dotNET Ideas
dotNET Ideas
Enough said!
http://www.techweb.com/showPressRelease.jhtml?articleID=X333718
Microsoft platforms (ASP.NET, ASP): 43.6% Java platforms (J2EE, JSP, WebLogic, WebSphere, Tomcat): 12.2% PHP: 5.2% ColdFusion: 2.7% Perl: 2.3% Python (Zope): 0.1%
Microsoft IIS: 53.7% Apache: 22.7% Other Web servers: 12.8% Netscape Enterprise: 10.8%
I think I made the right decision in my career path!
There has been an online petition (http://classicvb.org/) recently to try to
convince Microsoft to continue the development of VB6. When I first heard this,
I thought it was a joke. I have been using VB.NET and C# for about 3 years now
and I would seriously quit my job before working in VB6 again.
Some background about myself, so you won't think I am a Java or C++ guy. In my
career, I have worked in many languages. Here is a recap and some commentary.
BASIC (TRS-80 & Commodore) - 3 years - OK, this was before my career but that's
when I learned to code.
Ada (original) - 6 years - The original Ada was a lot like VB6 in syntax and
support for OOD concepts. Where VB6 is component-based, Ada is "package"-based.
Neither is object-oriented. I was fairly fond of the syntax but the tool
support around 1990 wasn't very good. It was about what you would expect for a
language designed by and used only by the U.S. government.
APL - 1 year - If you are a mathematician or statistian by trade, APL might
appeal to you. Otherwise, YUCK!
C - 3 years - I have always loved C's syntax. Crisp, clear and terse. Also,
it's as fast as anything you'll find.
C++ - 2 years - I never really liked C++ all that much. I think it was because
I was doing it in VC++ with Win32 and MFC. It just never felt like a coherent language.
If everything in your code is a macro, then you aren't really using the language
anymore.
VB (versions 3-6) - 8 years - By far, the majority of my gainful employment has
been in VB. I never cared for the syntax much because I am not fond of typing
and it is hard to find your logic in amongst all of those keywords. It's the
same problem I have with COBOL.
VB.NET - 3 years - It's just similar enough to VB6 to be confusing, but not
different enough to be worth learning, IMHO. If you are switching to .NET
anyway, learn C#. It will be easier to learn without all the VB baggage.
C# - This is where I found my true love. C# has the syntax of C, the power of
C++ and RAD environment of VB. I'm home.
Reading the FAQ on the petition web site, it seems like they are complaining
about several different things.
1. The end of "mainstream support". Obviously, Microsoft should be making bug
fixes and security fixes to VB6 as long as there is a significant install
base. The patches should be free and easy to get. As for free phone support,
you've had that for a long time now. It has to end sometime. 3 years after it
is superceeded is plenty long enough. Many appliances cost about the same as a
VB6 license and they only come with 60 or 90 DAY warranties.
2. "Microsoft made my data obselete." This is an obvious attempt to get the
attention of business decision-makers. No "data" has been made obselete. Some
CODE has. There is an argument on the site that code and data are the same
thing. Sorry, I must say bullhockey to that!
3. Existing VB6 applications may not run on future versions of Microsoft's
operating systems. I haven't heard Microsoft announce the end of COM support in
Longhorn, so I don't think that is going to happen in the next 5 years or so.
Further out than that, you're on your own. Just like you always were.
4. VB6 code won't be able to access all of the platform features introduced in
later versions of Windows. Of course not! It wasn't designed to run on that
platform. If you want new features, you have to upgrade. No surprise there.
What this all boils down to is that these folks want Microsoft to treat VB6 as
though it were an entirely different product from VB.NET, not just its predecessor. If
Microsoft felt they could make money doing that, I am sure they would, but they
don't. The vast majority of pre-VB.NET apps have already been written. Now they
are just being maintained. Therefore, the license revenues will drop sharply
over the next few years regardless of how much money is poured into it.
I spent years defending VB against the VC++ bigots who considered it a "toy"
language. There are a few things VB6 can't do that VC++ can. How often do most
of us do them? Most of the VC++ code I have ever seen was written in VC++
because the guy writting it knew VC++ and didn't want to use VB.
VB.NET is finally a full citizen among MS languages
and that's great. I can do anything in VB.NET that I can do in C#. I just like
C# better.
If you are concerned about supporting existing applications, by all means, push
MS to supply bug fixes, but don't expect full support. It just isn't practical.
Get involved in the online communities and solve your own problems. Don't
expect Microsoft to solve them for you.
If you are concerned about new language and platform features, you shouldn't be
using VB6. Sorry to be blunt but there it is. VB6 isn't dead yet, but it is
retired.
If you really feel that Microsoft is missing the boat on this, here is my
suggestion. Get together with some like-minded friends and come up with a
business plan. Contact some venture-capitalists and propose to buy the product,
including support commitements, from Microsoft. (I'm sure Bill G. would be
thrilled.) You could continue selling the product, developing new features and
providing ongoing support for its customer base. The problem you will run into
is that your business will NEVER show a profit. That's why MS isn't doing it
itself.
I downloaded Firefox today out of curiosity. It seems pretty straight-forward and stable so far. I like the downloadable themes (mostly because I don't like the default one).
What have other people thought of it?
Here is a Dice ad I saw today:
Why would anyone try to recruit programmers with an ad composed of badly formatted C code with a GOTO statement in it? The GOTO isn't even syntactically correct.
In case you missed it, EntLib was released on 1/28. Here is a link: http://www.microsoft.com/downloads/details.aspx?FamilyId=0325B97A-9534-4349-8038-D56B38EC394C&displaylang=en
You will have to register, but that's life.
According to sources close to the project, Enterprise Library should be made publicly available on January 28th.
Avanade and MS have teamed up to reengineer the Patterns and Practices group's Application Blocks. The new release is called “Enterprise Library”. Avanade gets to use them exclusively for 90 days starting 10/27/04. After that, they will be made available under the same terms as the App Blocks have been.
Since I work for Avanade, I have had a chance to see and load them up. I haven't gotten through all of it yet, but it looks very nice. It will be good to have all of these parts made consistent and interoperable.
For more info:
http://www.microsoft.com/resources/practices/comingsoon.mspx
For fellow news junkies, I offer a couple of links:
“Watching Microsoft Like a Hawk“
http://www.wmlah.com/
“Watching Google Like a Hawk”
http://www.watchinggooglelikeahawk.com/
I always take vendor commissioned studies with a huge grain of salt:
http://www.eweek.com/article2/0,1759,1645550,00.asp?kc=EWRSS03119TX1K0000594
but this one is interesting for two main reasons:
1. It was done by a leading JAVA company (http://www.theserverside.com/).
2. It matches my own experience very well.
I have been thinking a lot lately about the virtues of being a generalist (as I
am: MCSD, MCDBA, MCSE, ad nauseum) or a specialist.
I define a specialist as someone who concentrates on a specific technology and
becomes extremely-proficient at it to the expense of their other past, present and
potential skillsets. These are the people who know every registry setting and
command line in Windows or every odd behavior in Exchange. My usual retort
when someone makes the "be a specialist" argument is the old saying "If you only
know how to use a hammer, everyting looks like a nail." The advantage to being
a specialist is that, while the specialty is new and in demand, you make more
money this way.
I define a generalist as someone with a broad knowledge of various technologies
and good proficiency in a few selected ones. I am a generalist on the Microsoft
platform. When I was just a baby programmer, I worked on UNIX. As an Architect,
I see generalization as the natural course. An Architect's job is to look at a
solution and identify the correct technologies for the job. How can you do that
without the broad knowledge that make you a generalist? The knock I hear
against generalists is that we are "a jack-of-all-trades, master of none." I
have never heard that said about an experienced generalist after a project, only
about generalists in general (pun unintentional). Also, when someone is
starting out and hasn't gained enough experience with anything to be a
specialist, they sometimes get dumped into the generalist basket, making us
look bad.
I have always resisted over-specialization because nothing lasts forever. Any
specialty you chose could seem very quaint in a few years. For example, I refer
you to all of the COBOL, LISP, ADA and Lotus Notes specialists out there. (No
flames please. These are just examples.) Of course, there is still work to be
had in these specialties but there are also lots of people that can do the work.
One product that I specialized in back in 2000 was BizTalk Server. I loved the
way BizTalk handled all of the mundane tasks associated with EAI and left the
interesting stuff to me. I got really deep into its inner workings and even did
some writing about it. Unfortunately, it didn't really take off the way I
thought it would. Some of that was the economy, but some of it was that I don't
think the product was promoted and positioned very well by Microsoft. Most of
the work I got in BizTalk was EDI or HIPAA oriented.
When BizTalk 2004 came out, I had a dual reaction to it. On the one hand, it's
an awsome product! On the other, most of what I knew about BizTalk 2000/2002
went right in the toilet. I am finishing up my first BizTalk 2004 project now
and I still love the product. I am about half way through rebuilding my skillset
but it is frustrating because it is SO different. Also, the fact that the HIPAA
Accelerator for 2004 is still not available has screwed up more than one
potential project for me.
Was I right to have specialized in BizTalk or did I waste my time? I am torn. I
am now seeing a big demand for BizTalk 2004 skills but I feel like I had to
start over with everyone else because of the new design. Was all of that work on
the ealier versions was wasted? As a generalist, I'll take that experience, tuck
it onto my bookself and wait for it to come in handy.
Now I am hearing from recruiters that want me to specialize in BizTalk 2004. Do
I bet my career on BizTalk and take the risk that MS will faulter again or do I
miss out on some really cool oppurtunities trying to maintain my generalist
"street creds"?
Anyone with an opinion is welcome to reply. I would like to hear your thoughts.
I guess it is a rite of passage to find this particular problem in VS.NET.
We were getting compile errors that didn't make any sense in our VB.NET project. The errors flagged were for symbols that weren't in our code on line numbers that didn't exist. (I don't think you should ever have 167075 lines in a single source file, BTW, but that's what the error message said.) And it wasn't just me. It was everyone on the team. We originally just rebuilt everything and it went away. Then we had to begin restarting VS.NET to make it go away.
This morning we stopped being able to compile all together. So, we did what any good Gold Partner would do. We called Microsoft and yelled at them! Actually, that's what I wanted to do but they were too helpful for it to seem right. Within about a half an hour they had gotten us in touch with the development team and they had figured it out. It is a bug in VB.NET that only affects very large applications like ours.
I had no idea we were over 100,000 lines of code! Yikes!
The Knowledge Base article is 822690 if you are interested.
Long story, short...We installed the Hot Fix that MS sent to us and we are on the road again.
Thanks, MS.
I have noticed a tendancy by Microsoft (whom I love dearly) to make software development more of a commodity. Here are a couple of examples:
“Express“ developer products:
http://www.eweek.com/article2/0,1759,1635813,00.asp?kc=EWRSS03119TX1K0000594
Software Factories:
http://www.eweek.com/article2/0,1759,1633764,00.asp
Soon, we will all work in "Express Software Factories". These will be places where high-school kids that know how to write crappy VB code will work on an assembly line to create bloated, slow, crap-ware for $7 an hour. I picture this conversation at a drive through in 10 years:
Intercom: Thanks for stopping at "Objects are Us". Can I help you?
Driver: Yes. I'd like a rock-solid line-of-business application, with an innovative user-interface and industry-wide interoperability.
Intercom: Would you like any security with that? It costs extra.
Driver: Of course, I need security! What kind of question is that?
Intercom: Most people don't bother paying for it since it never works anyway...
You may have noticed that I haven't been blogging much since I got back from TechEd. Well, I am back down to 3 (active) projects now, so I thought I would reawaken my blogging spirit.
Here's a brain teaser for you.
I have two sets of .NET components that use transactions. They are properly flagged and inherit from ServicedComponent. Component 1 uses Component 2 to carry out part of its transaction, so the transaction context has to flow across the call.
I have this compiled and working fine on one box with Components 1 and 2 in separate COM+ applications.
I need to access Component 2 remotely from Component 1 without losing the transaction context. I assume that it would work as expected if I exported the COM+ package for Component 2 and installed it on the server with Component 1. This is just normal COM+ Services stuff. I am looking for a easier way to get this done.
I want to avoid the overhead and configuration hassles of COM+ packages (i.e. DCOM). Also, I don't like that component 1 is then stuck using COM interop to get to another .NET component.
Everytime I try to use Remoting or Web Services, I lose the transaction context.
I have considered coding directly to the DTC API, but that defeats my goal of finding something “easier“ (read as “more elegant“).
WS-Transactions doesn't seem to be an option yet. I haven't found an implementation I can use.
Anyone have any ideas?
Yes, I know that Web Services can do this in Indigo, but that doesn't help me in the real world.
For those of us still trying to get our heads around Service-Oriented A-Thingamabobbies.
http://msdn.microsoft.com/architecture/soa/default.aspx
I have noticed something interesting in the last month or so. I am getting a lot of calls lately looking for BizTalk 2004 talent. As I said in my previous post, I am fully booked and two of my current projects are BizTalk-related.
Over the last couple of years, I have told people that I haven't seen much demand for BizTalk skills. While I have always liked the product, it just didn't seem to be catching on. Now that is suddenly changing. Most of the interest has been in HIPAA or general EDI work.
BizTalk 2004 is certainly a big improvement over 2002, but I don't see why the sudden upswing in interest. It's doesn't seem like Microsoft has unleashed one of their giant marketing pushes for it.
If anyone out there in blog-land is looking for BizTalk work or interested in training up for BizTalk, by all means, drop me a line.
Man! I come back from TechEd and I'm suddenly buried. 2 projects that had been sleeping (no client urgency) are now hot and now I am on 2 NEW projects. Yikes!
If there are any .NET Developers in or near Omaha that are looking for contract work, please post!
OK, enough whining...
MS posted an article on rearchitecting ASP.NET 1.1 sites into 2.0. Looks like a good article.
http://msdn.microsoft.com/asp.net/default.aspx?pull=/library/en-us/dnaspp/html/rearchitectv20.asp
I got my evaluation forms back from InfoTec today. It's always interesting to hear what people have to say in a situation like that. Expectations vary a lot at InfoTec and I never know what to expect.
When I presented at InfoTec in 1999, I gave a talk on Data Warehousing. I talked about multidimensional data, star schemas and so on. What I ran into was a room full of end-users who couldn't care less about the how or the why. All they wanted to talk about were the really slick front-end analysis tools business analysts could use to analyze the data once it was in a cube. That wasn't what my talk was about nor something I knew all that much about. I ended up showing them a little bit about using Pivot tables in Excel and generally bombed.
This year's talk on SharePoint went a lot better. I think I got the right crowd for what I was presenting. I think it was too basic for some of the people in the room, but the talk was titled “Introduction to SharePoint” so only one person said the content was “light”. To be fair, I ran about 15-20 minutes shorter than planned, but I had run over by 30-60 minutes in practice so I thought I had it about right. I could have blabbed on all day.
One comment I got was a hit to my presentaion style. Anyone who has seen me present before knows that I typically walk around in front of the screen pointing at things and gesturing more than I need to. (This can be fun to watch since I weight 430 pounds!) It's how I work off the nervous energy that comes with presenting. This time I sat behind a table with my laptop and didn't get up at all. One person wrote “STAND UP STEVE!” This wasn't planned on my part. The reason for it was that I sprained my knee just before the presentation while I was at plaNet's booth in the exhibition hall. I could barely stand, let alone walk around. (Maybe I'm not cut out to be an 'exhibition'ist?)
Aside from a sprained knee, I think it went pretty well.
A friend of mine, Phil Canarsky, pointed out something to me today that I wasn't aware of.
If you catch an exception and then rethrow it like this:
try
{}
catch (Exception ex)
{
// Do Cleanup
throw ex;
}
You lose the original stack trace. You can avoid this by just doing a “throw;“ without specifying the exception.
Here is an article: http://dotnetguy.techieswithcats.com/archives/004118.shtml
Thank, Phil!
I am making good progress on the biology project for my wife. Here is a shot of the UI.
The light and dark bands are potential intron and exon sites.