Zugg Software :: View topic

Magician Joined: 13 Oct 2000 Posts: 463

One very important thing that I don't see addressed yet in ZML (or at least, not in the demo apps or tutorials I've read) is control anchoring. Anchoring is what allows you to resize a window, and have the controls behave the way you would expect them to.

Right now, you only have two anchoring methods... topleft anchoring, and all. Allowing user control of anchoring will go a long way towards allowing usable, good looking applications in ZML.

As for how to allow users to control the anchoring, I would recommend copying the way .NET does it... it's quite easy to use, and reasonably powerful.

Wizard Joined: 10 Oct 2000 Posts: 1551 Location: Australia

Borland invented anchoring, dot net took years to copy it :P

Zugg's been using anchored controls since the dawn of Zmud, I would have thought, he probably hasn't exposed the properties yet, but they're a standard part of Delphi ;)

Magician Joined: 13 Oct 2000 Posts: 463

Heh, I can only discuss what I've seen. .NET was free, Borland was not. :-)

Posted: Mon Jun 21, 2004 9:33 pm

Zeus already has this. Take another look at the demo apps. You'll see several components that use the "Align" property to control this. There is also an "Anchor" property that controls anchoring at an even lower level. The calculator demo is an excellent example of some of the other anchoring properties that I've added on top of the normal Delphi "Align" and "Anchor" properties (both of which are available for any control). I added stuff like "AutoTop" and "AutoLeft" and also "DeltaWidth" and "DeltaHeight" that let you easily tweak the position and size of an aligned control very easily.

Anyway, it's another case of not having any documentation for this yet. As Rainchild mentioned, Delphi has always been years ahead of Microsoft when it comes to component design and user interface and .NET copies a *lot* of stuff that Delphi already has. Since Zeus components are Delphi components internally and Zeus gives you access to any of the Delphi properties, there is a lot you can do that you might now be aware of if you are not a Delphi programmer.

Once of the challenges I have in documenting Zeus is doing it in a way where I'm not reproducing the Delphi documentation at the same time. I could probably keep someone busy full time for many weeks just doing the documentation. I have to split my time between that and the programming that is needed to make this into a viable product to start bringing in the money needed to keep things going.

Also, there *is* a free version of Delphi, and has been for quite some time. It's very functional. Check out the borland.com site...it's in there some where. Also, if you really want to do .NET programming, the latest Dephi v8 is for .NET. That's a bit disappointing for people like me who don't care about .NET and am sorry to see Borland move away from the Win32 world, but for people who believe the marketing and think .NET is the next greatest thing, the Delphi 8 is a good step in the direction of applying the power that Delphi has always had to the .NET world. And since a lot of .NET is just Microsoft's variation on the VCL that Delphi has always had, the Delphi model fits quite nicely with it.

At least that's what I've heard. I bought Delphi 8 just because it came with a copy of Delphi 7 for Win32, but I haven't installed v8 myself because I believe I can write betters apps in Win32 than I can in .NET given all of the 3rd party components that are available for Delphi Win32.

Magician Joined: 13 Oct 2000 Posts: 463

I agree Win32 is more powerful... no argument, hands down. .NET is merely EASIER. It's the new VB, and unlike VB it's not a completely crippled programming environment. It has advantages and disadvantages... most of the advantages are relating to ease of use, most of the disadvantages are about flexibility and power.

But I'm not gonna get into a religious war... .NET is overhyped, but it's also fun to program in, for me at least. The built in object hierarchy is relatively comprehensive and easy to navigate. But I program as a hobby, not as a vocation. I'm sure that gives me a different perspective on the issue.

Posted: Tue Jun 22, 2004 8:16 pm

I think we've discussed this in another thread before. What what I have seen, basically there are two views on .NET:

a) If you have never used Delphi, .NET looks and *is* a great improvement to anything you have had before.

b) If you have used Delphi, then you look at .NET and wonder why it took Microsoft so long to catch up because basically .NET doesn't add anything that you couldn't already have done just as easily in Delphi for a long time (Delphi just calls it the VCL (Visual Component Library) instead of the .NET object hierarchy).

So yes, for most of the world (who doesn't know about Delphi), .NET is a wonderful thing. But for us Delphi programmers, it's nothing new. Everything you mentioned about it being "fun", for example, is exactly why I have always programming in Delphi instead of anything else. Programming in Delphi has always been easy, powerful, and a lot of fun. I also started programming as a hobby and that's when I fell in love with it.

I guess we can blame Borlands poor marketing over the years (or Microsoft's good counter-marketing) for not getting Delphi the exposure that it needed.

Magician Joined: 09 Oct 2003 Posts: 459 Location: USA

There are a lot of things about .NET which aren't really talked about or understood by most people and therefore, overlooked. One of the major things about it is that it has opened up a better way to see what functions and members that other objects have and has established common variable types across all .NET programming languages.

This allows you to know for certain that if you're supposed to toss a string to another executable, you don't have to worry about what format the string needs to be in based on what the original programming language was. It also makes sure that you get back the kind of variable from the other executable that it said you were.

Another layer of .NET is the operating system independence of managed code (or, the Common Language Infrastructure). If you were running a BSD or Mac OSX box and compiled a copy of the shared source CLI from microsoft called 'Rotor', then you would be able to run any pure .NET application compiled. The major problem with this at the moment is that most applications compiled using Visual Studio and other .NET programming tools is that they make use of Windows specific sources as well (such as Windows Forms). There are a couple Linux tools which are being developed to bypass this such as Mono (a .NET framework clone that also makes use of the WinE windows emulator for windows forms emulation).

There are a lot of interesting parts of the CLI which all .NET applications run on, since all .NET apps are essentually running in a Java like virual environment. Garbage collection is handled by the framework as well as buffer overrun protection on a framework level.

A lot of people bash .NET and claim that it is a lot of hype, but most of the important aspects of it aren't the improvements in the Visual Studio tools, but the actual framework itself that the applications run on.

I'm rather looking forward to what the future of .NET is going to bring.

Posted: Tue Jun 22, 2004 10:33 pm

quote:
Originally posted by Zugg

I think we've discussed this in another thread before. What what I have seen, basically there are two views on .NET:

a) If you have never used Delphi, .NET looks and *is* a great improvement to anything you have had before.

b) If you have used Delphi, then you look at .NET and wonder why it took Microsoft so long to catch up because basically .NET doesn't add anything that you couldn't already have done just as easily in Delphi for a long time (Delphi just calls it the VCL (Visual Component Library) instead of the .NET object hierarchy).

So yes, for most of the world (who doesn't know about Delphi), .NET is a wonderful thing. But for us Delphi programmers, it's nothing new. Everything you mentioned about it being "fun", for example, is exactly why I have always programming in Delphi instead of anything else. Programming in Delphi has always been easy, powerful, and a lot of fun. I also started programming as a hobby and that's when I fell in love with it.

I guess we can blame Borlands poor marketing over the years (or Microsoft's good counter-marketing) for not getting Delphi the exposure that it needed.

And most of the .NET world keeps ignoring the fact that Anders Hejlsberg is the creator of both Delphi and C#. When I first saw C#, I said "Gee, this looks a lot like Delphi." I dug a bit and found the tidbit about Anders. Very Happy

Posted: Thu Jun 24, 2004 6:35 pm

Wow, I didn't know that! Thanks for the trivia info on this BortaS.

GURU Joined: 10 Oct 2000 Posts: 867 Location: USA

quote:
Originally posted by Zugg

Wow, I didn't know that! Thanks for the trivia info on this BortaS.

Yeah, MS hired him away from Borland circa 1996, IIRC, along with Borland's Chief Developer and about 30 others over the course of two years or so. There were some conspiracy theories at the time (and a lawsuit or two) that MS was trying to steal Borland's next-generation stuff. Funny, it was the next release of VB after that (1997) that suddenly didn't suck any more :)

-Tarn

Magician Joined: 09 Oct 2003 Posts: 459 Location: USA

Recent interviews with Anders about what he's been doing at MS:

Tour of Technology: http://channel9.msdn.com/ShowPost.aspx?PostID=10116

Generics: http://channel9.msdn.com/ShowPost.aspx?PostID=10189

Origins of C#: http://channel9.msdn.com/ShowPost.aspx?PostID=10502

C# 3.0: http://channel9.msdn.com/ShowPost.aspx?PostID=10276

Common Language Runtime (CLR): http://channel9.msdn.com/ShowPost.aspx?PostID=10804

Posted: Fri Jun 25, 2004 12:55 am

quote:

No, I do not work for Microsoft....... YET!

Ahhh, but I do. Wink

Unfortunately, I'm not in Redmond. I'm be in Seattle August 23 to 26.

GURU Joined: 10 Oct 2000 Posts: 4379 Location: USA

Ahh... just a week after I leave the campus here at Redmond.

Now, let me leave the site before the filters here catch all this anti-.NET talk. Razz

Wanderer Joined: 13 Jan 2002 Posts: 76 Location: USA

I'll be honest, when I first heard of the who .NET thing, i wasn't sure what to think of it. Sounded alot like they were copying the whole middle-ware Java idea. But the more I use it the more I have begun to enjoy it. It makes building quick (and sometimes dirty) apps a snap. And the more detailed apps are quicker to build as well.

Am I a Microsoft fan? Maybe. Am I a .NET fan? Absolutely...

Posted: Sat Jun 26, 2004 6:19 am

Could someone who knows *both* Delphi and .NET point me to something in .NET that should get me excited. I keep looking and I can't find anything that I haven't been able to do already for years. As far as an architecture, .NET might have an advantage in the next version of Windows, but for now it's adding another layer of speed/size that Win32 doesn't need, and the Win32 COM architecture seems to work just fine for me. True, you need a separate program to browse the COM objects on your system and see the properties, methods, etc, but it's available.

Sure, .NET is supposed to run on all platforms. I'll believe that when I see a *serious* interactive application (not one that just has web-output style windows and dialogs) running on both platforms properly. Java always claimed cross platform also and never really delivered this for complex applications. Last I heard .NET on Unix was still just a theory, but maybe someone has implemented this now?

Even still, anything that runs cross platform is by definition less efficient than platform-specific code. Sorry to be a devil's advocate here, but to me .NET still seems just like a way for Microsoft to convince everyone to buy new versions of software (and Windows) and developers to buy new development tools. That might be an improvement for VB and Visual Studio people, but for Delphi people I'm still struggling to see anything I should care about.

If you are going to put in another layer between the application and the OS, I'm much more interested in scriptable applications ala Zeus, Mozilla, (and whatever the new Microsoft stuff in Longhorn is called). At least that gives the end user the ability to customize applications to look and work the way they want it.

Beside, if you point me to something cool in .NET that I've missed, I can always add that to Zeus Wink

Adept Joined: 12 Nov 2000 Posts: 268 Location: Canada

I just want to take a jab at your one statement. You say that anything that runs cross platform is by definition less efficient than platform-specific code.

This is not necessarily the case for java. You can write java code once, and then compile it for specific architectures with the virtual machine compiled directly into the application.

I currently use a program called Azureus which doesn't even do this, it is actually interpreted at run time, and it is as fast, and efficient as any other program of it's kind (it's also beautiful).

And I've personally made programs cross-platform specific (compiled in java for specific architectures using some fancy tools) that have yielded equal speed times.

However this has nothing to do with .net so I will yield, and return.

Posted: Sat Jun 26, 2004 7:05 pm

While I don't dispute your actual results in the cases you tried, I'm having a hard time figuring out in theory how that could be. If your code is interpreted, or running on a virtual machine (which is essentially an interpreter at the machine code level), then you are executing more instructions than code that has been compiled directly into native machine code.

If you only compile java once, even with the virtual machine compiled in, the java code is still being executed by the virtual machine which is converting java code into machine code.

It's a somewhat different topic, but if anyone knows how Java supposedly performs this magic of being just as efficient as native compiled code, I'd be interested in hearing it.

Perhaps the virtual machines have improved tremendously. Several years ago I was playing with the Borland J++ compiler and was trying to do a MUD client in Java (thinking of possibilities for zMUD) and it ran so much slower than zMUD that I abandoned the whole project. That was compiled Java code running on the Microsoft Virtual Java machine that used to be part of Windows back then (it was Windows 2000).

Of course, this is a bit off topic and I'd probably be more interested to know how a .NET application performs compared to the same application compiled directly for Win32. Also what the memory footprint difference is (not counting all of the .NET libraries and runtime stuff that is needed)

Magician Joined: 09 Oct 2003 Posts: 459 Location: USA

A third-party group on behalf of Microsoft compiled a version of Quake II for .NET

http://msdn.microsoft.com/visualc/homepageheadlines/quake/default.aspx

There is a simple compile switch in VC++ where you keep all of your original non-.NET syntax and the compiler simply makes the conversion to managed code. This also means that the code is trying to do things itself that, if it were optimized, the framework would be doing for the application. So, a lot of cruft doing it this way... but it allows a simple way of converting from VC++ native code to VC++.Net managed code.

The time I saw it demonstrated it had an 18% drop in performance (60 fps down to 50 fps). However, this was with non-optimized code for dot net. (more details in the video on: http://msdn.microsoft.com/theshow/Episode035/default.asp in the 'Enter the Programmer' section)

So, you can do a real application in .NET as opposed to what I've seen with Java.

As for Unix implementations of the .NET platform, Novell is backing Ximian in producing a set for Linux, Unix and any other platform they can easily port to from there. ( http://www.go-mono.com/faq.html#platforms )

GURU Joined: 24 Sep 2000 Posts: 1237 Location: USA

Recently saw an article ( http://www.joelonsoftware.com/articles/APIWar.html
) where it's detailed why .NET matters -- in a round-about way: Microsoft is changing the Win32 API. Not just adding new stuff: CHANGING the way existing stuff works. No more backward compatibility in future OS versions.

What this means is that unless your compiler's vendor keeps up on new versions that adhere to the new rules in future Win OS's (and you re-compile for each new win version), you're going to be squeezed out of a market. Unless of course, you fall down in obedience of MS, use their own products, and endure their product gaffes.

THAT should get any non .NET programmer excited. Although probably not enthusiastic.

Wizard Joined: 14 Oct 2000 Posts: 1368

quote:
Originally posted by Darker

Recently saw an article ( http://www.joelonsoftware.com/articles/APIWar.html
) where it's detailed why .NET matters -- in a round-about way: Microsoft is changing the Win32 API. Not just adding new stuff: CHANGING the way existing stuff works. No more backward compatibility in future OS versions.

Remember that Win16 is replaced by Win32. Still Win16 pieces exist here and there. This does not mean Win32 will vanish all of a sudden.

What we see here is a kind of Win3.11 reaction. When Win95 came out a lot of people decided that it was really, really bad. Win95 was the worst OS Microsoft had ever made! Of course Win95 did have many bugs but it is so different from Win3.11 that they really cant be compared.

In the article(I just looked at some parts quickly) they mention WinFX to be the new API replacing Win32. You make it sound like Microsoft would change Win32 to break it. I doubt they would do that - perhaps they plan on replacing it with WinFX with time however.

quote:

What this means is that unless your compiler's vendor keeps up on new versions that adhere to the new rules in future Win OS's (and you re-compile for each new win version), you're going to be squeezed out of a market. Unless of course, you fall down in obedience of MS, use their own products, and endure their product gaffes.

THAT should get any non .NET programmer excited. Although probably not enthusiastic.

This is why I dislike programming languages made by companies, like Java. What if they decide to change next version too much, or what if they just close down the project? Still they might try to earn some extra cash by protection their trade mark so other people would have a hard time making Java compilers. Also the virtual machine would get outdated. A non standard development environment/language such as Delphi would also have this issue.

What I still don't get after reading this forum for some time is why Zugg dislikes .NET. Why? I made my zMUD translation tool using .NET and I suspect I would still be working on showing the forms if I had been doing it the win32 way Very Happy

. That is partly because the new devstudio has a much better development environment for .NET than standard Win32. Atleast in my opinion.

Magician Joined: 09 Oct 2003 Posts: 459 Location: USA

quote:
Originally posted by Rorso

A non standard development environment/language such as Delphi would also have this issue.

Delphi 8 is fully .NET compliant and has tools for moving from win32 to .net...

Also, .NET doesn't completely cut you off from having COM support and being able to be backwards compatible with COM.

*ponders the Borg for some odd reason, but shakes it off*

GURU Joined: 24 Sep 2000 Posts: 1237 Location: USA

quote:
Originally posted by Rorso
In the article(I just looked at some parts quickly) they mention WinFX to be the new API replacing Win32. You make it sound like Microsoft would change Win32 to break it. I doubt they would do that - perhaps they plan on replacing it with WinFX with time however.

But they would break it. It's a lengthy article, but it details 2 camps at Microsoft - the "Always Backwards-Compatible" camp that has reigned until recently, and the "MSDN Magazine" camp that has reigned since .NET came out, where VB.NET was not backwards compatible with VB.

Some developers screamed and MS made an attempt to appease them in a subsequent release, but they learned that not too many people screamed. That's their justification for breaking backward compatibility.

Sure, some parts of the Win32 API won't change, and will still work under the next version - but some will change -- not additional optional arguments and other "harmless" changes, but real live, worked in XP, doesn't in LH changes that require recompiles and maintaining multiple versions of code to support both popular operating systems before one or the other dies out (a span of years).

The whole article is worth a read.

GURU Joined: 10 Oct 2000 Posts: 4379 Location: USA

quote:
Why? I made my zMUD translation tool using .NET and I suspect I would still be working on showing the forms if I had been doing it the win32 way.

I agree that .NET allows you to create programs, fast. But I've also used Delphi before so I'll just save Zugg the trouble of saying the same thing. He does not need to get excited by this because he has been able to do it for years. Take a look at Delphi if you can, and then imagine that it's been like this for years. Yes, it's Win32, but the IDE and VCL makes creating programs fast and easy. Just like with .NET, you just create a new Form object, set it's properties the way you like it and call its Show method. That's all there is to showing a window on screen.

Keep in mind, what .NET really offers is not rapid application development, that has been available for years and it is what versions of Visual Studio before VS .NET were supposed to offer (Visual Basic came closer but you couldn't make a "decent" program on it). The real "gimmick" behind .NET is the code portability and interoperability (with other code in perhaps other languages) and the managed environment which "babysits" your code. However, they also decided to throw in a huge library of classes because they knew that otherwise they wouldn't get anyone to use .NET

Posted: Tue Jun 29, 2004 7:15 pm

Thanks Kjata, you saved me the trouble of repeating myself. Again I'll lament that it's too bad Borland wasn't able to publicize Delphi more heavily. I've been programming the ".NET way" for over 8 years now thanks to Delphi. I don't need a new version of Windows to do this. Microsoft is just trying to make everyone think that this is something completely new that they invented...it's not. If you were willing to use non-Microsoft development tools, you could have been having fun developing Win32 applications for years, which is why I've had fun with project like zMUD that would have been horribly "unfun" using something other than Delphi.

Anyway, back to Darker's point. If you've read detail on some of Microsoft's history, especially in the early days, you'll see that they got very lucky in many situations where things went their way. but might not have. I've always thought that one of these days their luck might end, and the article you mention points the way.

My prediction is: If Microsoft seriously damages it's backwards compatibility (in that existing WinXP applications do not work on Longhorn), then I think you will finally see the market share of Windows drop considerably. The only time Microsoft has ignored backwards compatibility is in the ability to run 16-bit Win 3.1 apps on Windows XP. That means that applications that have not been updated in the 8 years between Win 3.1 and WinXP "might" no longer work (and yet there is even still the compatibility mode in WinXP that can be tried).

If Microsoft was dumb enough to break existing Win32 applications in Longhorn, they will finally suffer the consequences.

Back in the Win 3.1 -> Win 95 days, or even more recently in the Win98 -> WinXP transition, things were *very* different. In both of those cases you were replacing operating systems that had severe limitations and problems with a better system that was more stable.

But now, WinXP *is* stable. There are not any problems with memory limits or disk space limits. It's a much more open architecture that actually works with new hardware. It is *not* the buggy limited OS that we had with Win 3.1, Win95, or Win98.

That means there is less of a reason for people to upgrade. The main way people will get the new version of Windows is via new computers. And I'll tell you, from being in the corporate world for several years, a company is *not* going to suddenly replace every computer and is even not going to upgrade Windows on every computer just because Microsoft decides not to provide Win32 compatiblity. If they break compatibility with any sort of major business application, companies that have contracts with Microsoft will be screaming bloody murder. Microsoft might have a lot of cash in reserve, but they still can't afford to loose their core business customers.

So, *maybe* we'll see a version of Windows in 8 more years that starts to remove some old Win32 stuff. But no time soon, certainly not in Longhorn. Microsoft would never take that kind of risk with their existing customers.

What you will see is that Microsoft puts all of their new development efforts into .NET stuff and WinFX. They will be constantly changing errr "improving" various .NET components. It will be like the days when every new Internet Explorer upgrade was changing the COMCTL32.DLL file. Microsoft software will keep updating .NET system components. Because of the design of .NET, this shouldn't be too bad, but I'll be surprised if they are able to do it without some problems with existing .NET stuff.

The Win32 stuff will remain frozen. They won't update it or "improve" it, they'll just leave it alone. There is no reason for them to remove something that provides them compatibility with previous versions of Windows. What they might do is finally actually remove some of the remaining Win 3.1 components that are still lurking in there, but I doubt they will touch Win32 stuff.

And you know, if they decide to leave the Win32 stuff alone, that's fine with me. I think the current system works fine and I'd like to see some stability in the OS so that developers can concentrate on their own software and not worry about continuously trying to update their apps to keep up with OS changes.

.NET is fine. I don't hate it. It provides a nice framework for people using Microsoft development tools and should improve the overall quality of software available for Windows. Someday in the distant future I might be forced to use "Delphi for .NET" just to keep up, but that's a long time from now. But my business is all about providing customer-focused software that runs on PCs and software that works the way the end-user wants it to work...and you know, despite what Microsoft would like you to belief, you don't need .NET to deliver that kind of software. .NET will help *more* developers deliver that kind of software, but with tools like Delphi you can still do just fine without using .NET.

I think Microsoft needs to be *very* careful with the compatibility issue. Without backwards compatibility, there is suddenly no need for anyone to use Windows instead of Linux or Mac. Compatibility is the *only* thing keeping the Windows market share alive right now. Microsoft has stockholders to listen to...they aren't going to do anything that stupid.

Posted: Tue Jun 29, 2004 8:04 pm

OK, sorry to post again so soon, but I just got done actually reading the entire article that Darker mentioned. You need to GO READ THAT ARTICLE! Here is the link again just to make it easy: http://www.joelonsoftware.com/articles/APIWar.html

This guy makes some *very* good points and I actually think I agree with a great majority of his article. I am not nearly as convinced as Darker, however, that the conclusion to the article is that Microsoft is giving up Win32 compatibility any time soon though and I'm not sure where that conclusion came from.

A couple of points of my own. In the article, he mentions that the "MSDN Magazine" team won when the new VB.net wasn't compatible with VB 6 and few people complained. I think he jumps to a huge conclusion...he is assuming a couple of things here:

1) That a lot of people upgraded to .NET...I think (as he even admits later in the article) that many people are just sticking with VB 6 and are not using VB.NET

2) That people didn't complain. I'm sorry, but if you read some of the VB forums, you'll see that there are some vocal people who *are* upset.

3) That VB actually matters at all. I mean come on...you don't write serious business applications in VB. I've got a couple dozen applications running on my computer right now and *none* of them are written with VB. For serious applications every developer knows that you use Visual Studio (or Delphi :). VB is a "toy" you use for learning, or something you use for a "quick and dirty" app, or something you use for consulting to write "yet another database application" for some small business (that's just a step up from using MS Access).

If Microsoft came out with Visual Studio .NET and made is incompatible with the current Win32 Visual Studio (and I'm talking about source code that compiled in Win32 that suddenly doesn't compile at all in .NET), then you would hear a *lot* more complaints about developers.

But the guy admits it later in his article. He admits that most developers are not keeping up with all of the new stuff Microsoft is tossing out there. Mainly because they talk about so much stuff that isn't even going to be useful till be have Longhorn. And I'm sorry, but some of us have real development jobs and have customers that we need to keep happy. I can't waste my time playing with some new technology that isn't going to be viable for 4 years and that Microsoft might even completely change between now and then. I have software to write, thanks. And for that I'll use existing stable tools, whether they are Delphi or Visual Studio.

Someone really needs to take a serious poll of *real* programmers that are doing this for business and find out just how many people are writing serious .NET programs yet. I think a lot of people are playing with it and testing the waters, but nobody is betting their company on this stuff yet (except Microsoft ;)

The final part of his article was very interesting and touched a part of me. I *also* do not want all of my apps to switch to being web apps. Even ignoring the huge security issues (which he also completely ignores in his article), I hate the user interface of web-based apps. I hate the slow response. Why are we moving backwards? I have specialized in user-interface design since I got my computer science degree back in 1983. In all of that time, I have seen user interfaces get worse more than I have seen them get better. I was furious at Microsoft when Microsoft Money adopted a "web look and feel" and became a crappy hard to use program. I only continued to use it because it Quicken wasn't much better and I was stuck with all of this archival data in MS Money. So now every week I dread balancing my accounts because the user interface is so bad in that program.

And don't get me started on web-based mail (obviously I hate that otherwise I wouldn't be writing eMobius).

But he is correct...there seem to be a lot of people who just don't care. They use web-based mail because it's cheap, and because then they can access their mail from multiple locations. They could care less that you could also access mail from multiple locations if there was a decent IMAP mail client available. They don't even know what IMAP is. They want something that works and is cheap, with "cheap" being the highest priority.

Well sure, lots of these kind of customers exist. They aren't my customers or the kind of customers that I develop software for. I'd call them the "AOL users" but that isn't right either since AOL users are used to paying for their service. But I will call them the "casual" computer users. The people that want free web-based apps are the people who really don't use their computer very much. They might play the latest games, but even for games they have probably started using their XBox and PS2 more than their computer just because computer games are too complicated to install. The apps these people use the most are web browser, email client, and instant message chat programs. They might have Office installed, but they only use it for a quick letter for those wierd companies that still require paper. Many of them probably have no real clue how to use Excel and wonder why Powerpoint is even part of Office. After all, why do you need Powerpoint when you can just send web articles to everyone you know.

For these "casual" computer users, I agree completely with this article that Win32 is dead and the current and future development is all web-based apps, with poor user interfaces, that are cheap or free. How developers will stay in business doing this is anyone's guess.

For the "business" computer users, they also don't really care. They spend all day using whatever corporate applications are required for their job. They really don't care if they are using Windows, Linux, or Mac as long as they can run their custom business application easily and efficiently. During lunch breaks (heh) they surf the web, and they use email. The company firewall probably blocks instant messanging. Their secretaries use Office a lot for memos, reports, graphs, charts, etc. But they mainly rely on their corporate IT department for their needs. Who cares if Microsoft changes the Win32 API in the next version of Windows...that's what we pay the IT department to take care of.

Then there are the "non casual" computer users. I hate to call them "hard core" users. But these are people who like to use their computer for lots of different stuff. They want software that works the way *they* want it to. They want it to run fast, use less memory, use less disk space. They want cool software. They have 10 programs running on their desktop at once and would love to make them all talk together if they could. They still run PC games because they are still better than console games ;)

"Non casual" computer users *DO* care about backwards compatibility. They will scream if Microsoft gets rid of Win32 compatibility. Are there enough of us left for Microsoft to worry about? We'll see. I think there are more of us hiding in the background than anyone considers, and I think we contribute a *lot* of the software sales money.

Think about it. The "casual" computer user isn't buying a lot. Even in the rare cases where they buy software, they will use that same version till it's dead. They do not buy upgrades as soon as they are announced. Some of these people *paid* for zMUD 4.62 and are still using it because it's too much trouble to upgrade.

Business users contribute a *lot* of money. This is where Microsoft makes their big chunk. It's all about support contracts. They upgrade to the latest and greatest stuff as soon as they can. It keeps the IT people employed. They don't have to think and innovate...they just need to keep the company software updated for the latest round of changes from Microsoft. They probably spend way too much of their time these days just reinstalling software on systems that have been trashed by viruses anyway.

But "non casual" users actually spend a lot of money too. Not as much as businesses, but more than the casual user. I'll actually buy a piece of Shareware that does something that I need. I'll even buy it if I think it's cool or innovative, even it I don't use it that often (just bought the entire StarDock set of tools for customizing WinXP for my entertainment computer downstairs...ObjectDock and WindowBlinds are just cool!). I keep my software updated most of the time. Sure, I skipped Delphi 6 and jumped from Dephi 5 to 7, but I still upgraded. I'm actually using Office XP instead of Office 95 now. So I don't upgrade *all* the time...only when it helps with something I'm doing. But I'd certainly rather pay for a quality piece of software that does something I need rather than using a free web-based application.

So, we'll see how vocal the "non casual" user ends up being. I don't think they (us) have really been heard yet. I think Microsoft is doing it's testing with the casual user and the business user. The casual user will use whatever Microsoft wants. The business user will get all excited about new stuff from Microsoft because it might keep them employed in their IT department.

So, after reading the article, I'm back to my previous position. .NET is fine. It helps a *lot* of developers. No problem with that. But if you take away my Win32 API, I'll start screaming and Microsoft will regret it in the end.

Thanks for posting that article Darker. It was a fun read.