The reason being is that I've left the devastatingly dull world of corporate web development and got myself a job in the games industry. And apart from the direct benefits of that move, I want to say how refreshing it is to not have to care about Microsoft's torrent of API's and frameworks any more.

Of course the game engine is dependent on APIs, but on mature ones like win32 and DirectX, not on the stuff that changes every few years because people at Microsoft got bored and decided they finally cracked the secret of how to make development *easy* at last, only to abandon that again when they realise they over-engineered some monstrosity like WF or Entity Framework, or ASP.NET Webforms. WF and EF got pretty-much rewritten, Webforms got sidelined by MVC, WPF is sidelined (as far as I can tell) by WinRT/Metro apps if it wasn't already by Silverlight. And WCF should have been taken out back and shot a long time ago.

The amount of developer hours that are wasted each year by people learning new Microsoft APIs and frameworks must be staggering. It's amusing that Channel9 put up video from PDC2003 - the one where the Longhorn technologies were revealed before all being dropped or demoted to installable .NET frameworks. Will the stuff revealed at Build be any different? Only time will tell. But I'm just so glad I don't have to care any more.

Why don't you go post in your games forums, rather than coming here and telling everyone why they are a bunch of losers, people tend to dislike that.

Move along now...

6 minutes ago,vesuvius wrote

@rhm: This post reminds me of when Dr Evil in Austin Powers shouts "1 Billion dollars" after time traveling while not realizing that the world has moved on

Why don't you go post in your games forums, rather than coming here and telling everyone why they are a bunch of losers, people tend to dislike that.

Move along now...

That was a mature response.

For one thing I wasn't calling anyone a loser. And secondly I thought people might be interested in the impression Build gave to someone who used to a huge .NET fan and look forward to ever shiny new toy that rolled out of devdiv. But of course you're not interested in anyone else's perspective.

What is it you do again?

5 hours ago,spivonious wrote

@rhm: I have an old book on DirectX 7 that I never really got into. Do you think it's worth going through or has enough changed where it wouldn't apply to DirectX9/10/11?

dx8 is where they added vertex and pixel shader support, and dx10 was a significant rework of the API (so much that many games still support separate dx9 and dx10 code paths in their engines), so I'd say no.

25 minutes ago,DeathBy​VisualStudio wrote

So when Metro/WinRT is a flop what do they spin up next? WinFormsXAML/WinRT? That way they don't have to say WinRT was a complete waste of time. They can also say that we benefit from it as we're not throwing away all of our API knowledge and it still XAML. We can also "port" all of our lovely Meto apps to the new, more powerful framework. Why does that sounds so familiar?

Yes, the problem is that MS doesn't really have a strategic vision; their 'strategy' is to see what others do, wait a few years, then react. This is why you get all these shifts and about-turns.

I enjoy learning new things, but if it only has a shelf-life of about two or three years (and two years is about as long as it takes to become a real expert in enterprise technologies) then I don't see the point in wasting my time. 

You don't have to be cutting edge to use the Microsoft platform; be grateful that Microsoft if moving at all and giving you the opportunity of using alternatives if you feel you need to.

Herbie

6 hours ago,DCMonkey wrote

*snip*

dx8 is where they added vertex and pixel shader support, and dx10 was a significant rework of the API (so much that many games still support separate dx9 and dx10 code paths in their engines), so I'd say no.

Thanks. I guess I'll look for an updated book.

16 hours ago,rhm wrote

For one thing I wasn't calling anyone a loser.

Yes you did.

"The amount of developer hours that are wasted each year by people learning new Microsoft APIs and frameworks must be staggering"

14 minutes ago,Harlequin wrote

*snip*

Yes you did.

"The amount of developer hours that are wasted each year by people learning new Microsoft APIs and frameworks must be staggering"

I take that as "sympathy" or "disgust" with Microsoft and not labeling anyone as a "loser".

12 hours ago,Dr Herbie wrote

I've worked in .NET exclusively since the first release (2002).  I got off the 'API treadmill' at.NET 2; we still use Winforms and ADO.NET Typed Datasets.  

You don't have to be cutting edge to use the Microsoft platform; be grateful that Microsoft if moving at all and giving you the opportunity of using alternativesif you feel you need to.

Herbie

I'm glad someone got the point of my post (and managed to present a decent argument against). I can't help wondering if the defensiveness of certain others belies their true feelings about what they do.

Anyway, I see where you're coming from - people could just stick to one generation of MS tools. I mean there are still MFC developers, VB6 developers even. But then why stick with Microsoft stuff at all if you're going to get off the treadmill that way?

14 minutes ago,rhm wrote

*snip*

Anyway, I see where you're coming from - people could just stick to one generation of MS tools. I mean there are still MFC developers, VB6 developers even. But then why stick with Microsoft stuff at all if you're going to get off the treadmill that way?

*snip*

How's that getting off the treadmill? That's simply staying on the old treadmill and not buying a new one.

-Josh

6 minutes ago,JoshRoss wrote

@vesuvius: It's more like Microsoft has this really super long treadmill-- like the ones at airports-- and instead of getting off, you just decide to stop walking forwards. Kind of like what Herbie said.

-Josh

Well, treadmills move backward and the airport "speedwalks" move forward.

On a treadmill, if you stop walking, you fly off of it backward. You have to run just stay in one place. On a speedwalk, if you stop walking you still move forward gently until you have to step off.

8 minutes ago,Frank Hileman wrote

Consider the "metro style" rules for applications. Is it worth paying lots of attention to the metro profile? It may disappear soon. We have no way to know.

Isn't that always the case?  Technologies progress.  People used Prototype, now they use CoffeeScript.  What will it be next week?  People used tables, then only divs/spans, now grids and flexbox, what will it be next week? 

This is why the core of a good developer should be fundamentals.  If you ask me to RoR, I can pick it up in a day or two.  If you want me to ASP.NET MVC, the same thing.  Was up and running with WPF in a few days coming from scratch.  HTML/JS -- literally felt comfortable in a few hours.  Haskell, give me a week to be comfortable.  Learned QT (from w/ C++) in a couple of days.

Sure I'm not a language lawyer in that time, but I'm "early" productive in that time, and learning.  But I've built the foundation and I supplement it with these languages and frameworks.

If you feel like you're on a treadmill, it's because you have no foundation.

There is a lot of API competition on the Java platform which gives you a lot of choice (perhaps too much choice in some cases). It also means that you new ideas come along much faster.

People seem to think somehow that because it has the Microsoft seal it is the best in the World, this seldom is the case. Take the Entity Framework that is programming terror all wrapped up in one. Juxtapose this to Linq to SQL or nHibernate and you just accept that sometimes people there prefer to keep their heads in the sand. If you used enterprise Liibrary before the fourth version or workflow foundation before .NET 4 or Patterns and Practices in Windows Forms there is a pretty wide ranging suite of failures that noone seems ever to be able to take the blame for.

That is why architects usually opt for stability and maturity in APi's, rather than the latest thing

Zipping a file is a very easy thing to do with the plethora of free zip libraries. I don't know why there is no simple way to achieve this with this library, it could be simpler.

Off topic: if all you need is to zip one file, the ZipPackage class is probably overkill; I would try a simple DefalteStream first.

But the difference is in the non-Microsoft world there is no "cathedral model", where a bunch of "high priests" come out every year to command some new form of development that everyone must follow.

In the Java world evolution happens in a distributed ("bazaar") style, where the developer community organically moves to various technologies as time progresses, and these technologies rarely comes from the original sponsor of Java (Sun). It works like natural selection.

That's why we got stuff like the Spring Framework which pretty much eclipses the "cathedral"-competitor EJB. It's more even more true in Python and Ruby communities which have nothing resembling a centralized commercial sponsor.

The development world is getting more and more interesting as there just simply is no one-way to do things anymore, having people concerned in whatever camp they belong (all for some of us) means better software for us all round

WPF is sidelined (as far as I can tell) by WinRT/Metro apps if it wasn't already by Silverlight.

I don't see WPF as being in any jeopardy:

http://joshsmithonwpf.wordpress.com/2011/09/18/winrt-is-your-friend/

3 hours ago, Bass wrote

The API treadmill definitely happens for non-Microsoft stuff as well. One example I think of is Java XML marshaling libraries, and web frameworks. Java has many of both, in various stages of "hotness" and approaches.

But the difference is in the non-Microsoft world there is no "cathedral model", where a bunch of "high priests" come out every year to command some new form of development that everyone must follow.

In the Java world evolution happens in a distributed ("bazaar") style, where the developer community organically moves to various technologies as time progresses, and these technologies rarely comes from the original sponsor of Java (Sun). It works like natural selection.

That's why we got stuff like the Spring Framework which pretty much eclipses the "cathedral"-competitor EJB. It's more even more true in Python and Ruby communities which have nothing resembling a centralized commercial sponsor.

What is the Microsoft world?  Is that Windows?  Or is it Microsoft dev tools?  Because I have a ton of different dev tools I can use for Windows, many that don't originate from Microsoft. 

If you're referring to MS dev tools then it is a tautology.  Every dev community has its own high preists.  Go to Rails and they are even more hardcore about "only one right to do things".  That's the whole notion of opinionated frameworks, "We know better, so listen to us, or go home". 

But the MS high preists have typically been quite good, compared to the open source world.  Even the great high preist of open source dev tools, gcc, just recently because decent. 

But the MS dev community doesn't blindly follow the preists.  When the preists came out and said, Visual Source Safe then later TFS, for SCC everyone said that is blasphemy.  And everyone used SVN, Mercurial, Git, etc...

The world is never as nice and neat as some would like you to believe.

I don't really think there is much a Microsoft ecoystem outside of what Microsoft produces. Microsoft has never that I recall "blessed" something from the community and integrated it with .NET - other than jQuery.  

If you look at the Java Persistence Framework, a official part of Java - but it's basically Hibernate. Would Microsoft ever do something similar with NHibernate? Would they adopt NHibernate as the "official" .NET ORM?

No way, they rather spend millions developing (and redeveloping) an incompatible and widely derided alternative, Entity Framework. And have their army of evangelists push it over NHibernate.

Do they profit from Entity Framework? No, they do not. Is it a competitor to Microsoft's business? On the contrary. So why go out of your way to compete with your own community? You compete with competitors not allies. Allies that are enriching your ecosystem with powerful software at no cost to yourself. Microsoft is throwing cash at a problem that's already solved. In order to gain absolutely nothing in return, except confusing their developers and perhaps spawning threads like this.

There is only ASP.NET and it's various flavors for web stuff. I tried to promote this competitor to ASP.NET called Manos de Mano, which IMO has a much simpler and more powerful design. But have yet to see any real world use of it. Compare to Java and stuff like Lift, Play, Spring MVC. None of which came from Sun but enjoy wide usage. Will a .NET web framework ever be successful if it isn't written by Microsoft?

PS: I know a lot of companies that use SourceSafe and STILL use it. It was a successful product.

http://www.codinghorror.com/blog/2007/06/supporting-open-source-projects-in-the-microsoft-ecosystem.html

2 hours ago, Bass wrote

@KDawg:

I don't really think there is much a Microsoft ecoystem outside of what Microsoft produces. Microsoft has never that I recall "blessed" something from the community and integrated it with .NET - other than jQuery. 

NuGet