where will you get it from?  I point this out
simply to highlight your apparent blindness to Microsoft's abandonment
of actively used tools (which you cite later as a reason to prefer MS
over open source projects).

> 4.  Don't like SQL Server? use ODBC.  I've done Access apps using Btrieve,
> MySql, Interbase/Firebird and others as the backend data engine, and Access
> as the GUI front end.  My development time in Access is 1/3 to 1/2 the time,
> and with fewer errors.

1/3 to 1/2 the time of what and with fewer errors than what?  And when
you've specified that, for what complexity of the previously
unmentioned?  Access, VB, and, in fact, most "RAD" development tools
work best with simple applications.  Their friendliness and simplicity
quite often become a liability for more complex applications.  It's been
a few years since I was forced to use Access, but as I recall, it wasn't
even possible to design forms (dialogs) with dynamic layout using those
tools.  Quite frankly I doubt there is much to be gained in development
time with using Access over a combination of Python, PostgreSQL and
wxPython/Tk or even a web interface.  But I don't doubt the lack of
flexibility one would incur with that choice.  

> 5.  JET is no longer being actively enhanced.  It IS being maintained.
> Eventually it will fade away, but the MSDE will still be there.  Access is
> now up to version 2003 just in case anyone is interested.  MSDE is up to
> version 2000 or perhaps more by now.  It's a free download from the MS
> website.  There ARE some licensing issues, so pay attention to them if you
> download.  

JET will fade away.  Gasp!  Yet another MS technology that won't be
around?

> Not everything in the world has to be or should be "free" or
> "open source".

Agreed.  But when it's already there, and it's superior to the
commercial version, why wouldn't you use it?  Bigotry?

> Don't get me wrong.  I love the idea of Open Source.  I make use of it quite
> often.  My problem with it is when the primary developer gets tired of the
> project, which often happens after a year or two, it generally dies because
> no-one is interested in taking over, or the project stagnates. MS never
> said they were in business to produce free or open source software, and at
> least they don't let their products stagnate.  It takes money to build and
> maintain a Word, or an Excel, or an Access, and sometimes I absolutely need
> to use a product I know is going to be around a while, and be supported.

PostgreSQL (started around 1986 as a research project, moved to its
current open-source status around 1995 [also when SQL support was
added]) has certainly been around at least as long as the equivalent
Microsoft products (SQL Server was released around 1989 on OS/2, 1992 on
Windows), so I fail to see your point here.  Also, holding Microsoft up
as an example of a company that supports its own products for extended
periods of time is probably ill-advised.  The original version of SQL
Server doesn't exist in any incarnation, nor does Microsoft support the
OS it ran on (OS/2), despite having hyped it as the next major PC
platform.  In fact, the current version of SQL Server is a complete
rewrite that actually qualifies as a completely unrelated product
despite having the same name.  You yourself have mentioned at least two
MS technologies that either no longer exist or are being phased out
(QBasic and JET). Now, unlike you, I'm going to be fair and acknowledge
that this is simply the nature of software.  Microsoft, and its open
source competitors, do well to occasionally abandon existing platforms
as new technologies, architectures, and plain old hindsight dictate.  

> 6.  Finally, MS is positioning Access to be a GUI front-end to any database,
> and the hooks are already there.  Look for a .NET version in the future.

Access is terrific for creating *Windows* GUIs for databases (whatever
database is chosen).  Of course if you need (or simply want) cross-
platform, very few Microsoft technologies are going to help you.  .NET
is being held up as such a solution but quite frankly, given Microsoft's
bait-and-switch tactics of the past, I'll maintain a wait-and-see
attitude until proven otherwise.

> Sorry about the long rant, but so much mis-information has been passed along
> in this thread and in  past threads, mainly due to so much MS hatred,
> deserved or not, that  I felt I had to speak up.  

Well, certainly opinions of MS are just that.  But all else aside,
Microsoft made my job miserable for enough years with shifting API's,
buggy software, inflexible/limited/incompatible tools that I feel quite
justified in my contempt for them.  And that's before considering their
company from any ethical standpoint.

> I get the impression that
> most of the info comes from people parroting other people, and not really
> knowing about the products that they are dis'ing.  

And I get the impression that people who defend Microsoft simply haven't
tried anything better.  I suppose we could continue to view each other's
positions through Monet-tinted glasses but I doubt that really gets us
anywhere.  Try foaming at the mouth just a little less and instead focus
on forwarding self-consistent arguments and perhaps you'll be a bit more
convincing.

Regards,
Cliff

-- 
Cliff Wells <clifford.wells at comcast.net>




More information about the Python-list mailing list