VistaDB.Net Logo

Windows Workflow a total rewrite in Dot Net 4.0?

by Jason Short 17 February 2009

I read the latest Visual Studio Magazine Editor's Note with some interest.  I agree with a lot of what Patrick says in his piece.  Microsoft is seriously damaging the Dot Net developer community and adoption in the industry with these half baked product releases and abrupt about-faces after shipping.  As I wrote in another blog, I have seen the adoption problems myself.

People outside the developer community are starting to view Dot Net as an unstable platform to build on, and that hurts all of us who make our living on the Dot Net Framework.

Losing Future Developers?

Dan Appleman takes on future developers in the guest opinion part of VSM, Where did all the developers go?  Dan asserts that the changes and pace of change is driving people away from development as a career, not just to other platforms.  The pace of change for a new developer looking to learn the Dot Net framework is pretty scary. 

Can you imagine just picking up the Dot Net framework at 3.5?  I can't imagine the confused look students would have trying to teach the framework to them at this point.  And if we start losing future developers, where are we going to get our new hires of tomorrow?

Windows Workflow a total rewrite

I was a little shocked to read that Windows Workflow in 4.0 is a total rewrite though.  Wow, another major shipping product being dumped by Microsoft? 

From one reader on the site:

Kathleen Dollard's article on the changes in the upcoming version of Workflow [Ask Kathleen, "Windows Workflow Changes Direction," January 2009] reminded me that the same thing happened with SQL Server DTS 2000/2005, as well as with the MapPoint object model and ADO. Now SQL Notification Services was dropped in 2008? Are you kidding? It makes you wonder: How many times do you have to relearn things you used to know how to do?

Well, the Windows Workflow and WCF are both way, way over complicated if you ask me.  But that doesn't mean they will be simplified in 4.0.  In fact I seriously doubt they rewrote it to simplify things.

It actually makes me happy that I never invested more into the Workflow model.  We occasionally get requests to write a Workflow provider for VistaDB.  Doesn't look like that is going to happen anytime soon.  How many other companies are now going to reject WF?

Sort of reminds me of Sync Services (another technology that was totally rewritten after release of the API to developers and pushed into production too early).   We invested several man months of time into Sync and were totally dejected with a CTP of the system was a major change in architecture and API.  We dropped it after that, and in hindsight it turns out to have been the right decision since it is just now appearing to finally be stable.  Well, maybe.  I haven't checked the 4.0 list to see if it is changing again.

Developers are all beta testers now?

Are we as developers all beta testing for Microsoft now?  I seriously am starting to feel that way.  3.5 was shipped way before it was done, as evidenced by the fact that SP1 is 200MB and takes about an hour to install.  That is a Service Pack?  Should have been called 3.6 since it added so much new functionality.  Come on, the Entity Framework is a patch?  I don't think so.

LINQ to SQL anyone?

I was at the developers conference when Microsoft announced that LINQ to SQL was to be treated as a stop gap measure until Entity Framework was shipping.  At that time I thought they were talking a period of weeks, this was over 2 years before the release of EF with SP1.  A lot of developers are really ticked off at the amount of time they spent just to have to reinvest in another technology in the same release of the framework (3.5).

Get it together

Seriously, this is Microsoft!  Not some small startup working out of their garage.  They spend man years per day on development efforts.  If they can't get it straight it makes us all look bad. 

Take a look at the latest VSM magazine.  It is about 1/2 the thickness of the magazine was a few years ago.  Hmm, maybe they are losing audience, or advertisers?  Or maybe they already wrote about all the 3.5 technologies (including a TON of articles on LINQ to SQL that never mentioned it was a stop gap measure).

Dot Net is a great platform, but I seriously think Microsoft has made some bad decisions lately on the development front.  We need stability from the leader, especially during this economic period.  Budgets have shrunk to the point where companies cannot afford to waste time on development that will be scrapped in literally months of time.  Where is the return on investment?

How many times do we have to go through the rewrite process before Microsoft gets it right?  It used to be said that you always waited until Version 3.0 of a Microsoft product before you bought it.  I still know lots of users who refuse to install anything Microsoft related until the first Service Pack is released. 

 

Comments

18 February 2009 #

I don't like this kind of changed too, but if they come with usefull improvement, they are wellcome, you are not obbligated to migrate to the new platform (tell me if i'm wrong) so you don't need to waste your old work, you just only don't get the new features so why not change thing's for better?


18 February 2009 #

"Microsoft is seriously damaging the Dot Net developer community and adoption in the industry with these half baked product releases and abrupt about-faces after shipping."


I entirely agree! Is it all to do with complexity and turnover (of staff?). I don't believe it is so much an architect at work, more a committee that is loosing the plot?


The advent Windows 7 is not going to improve things either. The kernel will feature an early version of ConcRT that is going to evolve over the next few releases. There will be knock on effects and the developer community is going to be hard pressed to keep up or change their ways?


channel9.msdn.com/.../Dave-Probert-In


18 February 2009 #

1) People make mistakes. The developers of WF liked it but it wasn't doing well at all. No one got it. So they are fixing it. The article does say both will be side by side at first. But by your reasoning they should just stick with it because its hard to learn new ways.


2) I don't know if i am upset by less developers. There were too many. I started college in 95 and everyone wanted to be a developer because dot com was soo cool. There are too many developers that have no idea whats going on. Let them die off.


3) Visual Studio Magazine is horrible. I had someone give me a copy last week and I started to read it. Expecting MSDN like quality. ALL BUT ONE article were long winded advertisements for third-party add-ons. ( a waste of my time)


18 February 2009 #

js_vistadb

Less developers over the long run means a lot less jobs because if there is no labor pool for the market they will move elsewhere (hint - overseas - and why Dan makes the point about all the people doing those learner queries are in China and India).


VSM used to be a good magazine. It appears to me that they are in a slow decline. Leading up to VS 2008 they had a ton of great articles all about 2 years before VS 2008 was released. So the magazine lost all its relavency to most people. Cover the problems of today, not tomorrow (especially not unreleased software of tomorrow). From what we have seen VS 2008 adoption is slowly taking off now that SP1 is out. Maybe people were just waiting for the SP before they started using it?


And actually we ARE compelled to migrate to the latest systems, or we lose customers. Since we are developer tool we have to integrate with the latest and greatest or people view you as stale, or outdated. Look at the number of WPF control ads from companies 18+ months prior to anyone being able to use them. They were selling "We are state of the art".


I agree that the APIs should be updated.  But look at the number of serious API changes that are, through Microsofts own notes, broken or non-standard.  Do you see them fixing any of them?  I have been reading the Framework design guidelines book and they point out all sorts of broken or nonstandard APIs that should be fixed, but won't.


 

js_vistadb

18 February 2009 #

If there is a problem, fix it and fix it early. What MS does is not a whole lot different from past, where 1.0 is about proof of concept, 2.0 improves and 3.0 delivers. It's better to make the necessary change early than sticking to a lame model and build on top of it.


WF / WCF / WPF / SILVERLIGHT and so on are basically at 1.5 and 2.0 stages. What us developers need to do is to recognize that and thus pick the right time to step in and pick it up. For me, WCF / WPF are the only things I've used in real work. I keep an eye on WF / SILVERLIGHT but won't touch them until more stable versions come out.


A word on LINQ. Maybe MS should have made it clear, but it's a functional language feature instead of an ORM framework from get go. It can query over Sql server as well as a whole bunch of other list data structures. Many people find it more appealing than TSQL only to mess up the DB performance w/ a bunch of dynamic TSQL LINQ produces. In my team the rule is avoiding Linq over Sql Server as much as possible.


18 February 2009 #

js_vistadb

You know Ben, sometimes I have thought about it.  Microsoft is making me more and more nervous.  Visual Studio is our only target right now, if they seriously screw with the plugin model for VS 2010 it would hurt us badly in terms of lost time.


I wish there was something better than ODBC for another database interface.  That is one interface I seriously cringe over when I look at the specs.


js_vistadb

19 February 2009 #

Perhaps, VistaDB in C++ might be a good investment?


20 February 2009 #

So easy to complain, so hard to do better. For myself I'd be happy making the same amount of mistakes as Microsoft or VistaDB, probably I make more. But I love .Net, WCF, WPF, Silverlight and VistaDB, but what the hell is SQL Server Tin Can Edition 32b with only 250MB download? Good work with your product, it deserves more success. Someone told me lately the Website of VistaDB does not look real professional, just as a hint..


20 February 2009 #

js_vistadb

Oh believe me, it is very hard to do it better.  Microsoft is THE leader though.  All of us model and process our development teams around MS best practices.  They are the example that most companies try to emulate.  Part of my point in the post was that if they can't get it right what hope do we as small teams have?  We base our stuff open their foundation of shipping products.  If we can't count on that foundation to be stable our products will be the same.


Thanks for your feedback on the site.  Anything specific you don't like?  Some people think professional is flash, heavy menus, etc.  Some people think that is amateurish and only like sparse design with CSS (no graphics or tables).  


I am still not happy with the site, but the reasons I am not happy have to do with information linkage from one to another page (navigation).  I personally find sites with lots of menus to navigate offensive because they usually are brittle to the browser showing them.  But sometimes you need menus to compress the number of choices on a single page.


js_vistadb

23 February 2009 #

Concerning the site: the person told me that it looks like someone (excuse me) used some template and has chosen the wrong fonts and color. But I guess because he is a Firebird Fan he was not objective in any way. But actually I think too that the Blog site looks better than the real website, anyway, I love the product.


23 February 2009 #

js_vistadb

OK, thanks.  Sort of funny though.  The blog is a template, but the website isn't. Smile


js_vistadb

2 July 2009 #

trackback

Entity Framework support - Be there first, or getting it right?

Entity Framework support - Be there first, or getting it right?

VistaDB .Net Database Blog

Comments are closed

Powered by BlogEngine.NET 1.5.0.99 and VistaDB 4

 Privacy Policy - Legal Nondisclosure - Refund Policy - Terms of Use

Log in