Python is a gem ,another successful day ....

G

Graeme Matthew

Hi all I previously made a post (
Python is a gem, you need to keep pushing it ....)

about how impressed I was with python and how im suign it in a proposed
commercial app.

Got a call on Friday a company with a very succesfull product in Visual
Basic (Front End) SQL (Backend) wanted to move from Microsoft products due
to all the problems with MDAC (ADO etc), XP networking issues .... and
concerns with microsofts control ///

Well after three hours looks like Python (and linux and mysql) may have won
again, explained perl (semantic nightmare but a great language) , ruby
(great language but lack of libraries) etc, Vb's weaknesses and the need for
vendor neutrality anyway, as things move ill put some updates on here ...
and tell you ho it goes, ....

Hey, im trying very hard to promote it :)

Cheers
 
S

Steve Holden

Graeme:

Congratulations!

I'd be interested in knowing some of the arguments you used to win over VB,
as I've been asked to put a couple of articles together suggesting why VB
users might consider switching to Python.

regards
 
R

Ray Smith

Steve Holden said:
I'd be interested in knowing some of the arguments you used to win over VB,
as I've been asked to put a couple of articles together suggesting why VB
users might consider switching to Python.

The first obvious reason is that VB 6 (which most VB users currently use) will
be unsupported by 31-Mar-2005.
http://support.microsoft.com/default.aspx?scid=fh;en-us;LifeDevToolFam

The change from VB 6 to VB.Net is a huge change and in some cases a complete
re-write.
For some features missing in .Net see:
http://www.mvps.org/vb/index2.html?rants/vfred.htm


Regards,

Ray Smith
 
G

Graeme Matthew

Hi Ray, Steve

The company that is interested in finding an alternative has provided a
software solution for 25 years now. It has been through a plethora of
microsoft conversions from dos, /BASIC through to WinXP /VB 6.0. As some
companies are small their version has an Access Database (dont ask me why
:)) and the medium to large sized organisations have SQL Server backends.

I am also doing a job for a partner company of theirs and they originally
wanted a VB6 front end onto a SQL Server backend. Luckily the head of the
partner company is an open minded I.T professional (going back to assembler
/ unix programming) so convincing him was not too difficult. The press has
been really good regarding open source and linux so I think more people are
realising that there are better products that exist that dont bear the
Microsoft name.

Ray is correct in what he says, VB6 has a short lifespan and the migration
to VB.NET is not that simple.

This company is now having untold problems with WinXP as long running
network connections that are idle are automatically disconnected, their
application for small users open a permanent connection to Access and this
disconnecting "feature" for the sake of network performance is causing
corruptions.

The issues with MDAC (ADO) is also causing support nightmares, i.e
downgrading upgrading , but im not to up to date with all the issues
regarding this. Python solves this as modules are text files and through
sys.path one can control the search path

DLL versions etc have been a nightmare since day one ....

My first motivation was to get them away from Desktop dev and move to
intranet based applications. This they realise and can see the benefits of
moving to this new ("Yet old") alternative. So the question then was What
language / technology to use ?

My first issue to highlight was that if we went with Microsoft then the
costs would be greater not only in licensing but also server specs, and
from fellow colleagues doing .NET development they all say its hungry on
system resources. I also emphasized that Microsoft will be "coercing" them
to keep upgrading and that we should remain vendor neutral and get out of an
environment where choice is limited ....

Having programmed in Perl on Linux, I remembered how frustrating it was to
come back to your code or others , 6 months later and not understand it :)-)
not that bad) I think perl is great, so I shifted to python which I know
offers way more than VB 6, even VB.NET especially if multiple programmers
are to work on a project. I also emphasized the importance of a language
been controlled and progressing through programmer enhancement proposals and
computer science principles rather than a marketing panel, I mean how many
times do they change concepts and badge it under something new in VB ? I
also love ruby but I still feel that there are too few libraries, datbase
interfaces (we need this for possible integration.)

Visual Basic has also been a nightmare for memory management and its 1MB
stack space, so recursion is out, in a recent project I have had to write
recursive and non recursive quicksorts and binary searches so I can switch
on large array sets.

I also feel that microsoft's focus on marketing concepts without fixing or
enhancing the core language, even in VB6 now arrays dont have pop so one has
to hold the entire array in memory even when your down to the last few
elements. There are ways to remove items but a lot of work. I feel that its
things like this rather than Data Environments, Class Wizards and
containment instead of inheritence is what has made it into the language
that it is.

Not sure if I am making sense, one could write forever :)

Cheers

Graeme
 
S

Stephen Ferg

I'd be interested in knowing some of the arguments you used to win over VB,
as I've been asked to put a couple of articles together suggesting why VB
users might consider switching to Python.

I really hope you're able to do those articles comparing Python and
VB.

I'm trying to promote Python within the Federal government, and my own
agency, the Bureau of Labor Statistics. Right now, the chief
alternative at BLS to Python is Java, but not too long ago it would
have been VB and PowerBuilder.

Folks are moving away from PowerBuilder mostly because of concern
about the vendor's longterm support/viability, but VB actually may
rise again, as there is a small but noticeable groundswell of
discontent with Java's clumsiness and verbosity.

Bottom line: If we can develop materials making a convincing case for
Python compared to VB, there are situations where such material might
be of real help.

-- Steve Ferg
 
N

Nick Vargish

I'm trying to promote Python within the Federal government, and my own
agency, the Bureau of Labor Statistics. Right now, the chief
alternative at BLS to Python is Java, but not too long ago it would
have been VB and PowerBuilder.

Sounds like the situation here at the <Federal Agency Where I Work>
where I'm trying to promote Python as a viable alternative to Java or
Visual Basic.

I'm currently seeing the most resistance in these areas:

* Who makes Python? A lot of the culture is still mired in the idea
that a corporate entity is required for a product to be fully
supported, or viable in the long term. I like to remind them how
well Sun's stewardship has "served" the Java community, but I need
more than negative counterexamples.

* Who provides training? I mean no disrespect to my coworkers, and
present company is of course excluded, but a lot of the federal
workforce has no incentive to learn on their own. Unless a
training course is provided, with certifications and free tote
bags, there is very little motivation for people to learn new
technology.

* How do we deploy Python? Until we upgrade all the systems here to
recent versions of Redhat Linux (joke!), installing a Python
environment is an extra step. I can point out that finding an
appropriate JRE is a worse nightmare, but as above, I need more
than negative counterexamples.

We're currently in something of a holding pattern while a bit of
reorganization takes place in the IT departments, but things will
probably be changing rapidly (as rapidly as any change occurs in a
government institution) once the organizational situation is
stabilized.

There was some talk of forming a SIG for Python in the US Government,
did that ever go anywhere?

Nick
 
S

Skip Montanaro

Nick> * Who makes Python? A lot of the culture is still mired in the
Nick> idea that a corporate entity is required for a product to be
Nick> fully supported, or viable in the long term. I like to remind
Nick> them how well Sun's stewardship has "served" the Java
Nick> community, but I need more than negative counterexamples.

Is Apache used by your organization?

Nick> * Who provides training? I mean no disrespect to my coworkers,
Nick> and present company is of course excluded, but a lot of the
Nick> federal workforce has no incentive to learn on their
Nick> own. Unless a training course is provided, with certifications
Nick> and free tote bags, there is very little motivation for people
Nick> to learn new technology.

There are people (like Mark Lutz and Wesley Chun) and larger organizations
(perhaps ActiveState?) who do Python training.

Nick> * How do we deploy Python? Until we upgrade all the systems
Nick> here to recent versions of Redhat Linux (joke!), installing a
Nick> Python environment is an extra step. I can point out that
Nick> finding an appropriate JRE is a worse nightmare, but as above,
Nick> I need more than negative counterexamples.

It should be a small extra step. If you can't use Sean Reifschneider's RPMs
directly, you can probably build RPMs for your version of RH Linux using his
spec files, then install that across your organization. Similarly, for
Windows users the installer is easily run.

Skip
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,769
Messages
2,569,577
Members
45,054
Latest member
LucyCarper

Latest Threads

Top