Don't you have other Pharma sector companies other than this particular
one? I can safely say that as long as you are cool with relocating
wherever your job takes you, proficiency in Java Technologies will get
you into one of the pharma companies.

-cheers,
Manish

That's a commercial question, not a technical one.

There's no significant technical benefit of C# over Java; C# is simply a
copy of Java as close as Microsoft can get without violating Sun's various
contracts and intellectual property protections. C# does benefit a little
from coming later and learning from a few of Java's mistakes, but not in
any significant way - most of the things which are wrong with Java are
identical in C#.

If your employer is a Microsoft house, sooner or later they'll move to C#.
If your employer is not a Microsoft house, or wishes to keep their
long-term strategic options open, they won't.

If you are looking at big companies, my bet is they cannot afford to be
strategically dependent on Microsoft, and so will stay with Java. This
isn't simply a choice between Microsoft and Sun, since Microsoft are the
only main suppliers of the C# environment, whereas Java can be
second-sourced from IBM and others. It is this relative openness which,
for large companies at least, will continue to give Java its relative edge
over C#.

My experience is in biotech, so it may be slightly different in the big
pharma, but here goes anyway.

We typically used C++ for individual number crunching tools that are
computationally intensive and operate on a particular kind of data and
Java for database-related tools that are more integrative and operate
across many kinds of data.

So, for a chemistry example, if you're doing some sort of 3D protein
folding or something, you'd typically use C++ over Java.  If you're
writing the code that tracks the chemical inventory, and which
compounds had been tested against which targets, you'd use Java.

My impression is that roughly 1990-2000, a tremendous amount of push
was made for the C++ type of programs.  Number crunching, replacing
experiments with computations, etc. So the idea would be that you'd get
rid of labs and just have big computers. Algorithmic focus.  These
things would be labeled computational chemistry, etc.  And I think that
to the extent they still exist today, these are still using C++.

Eventually, people realized that wasn't working nearly as well as had
been hoped, and the new approach is to use computers to tie existing
knowlege together.  So computers become more integrative of existing
data, and their role is to help make the lab processes more efficient,
and tie their data together, rather than to replace labs.  This is more
IT or informatics.  Data focus.  And Java has been great in this latter
field.  Depending on a given company, they might use Java with various
frameworks or go for C# and .NET.

Michael

Probably nobody will know the answer...

(In a working newsgroup, a lot of the communal knowledge at any one time has
been learned from reading previous posts -- so a genuinely new question might
stump the whole community ;-)

   -- chris

Occasionally, and specific bits. I wrote a course on Tomcat earlier this
year, which you can find here:

http://www.weft.co.uk/library/tomcat/

and am working on a course on the Location API for J2ME.

My Java/database/XML tutorial from 2000 is here
http://www.weft.co.uk/library/inet2000/

...but it contains a series of absolute howlers in slides 86 - 89 of day
two - the example given just happened to work with the version of Xerces
available in 2000, but is not compatible with the DOM API and you should
not follow it!

There are petroglyphs with clear evidence of having been
created using a LayoutManager.

Larry Coon
University of California

Find a concrete real world problem that needs solving. Choose a toolkit on
which to base your work - personally I tend to use things I've rolled
myself, simply because I've been doing this longer than most people and
when I started there weren't good toolkits out there so I had to build my
own; and now I'm so practised with my own toolkit that there's little
benefit in trying something new. But if I was starting over I would
probably start with Struts:

http://struts.apache.org/

I'd then find an open source project built with that toolkit and pull it
apart looking for patterns to learn from.

Personally I very much like Tomcat. As to databases MySQL is getting better
but it's still not as robust as Postgres, which I would recommend in
preference.