Saturday, February 07, 2009

.NET is dead

Windows - Losing market share fast to Mac and Linux and Microsoft is losing market share
Office - Google eyes the enterprise market
Internet Explorer - Internet Explorer Loses Market Share

With the continuing decline in market share of all things Microsoft it's becoming obvious that .NET is not long for this world. With current trends extended, OS X will reach parity with Windows in 2022. .NET will have long since lost relevance before then.

So, if you're working on legacy applications in .NET, that's fine, but you'd have to be crazy to recommend it for a new project since you know that the future pool of developers will be greatly contracting. Microsoft will of course find ways to increase pricing to make up for the shortfall in sales, so the cost of all those applications built on top of the proprietary .NET platform will become more and more expensive.

If you are a .NET developer, now would be a good time to start looking at a cross platform language with open source options like java and ColdFusion. At least then you might be able to find a job in a few years with the growth in cross platform solutions.

Ok, this isn't a serious post, even though it wouldn't bother me if it was true. I just wanted to point out how silly all the "ColdFusion is legacy" arguments are when they are uttered by people who know nothing about it. It really irritates me when I read posts by people that say they would think less of a company for proposing a ColdFusion solution for anything new. That's just stupid. CF has been in constant development since it was owned by Allaire. It has and continues to grow greatly in terms of capability. Knocking out a technology just because of something "they" said robs you of the chance to get a good, maintainable solution at a decent price.

I guess it's time to pull off a few snazzy Flex projects using CF as a service layer.

23 comments:

  1. LOL LOL
    This pop up on a filer of mine looking for ".net is dead".

    Keep the faith.

    Brother in Arms.

    ReplyDelete
  2. .NET, like other MS technologies such as SQL Server, appeal to Microsoft shops, and little else. As long as there are Microsoft shops out there (small interactive firms, etc) these technologies will hang on.

    ReplyDelete
  3. I am NOT dead, you seal-clubbin' wingnut. Next time I see you, you're gettin the tire iron!

    ReplyDelete
  4. @ .NET 3.5 SP1

    Shut up Marc. And I mean that in the nicest way.

    ReplyDelete
  5. hahaha. That's a funny title. It's about time someone said it. :)

    ReplyDelete
  6. Have recent events not made an impression? I walked twice as far as i did yesterday which means by 2022 i should be able to walk to australia in an afternoon.

    ReplyDelete
  7. changing .net to cf? ha ha. it is like changing corvette to civic.

    ReplyDelete
  8. @anonymous #2 - Not a very good analogy there. How about you walk twice as far as you did yesterday for 3 or 4 years. Then we'll talk.

    @anonymous #3 - More like changing from a corvette to a little nash rambler. You're in trouble if we ever figure out how to get out of second gear.

    Consider this: ColdFusion was around when Microsoft was pushing the use of idc/htx. Which one is still going strong? ;). I'm starting to think that CF is the product that WON'T die. It just keeps getting better.

    ReplyDelete
  9. of course it won't die. adobe is baking cobol #2. in 20 yers the zen of programming will state: you can program in any language but not in cobol and coldfusion.
    and that famous number - 500,000 developers, ha ha.

    ReplyDelete
  10. @anonymous 4 - You're never going to run windows on a mainframe. Consider this: despite the 20+ year forecast of the demise of COBOL, the amount of computer involvement in business processing continues to increase. At one point, COBOL we expected to be replaced by C. That obviously didn't happen and COBOL has proved to be a lot more resilient than the "experts" predicted.

    There are a few things that keep COBOL relevant. One is that the standards are very stable, but do evolve when necessary. Recognition of XML was added to the standard in 2006 and Object Oriented COBOL is currently in the draft phase. Another is that a lot of this code has been around for a LONG time. It's very reliable and tends to work consistently. There is a lot of value in that sort of stability in business systems.

    So, if Adobe is rolling COBOL 2, that would be great!

    ReplyDelete
  11. ASP.net loses yet another direct competition against ColdFusion.

    http://www.adrocknaphobia.com/post.cfm/coldfusion-8-wins-web-development-award-in-india

    ReplyDelete
  12. The thing that troubles me most about ColdFusion's future is that it's always playing "catch-up" with all the other languages. I mean until version 8, debugging was a 3rd party add-on! (Sure you had it in 'ye olde CF' but why would anyone let that feature drop off?) .NET and Java had stable ORM frameworks when CF was still learning what object inheritance was. And they're building all their tools on Eclipse now? Even the Java people have abandoned the Eclipse project!

    I'm not saying that CF is dead. With the sorry state of PHP libraries, CF is probably your best bet for loose-architecture rapidly-built apps, despite the price tag. But when it's time to get serious, CF is late to the party and bringing the same old stuff that .NET folks have been using for years. If you're going to drop 2 grand on a software platform, why would you spend it on 2-3 year old technology?

    ReplyDelete
  13. Sorry to burst your bubble...

    I work with CF and .net on a daily basis so I have a very good understanding of the strengths and weaknesses of each. Believe me, CF is not even close to being a serious competitor.

    ReplyDelete
  14. @Dan and Anonymous6 - CF runs on my Solaris box. Pbbbbbt.

    ReplyDelete
  15. @Mike: So do PHP, RoR, Mono, and Glassfish. Being cross-platform isn't a special feature anymore. It's standard practice. (And yes I know that Mono is not caught up to MS.NET yet, but it's still very powerful and moving forward way faster than CF is.)

    ReplyDelete
  16. @Dan: I don't see .NET in that list of cross platform technologies listed there ;)

    Ok, seriously folks, I was just having a little fun with that original post. It was tongue-in-cheek from top to bottom. I intended it primarily as a friendly poke in the eye for all those CF developers that are constantly worried that Adobe/Macromedia/Allaire is going to go out of business or pull support for the language and feel they have to write a "CF is Dead" post on their blogs.

    ReplyDelete
  17. @Mike: Mono is .NET. Most of it, anyway.

    ReplyDelete
  18. Mono is actually looking better each day. Its amazing what they've done.
    As for ColdFusion it is, or was, a very productive webdev framework. In fact, it was the 1st language I ever felt productive in. At some point, when it took the java route (rel. 5.0?), its performance started to lag. No, I don't have numbers. Just a perceived sluggishness compared to previous releases.

    The advantage of .NET/Mono is that it supports both thick and thin apps.

    ReplyDelete
  19. I took a look at mono briefly after the comment here about it. Previously I had only heard about it occasionally. Based on my complete, in-depth review of it (ok, I looked at the website for about an hour), it does look pretty amazing. If I had nice 36 hour days and mono took off, I'd be on that really fast.

    As far as cf performance, the platform has gotten noticeably faster with each release. I think a lot of the sluggishness you might be seeing now has to be attributed to the developer of the app you're looking at. Any competent developer can grind a web server into the ground. A lot of times, when I move into doing some maintenance on an existing program, there is a lot of what I call "programming plaque". Stuff that's just hanging around all the time hoping to be used eventually. It just gums up the works.

    One of the main reasons I tend to tread lightly when it comes to a lot of the cf frameworks is because of this programming plaque. Some of them just do so much on every call to the server that there just has to a performance hit, although some are really good at caching and keeping that overhead to a minimum. I just like to be careful there.

    ReplyDelete
  20. Funny, I did a google search for Calvin pissing on a Ford, Dodge, Chevy, and Toyota and this came up. :)

    ReplyDelete
  21. The long term problem I see with CF is that it runs on Java. While CF runs on Linux, OS X, as well as Windows, it still feels like a Windows product. Shops are either going open source or they are going Microsoft. Adobe is putting CF in a tight spot by keeping it on Java. CF will never succeed in the open source space, which is where Java is migrating to. CF's target audience is the shops that run Windows. It seems to me that for CF to survive, it will need to migrate to the .net platform for long term survival.

    ReplyDelete
  22. How exactly will OSX reach "parity" with windows when globally windows probably has a 95%+ of all desktops? PLEASE, I'm a java developer and hate .NET and even I think that's a joke.

    ReplyDelete
  23. @ last anonymous: You're really a .NET developer aren't you. You're trying to make java developers look bad because of their poor reading comprehension. Nice try, but I know a M$ shill when I see one.

    ReplyDelete