Submitted by “The current Win32-based Windows UI graphics subsystem, found in Windows XP, has been around for nearly 20 years. It’s aging and limited, and as a result, user interface development has been somewhat, well, constrained at best. The Windows Presentation Foundation, which is built on the .NET Framework, provides new techniques for developing applications, and makes better use of current hardware and technologies. In this article, we’ll show you 10 of the most significant advances that make WPF superior to its Win32 forebears.” In related news, here are shots of Vista server.
Okay so I just skimmed the article, but still. Basically it was saying that you no longer have the dodgy limitations and hacks that were in the previous model.
Presumably you will have a whole new set?
I wish they never touched the graphics component of windows and devoted its development time to rewrite or redesign the security features of windows.
If they have extra time I wish they spend it trying to prevent applications from conflecting with each other.
I wish they never touched the graphics component of windows and devoted its development time to rewrite or redesign the security features of windows.
If they have extra time I wish they spend it trying to prevent applications from conflecting with each other.
You are aware that you can have more than one group of people working on windows right? They have people focusing on the things that you want as well as (hopefully) fixing the GUI API.
“They have people focusing on the things that you want …”
Actually they don’t. Could you please give me evidance of that; as I was able to infect longhorn vista 5270 when going to a malacious site and made software conflict with 2 optical disk emulation softwares. All MS wants is to produce good looking GUI to get reason to sell its newer products.
Actually they don’t. Could you please give me evidance of that; as I was able to infect longhorn vista 5270 when going to a malacious site and made software conflict with 2 optical disk emulation softwares. All MS wants is to produce good looking GUI to get reason to sell its newer products.
Spot on.
As Bill Gates says:
“If you can’t make it good, at least make it look good.”
Microsoft, in general, has never done anything the user wants.
(1) They send spies to businesses to figure out what they want in terms of an interface in a work environment. Why not ask customers directly? Why do you have to do research in such a discrete and unconventional fashion? How hard is it to walk up to someone and ask?
(2) They only start to adopt something when it starts to affect their market share or potential income.
Example 1 : Tabbed browsing. (Introduced by someone else, was it Opera? And made popular by Firefox…Now IE7 is getting it.)
Example 2 : There was previously no PDF export function built into Office, the upcoming version will have it, why? Have a think about it.
I’m done with looking articles like this. OSNews used to pretty cool, Eugenia did a pretty good job…Now, its just start to be another tentacle to Microsoft’s PR machine.
This looks like a big step in the right direction.
Has anybody seen WPF in action ? How is performance ?
Can’t wait to see the design tools…
On the whole this looks interesting. Might lower the barrier of entry for people writing UI intensive programs.
And the clear separation between code and UI means that developers can finally offload the UI design stuff to designers.
> I wish they never touched the graphics component of
> windows and devoted its development time to rewrite
> or redesign the security features of windows.
>
The security model of .NET is mature and actually quite good. And it also solved the “dependency hell” problem you seem to speak about. So both of these issues are already solved.
But WPF is managed code and there´s no guaranteed compatibility between different versions of .NET and this is what I think It´s worong to write core parts of the operating system in managed code. Will it be a WPF for .NET 2.2/2.2/3 etc. What will Microsoft do?
.NET runtimes can run side-by-side.
You can even tell what app to run on which runtime version.
they’ll do exactly what they did with directx up to date: duplicate the old functionalities in the new api and “carry in the bag” the old version. imho, of course.
Oh, so they got around discovering GUI practices that have been around for ages elsewhere ?
Will that mean that they at least will stop inflicting us small fixed size dialogs where we have to scroll around or use tooltips everywhere to be able to get at the information we need ?
(aaaah, the joy of checking the list of libraries that a visual c++ 7 project is linked against)
Around for ages elsewhere? You mean like X, which doesn’t yet even have proper hardware-accelerated alpha channel support? Yeah, of course.
The only other OS that comes close to Vista’s capabilities is the latest OS X — and Vista will have the lead here, at least until we see what Apple demonstrates with 10.5.
I like how you probably didn’t even read the article and just decided to belittle Microsoft’s work. 🙂
No, I was talking about the useful part of the article, the one about layout. I don’t care about the eye candy.
Belittle Microsoft’s work ? They do it themselves.
Reading in 2005 that one of the “top ten UI development breakthroughts” in their future UI toolkit is adaptable UI layout when there were toolkits doing this ten years ago on amiga (and probably even longer elsewhere) is utterly ridiculous.
But hey, they’re Microsoft, they’re the largest developers on the planet, who am I mere mortal to laugh at them ?
> No, I was talking about the useful part of the article, the one about layout. I don’t care about the eye candy.
Too bad the rest of the world does. Stay with your command-lines, then. What’s another 5 years behind the race if you’re already 5 years behind?
> Belittle Microsoft’s work ? They do it themselves.
Spoken like a true zealous Linux user.
> Reading in 2005 that one of the “top ten UI development breakthroughts” in their future UI toolkit is adaptable UI layout when there were toolkits doing this ten years ago on amiga (and probably even longer elsewhere) is utterly ridiculous.
As always, the “Microsoft has done something cool” articles are usually taken out of context by you Linux zealots. You always twist it around into meaning something else, and then compare it to another product that existed years ago. I wouldn’t be surprised if it was the same case here.
It’s okay, though. The plastic bubbles are all nice and comfy. Microsoft isn’t ahead of Linux. Oh no — they’re very behind. Just keep believing that, and you’ll definitely be the leader. 🙂
Speaking like a zealot indeed.
Run through the god damn list, and show me how the current version of Win32 is ahead of GTK+ in the specific situations mentioned. Otherwise, stop spouting nonsense.
It’s clear when you have to start reverting to such foul-mouthed communication that you’re immature and insecure in the reality of the situation. You really need to investigate things in WPF before you show your rancid mentality like this, as it makes you look stupid…
Run through the god damn list, and show me how the current version of Win32 is ahead of GTK+ in the specific situations mentioned
It’s not on the list but:
1) Win32 is already installed on 90+% of the desktops out there.
Static building with Gtk is possible; moot.
No, I was talking about the useful part of the article, the one about layout. I don’t care about the eye candy.
Too bad the rest of the world does. Stay with your command-lines, then.
Whenever I make this kind of statement, people always seems eager to try and prove that they are mentally challenged.
I will spell it out for you since you seem to have a hard time working it out: that I dont’t care about eye-candy doesn’t mean that I don’t care about usability.
Again, I don’t care about the whole animated eye-candy thing, regardless if it’s on windows, macosx, linux or whatever. I don’t believe that it will change anything in the long term, and I don’t think it’s a particularly exciting technology either.
I don’t care who of microsoft or linux will first have the software infrastructure to do this crap.
> Belittle Microsoft’s work ? They do it themselves.
Spoken like a true zealous Linux user.
You’re the one bringing up linux here. Are you some kind of paranoiac thinking that everyone disagreeing with you is a linux zealot out to get you ?
As always, the “Microsoft has done something cool” articles are usually taken out of context by you Linux zealots.
They have not done something cool. They’re fixing some of their most obnoxious usability problems (please note that again, I’m not talking about the eye-candy crap, but only focusing on what I consider a usability fature, ie the layout system)
You always twist it around into meaning something else, and then compare it to another product that existed years ago. I wouldn’t be surprised if it was the same case here.
No, it’s just that I can’t stand bullshit. And I compared it to a product that existed years ago, but of course I could have mentioned Qt, GTK, WxWindows, FLTK, or about any other existing GUI toolkit that already do that.
It’s okay, though. The plastic bubbles are all nice and comfy. Microsoft isn’t ahead of Linux. Oh no — they’re very behind. Just keep believing that, and you’ll definitely be the leader. 🙂
Again, you’re the one bringing the whole issue up. I don’t give a shit either if linux is leader or not, I’m not insecure enough as to need that a majority of people make the same choice of OS as me.
However, if MS have to stay ahead, then I would like them to fix the most outstanding problems that make me curse them daily.
The obvious usability problems are one of them, and I hate people drinking the microsoft kool-aid who claim that they are somehow a reference when it comes to usability.
Seeing that I use windows about 8 hours for each hour that I use linux, if it was so perfect by now I should be so used to it that it’s using linux that should annoy me.
They’re fixing some of their most obnoxious usability problems (please note that again, I’m not talking about the eye-candy crap, but only focusing on what I consider a usability fature, ie the layout system)
You consider a layout engine to be a usability feature. You really are a Linux zealot, aren’t you?
You consider a layout engine to be a usability feature. You really are a Linux zealot, aren’t you?
Fixed-size dialog are a usability problem.
Dialogs that can’t resize properly to use a bigger font are a usability problem.
Dialogs that can’t adapt their size properly to the length of translations into other languages are a usability problem.
Do I need to spell out that having a layout engine allow to solve these in a way that also makes the developer’s life easier ?
I’m sorry to be upset to have to scroll horizontally most configuration dialogs trees when I’m using a french version of visual .net, and a million other similar annoyances.
Do I need to spell out that having a layout engine allow to solve these in a way that also makes the developer’s life easier ?
So what you meant to say was that a layout engine is just a tool that may help a developer with enhancing usability (or just help him fill the screen with a mass of incoherent widgets). And according to you this is the only feature being listed in the article that can help the developer when dealing with usability in an application. And this is somehow bad because other systems have had this tool for a while now. So the addition of this one feature to WPF means that it is hopelessly out of date and behind the times.
Linux does not have a GUI tookit
. Would you like me to point you to the source so you can see for yourself 
.
So, what, exactly, makes you think he’s a Linux zealout from the post you’re replying to? Or, are you simply attempting to flame him?
The reason a layout engine is a usability feature is that same reason that size checked arrays are a security feature:
Programmers neglect things and make mistakes.
A layout engine forces developers to actually think their design through in a non-static sized way. It also handles 90% of the grunt work for them.
And yes, many of us use programs in many different ways. Myself, for example, I only maximize windows if my screensize is 800×600 and it’s a browser window. I often use vertical only maximization. I manually place windows into sections of my screen so I can easily find and recognize them by location and watch the useful parts of each one at the same time.
I’ve known people who maximize every window. Even their AIM chat windows! Even their AIM buddy list!
Maybe you’re confusing accessibility and usability and that’s why you can’t see how it’s a usability issue?
hehe
It is really easy to identify the insecurities of someone name “linux is poo”.
You might try reading the article
Hardware accelerated alpha-blending is just one of the 10 points contained within. The rest are about far more basic things, like managed layout of widgets. It’s no secret that the Win32 widget set is painfully primitive in these regards, compared to more modern toolkits like GTK+ or Qt. 9, 8, 7, 6, and 5 are all basic parts of GTK+. 4-1 are specifics of XAML that can mostly be duplicated in GTK+, perhaps with the exception of 3. 5 is something Smalltalk-based UIs were famous for, and something that has made a reappearance in Qt 4.x. 1 is something that CLIM and more recently DUIM (the Common Lisp/Dylan counterparts to AWT) did in the mid 1990s.
GTK is terrible. Don’t … don’t even compare GTK to what Vista is doing.
:-/
I was just responding to an obvious Microsoft FUDster.
GTK+ has a mediocre API, but in terms of features, it’s quite a bit more advanced than Win32, and very comparable to what Vista is doing (XAML) aside. All that stuff about rich text layout, fewer widget constraints, etc, are all part of GTK 2.x. Microsoft is definitely playing catch-up in this realm.
Yes and no. I’d say they’re playing catch-up, and are past the catching up part and on to the flying by part.
Gtk+ is an excellent toolkit (Linux Is Poo is just anti anything FOSS), but it’s not up to the hype about WPF. Now, we may learn that WPF has its own stupid exceptions and things we can’t do that don’t fall into the elegant marketing we heard; but the marketing for what WPF can do does sound nice. Any widget inside any other widget; wonderful: You can do that throughout a lot of Gtk, but not all of it; toolbars are an example.
My understanding of WPF is that it’s still absolute positions (from the mouth of Microsoft’s marketing, err I mean a Microsoft programmer). This, IMHO, is stupid. How can you call a widget a child of another widget and still have its position determined by the window?
It doesn’t have to be absolute positions. It’s pretty much similar to what HTML can do for position. Also, XAML (I don’t know how much of this is XAML and how much is WPF) does grid positioning, which is really cool and useful.
Oh really? In what way? I would say that GTK+ is more advanced in some areas than what Microsoft is offering.
In what ways would that be? 🙂
Thats what it really is.
All the rest is hyperbole.
“This doesn’t stop some designers from designing a UI that only works at one size”
Don’t blame others, look at yourself.
I hate these awkward fixed-sized 5-line listboxes with a thousand of lines to scroll designed by MS guys.
Almost anything in MS windows is fixed-sized crap.
/me prefers fixed sized crap.
minimalistic. one instrument, one function, one interface. it’s just about taste
/me loves BeOS for example :°(
Some of this stuff looks really promising from a programmers standpoint, though I can already hear the die-hard XHTML nuts getting their panties in a twist with all that formatting inline in the document…
But as someone who hates window and text transparancy effects because they make it impossible to read ANYTHING, program skins that have goofy shaped windows because they don’t fit together neatly, and a bunch of other ‘trendy’ visual effects, I can also see some of these neat features ending up being abused – usually by programmers who are all flash and no substance; You know the type, the ones who spend all day futzing with the GUI of an application before they even have it functional, or worse changing it all around ‘because they can’ without actually changing what it does.
Still, it is good that they are abandoning the rather dated GDI for something a bit more streamlined and easier to program… This looks to be a FUNCTIONAL change, which made for an article that seems quite refreshing in the face of all the meaningless articles about them futzing around with another goof assed theme/skin (which of course we all know is coming as well)
You can have multiple versions of the .NET framework installed at the same time. An application normally uses the newest version of the .NET framework, but if for some reason it requires a particular version of the framework, it can say so in its configuration file.
Basically you will have multiple versions of the WPF dlls installed. Most applications will use the newest, but if some application is incompatible to the newest version, the application vendor can configure it so that it uses an older version.
By the way: the 2.0 runtime can load 1.0 and 1.1 dlls, so you will normally not have multiple instances of the runtime installed simultaneously.
It’s all fine and dandy that Windows now lets you have overlapping semi-transparent buttons. But why in your right mind would you want your UI buttons to overlap? I just see it as a total usability nightmare. Think of how much trouble a lot of less computer savy people have keeping the mouse steady as they click. Now overlap your UI widgets and watch them try to click the right one.
You can overlap widgets now, so what’s the difference? It’s simply about giving the programmer/designer as much control and freedom as possible.
People have been misreading this story for too long. Now, finally, the truth is coming out. The basis for all new MS products is the .NET platform, however many legacy interfaces will be available. A version or two after Longhorn, I still can’t call it Vista, MS will really get serious about managed code (.NET). It is a matter of MS moving their developer base to the new platform.
well, Windows GDI is first found win Win95 and it was “born old”, as the functionalities are were too limited even at that time.
WinXP (with backport win win2k) finally sees an update, in form of the GDI+. Not yet accelerated, but finally suppot for some more functions, most used ones being OS support of JPEG and PNG formats, duplicated then in DX. (That’s why the new old paint supports them, eg).
In the end, finally something new and possibly well done (they have the experience of dx now) for gfx foundations.
And so even those who never understood what opengl is and that it was supported since Win95 will (maybe) have decent gfx performance
GDI+ came out with the release of 2000.
Windows GDI pre-dates Windows 95 by a LONG time: it was there before Windows 3.0, and anything before Windows NT 3.1 or Windows 95 is 16 bit. That is why (from your perspective) it was “born old” is because it was already old long before 32 bit versions were created.
The changes are welcome, as the programmatic aspect of things had gotten rather snarled over time, though I’ve not read enough to know exactly how everything will work from the programmer’s point of view. It’s clear that the only way they could do the compositing engine to have things work efficiently is to decouple as much from each GUI application as possible in terms of updating the GUI that’s displayed: the fewer trips to/from a process to refer to drawing instructions there are, the much higher performance that can be achieved, even with a hugely complex user interface with all applications on screen at once. What I’m most curious about is how this will affect performance when there’s a mix of old Windows GDI-based applications and newer WPF applications.
Jonathan Thompson (writing Windows apps before Windows 95 or Windows NT)
I think windows developers are way to logical… they let the programmers do the UI designs which I think as a regular common windows user, is just plain WRONG, because everithing looks like “Standard”.
Also I hate the fact that windows does not support alpha channel in IE6 not to mention the whole OS.
Besides, what’s with the colors? I’m not an artist or graphics designer, but those color schemes or they are so color blinding (http://presentationzen.blogs.com/presentationzen/2005/11/it_was_one…) or so tasteless (http://activewin.com/screenshots/vista/5270serv/Pictures.JPG)
Maybe that’s why it took them 20 years to update their graphics engine, I think those 4 eyes programmers are just plain blind. And Steve Jobs was the first one to notice that.
Great so their UI’s are in C#. I still don’t see the big deal and it’s bulkier. I dunno. Is C# really that great over managed C++ libraries that are more open source friendly which is the future by the way. I see more vulnerable security risks because of the closed source nature of .NET.
I’m not trying to bash C# but I dont see why they are bashing C++ stuff like QTools, GTK+ and Fox which are very simple to use.
Great so their UI’s are in C#. I still don’t see the big deal and it’s bulkier. I dunno. Is C# really that great over managed C++ libraries that are more open source friendly which is the future by the way. I see more vulnerable security risks because of the closed source nature of .NET.
I’m not trying to bash C# but I dont see why they are bashing C++ stuff like QTools, GTK+ and Fox which are very simple to use.
You should read about .NET’s architectual details and code access security. This should give you a more accurate view of C#/.NET and provide the reasons why they chose to implement WPF and some other user mode systems in managed code.
There’s nothing to be gained by using managed C++ over C# in implementing managed code. It all compiles to IL, then native machine code before execution. It’s also incorrect to state that managed C++ would be more OSS friendly. If it’s managed code, you’re going to need a compatible runtime to use it. Last, the article doesn’t even mention the other toolkits. If you’re just asking again, “why not use C++?”, the primary answer is security. Toolkits such as the ones you mentioned are vulnerable to coding errors like buffer overruns because they are implemented in languages like C++.
Basically I meant managed in terms of managed libraries not MS’ definition of managed code. Also their definition of safety is awkward to me. Safety means a coding error? I would think safety should mean code security like if someone hacks it.
My 2cents. Maybe C# is better as I am coming from C# into C++ but am liking C++ better because of the easier access to the managed portion of the code. We’ll seee…. I like PHP too because of the ability to jump in and out of managed code on the fly. It’s really rising in usage after PHP 5.
Basically I meant managed in terms of managed libraries not MS’ definition of managed code. Also their definition of safety is awkward to me. Safety means a coding error? I would think safety should mean code security like if someone hacks it.
Safety means verifiable. The system can tell what the code is doing and restrict it to only performing operations it’s allowed to do. However, being able to catch more errors at compile time also adds to code safety. I don’t understand exactly what makes you think C++ is safer than managed code. If someone is able to co-opt an application written in C++ or another unmanaged language, they can execute code in the context of the user running that process. If the user is an administrator, the code will be able to do anything. Given the same situation with a .NET application, the malicious code could be prevented from running. or prevented from running with enough permissions to do anything harmful because it would be stuck with whatever permissions the co-opted application was given by the developer and according to the context in which it is run. Even if the application is being run by an admin, the application could be limited to a more restrictive set of rights than the user has.
So that type of safety deals with the entire Vista? Weren’t they taking .NET out of the Kernal? How would this work in all instances? Souds fantastic but seems far-fetched to me as it’s not a complete system like C, C++.
Yes, I studied this a a little and understand the CLR to an extent. Seems great for UI but I have noticed allot of bugginess in Interop although not entirely un-useful. Are they expecting, for instance, game makers to used managed C++ and have their games go allot slower as well as unoptimized? The large players are not going to use managed DX or at least be happy about it.
That’s me I signed in now.
So that type of safety deals with the entire Vista? Weren’t they taking .NET out of the Kernal? How would this work in all instances? Souds fantastic but seems far-fetched to me as it’s not a complete system like C, C++.
.NET was never in the kernel. It’s a runtime and a set of class libraries (all user-mode) and the system as a whole is actually more complete than C/C++. It’s a platform built to support multiple languages (including fully managed C++, e.g., C++/CLI, or interop with regular C++).
Yes, I studied this a a little and understand the CLR to an extent. Seems great for UI but I have noticed allot of bugginess in Interop although not entirely un-useful. Are they expecting, for instance, game makers to used managed C++ and have their games go allot slower as well as unoptimized? The large players are not going to use managed DX or at least be happy about it.
The actual performance differential between unmanaged and managed code is about 10 – 20% (using version 1.x of the runtime — v.2.0 has had significant perf improvements over 1.x). All managed code is compiled to native code before execution. It can even be pre-compiled, either at install time, or prior to distribution (C++ style, though you incur the same platform support penalties as C++ if you do this). The primary cost vs unmanaged code is some overhead from runtime services and the garbage-collector, but the benefits usually outweigh the cost, and the cost isn’t that high. In some cases, the managed code has shown to be faster than unmanaged. For todays games, most heavy lifting is done by the GPU, making the differential even less of a concern. There are games/game engines already using managed code/MDX. There are also support tools (editors, etc.) being built w/ managed code.
I expect adoption to continue, however, it really depends on the business case for each individual developer. I wouldn’t expect id Software, for example, to adopt managed code for development anytime soon because they do cross-platform work. Managed code would benefit them greatly here because they could have one codebase to maintain that could compile for all platforms, but they would need to evaluate the performance and capabilities of Mono or secure some other cross-platform implementation of the CLI. There are already OpenGL libraries for managed code (not sure about cross-platform sound/input libraries but they’re probably out there). UI can be customized on a per-platform basis, or they could use a cross-platform GUI library.
Vertigo Software created a managed port of Quake 2 for Microsoft. It runs on MDX 1.0 and .NET 1.x, and took them just a week to port the code and make a managed code feature addition. They could’ve gotten better perf if the game was actually written for managed code, but performance of the port was acceptable (within 20% of the original) and allowed them to make quick feature additions. It’s available for download ( http://www.codeproject.com/managedcpp/Quake2.asp ). An example of a game/game engine built in managed code is Arena Wars ( http://arenawars.krawall.de/eng/index.html ).
The obvious wins for managed code development, for games and other software, are security, uniform coding guidelines (aids readability/maintainability), memory management (this is good for security and productivity), and IL (one codebase, multiple platforms, one distributable, 32/64-bit.
RE: Vista
Vista is not written in 100% managed code. The kernel and kernel mode systems are still C/C++, as are some lower-level user-mode systems and legacy APIs (win32 etc.). WinFX, the new set of APIs that sit alongside and will eventually replace Win32, is managed code. Most apps targeting .NET or WinFX will get the greater security benefits of managed code. You can write unmanaged apps that also target WinFX, however in many cases, the managed -> unmanaged interop story (and possible, slight perf hit) that exists on current versions of Windows will be reversed. As time goes on, more of Windows will move to managed code, supporting unmanaged code via interop. Managed apps will also have more flexibility (and do currently) in being able to run in some contexts that unmanaged code won’t due to security concerns.
There is a lot more to Vista security than just managed code, however. A lot of work has been done to curtail what unmanaged apps can do without being trusted by the user. This goes beyond normal account permissions controls. Unmanaged apps may need to do more work, in some cases, to give users a “just works” experience under limited rights scenarios, but it’s still fully supported.
Edited 2005-12-23 22:41
Ok so there are ways to add type safety to C++ in Vista so there are pros and cons especially C++ being fully opensource which adds allot to security as well.
I haven’t had any major hiccups with safety of C++ on Linux. Linux’ opensource cabin or bucket approach to desgn prevents viruses from getting to far as because of being able to see and modularize the code it is much more portable. Allot of .Net is opensource but not all of it hence Mono but I think there are still contracting issues with that as well.
Also OpenGL is not to be supported by Vista and will run allot slower if used which I think is foolish considering that OpenGL also has an opensource element in Mesa which DX doesn’t have. I’m just saying there is a large audience you will loose there for a Vista purchase.
Also Vista will not support anything older then DX10(WGF). Major loss of consumer and hobbyist support there. It will be emulated like Linux currently. They might use a form of Wine but I doubt it.
Native environments don’t matter as when using Interfaces and Generics there is still the issue of timing that does create unpredictable inconstancies in the code that I have even seen myself. This has also been stated in MS’ C# step-by-step. This filters into Interop and then into other stuff as well. Now that’s for hardcore coding. If you like C# it seems to run great for basic UI and maybe even paint.
I haven’t seen the video yet here but they are trying to better the reliability factor with a fully C# OS but it will never be perfect in the timing sense like C++ needed for Scientific experiments and even mature game titles or purist software.
http://channel9.msdn.com/ShowPost.aspx?PostID=68302
Also, Id doesn’t want it because they like real pointers mainly. Plus no major coding guidelines like in C#. You can break up your code more in C++. They do allot of timing things with weaponry and special effects.
Actually I’m not sure if he uses C++ as more C/Assembly.
Also everyone check out the new QT designer which rocks! h files don’t seem to bad as you can track and manipulate your libraries pretty cleanly.
Thanks for the debate.
– After all these years, XP still doesn’t have enough marketshare to make dropping all the others feasible (i.e., its “advanced” features are mostly useless);
– .NET has so close to zero marketshare that it’s a nonstarter for anything but hobbyists and internal apps, and will be at least until it finally ships with Windows;
– Everything I’ve seen about XP SP3 (oops, Vista) suggests that its uptake will be slower than XP’s was;
– Assuming WPF actually ships with Vista, it’s highly unlikely to be usably complete/stable for the first year or two (nothing else ever has been.)
So…
Call me a troll, but will I get to actually ship something using this before I retire or not?
– Dropping the other what?
– Not true. It doesn’t have a very large market share at all, but it’s slowly being adopted. Just go look around the job market, and see how many .NET jobs there are.
– The fact that you called it XP SP3 shows your ignorance on the whole issue
– Why wouldn’t it ship with Vista? It’s probably the biggest thing they’re putting into it, and they can’t afford not to.
Personally, I wish WinAPI was similar to Qt API. With delegates in C# one wouldn’t need MOC. In current state, I find WinAPI extremelly inconvenient to code with. Setting up simple window is a nighmare, but I can do it in few lines of code using Qt.
Does anyone remember WinG ?
Dont much care for the GUI however features like XPS and a bigger better improved colorspace for RAW photograph shooters and having support built right into the OS are quite appealing. I think that belongs in the graphics side of things.
Was looking at some of the server screenshots, I thought the server version would become a serious piece of OS, but WTF are Pictures and Music doing in the start menu!? This is why I still can’t think of Windows as a serious server OS. They just cater to the user crowd too much and don’t enable a server configuration that a real administrator would use. (which is why I often use nlite customized Windows installs)
“The resolution of flat-panel displays has been gradually increasing. 150 dpi displays are commonplace, 200 dpi displays are readily available, and resolutions are likely to continue increasing.”
Unfortunately, the resolutions have stayed the same, with >100dpi (dots per inch) displays hard to come by. Still, resolution independence is crucial, as it will enable us to use high dpi displays effectively, which in turn will lead to availability of high resolution displays.
How I wish I had a 150 dpi display! Font display issues would be soooo much less irritating, and even small SVG icons would be crisp and clear.
Edited 2005-12-23 00:31
Displays with > 100 dpi are quite commonplace these days, in the laptop realm (remember, half of all computers sold are laptops).
People need to realize that most of these “introduction to Vista” articles are written from the perspective of Windows users who know next to nothing about other operating systems. When they say that something is new, it is new to them.
I think that it’s perfectly fine to say that it was implemented in other operating systems earlier than in Windows – because that keeps people from thinking that it was invented by Microsoft – but please do it politely, OK?
And you should be happy when Windows improves because it just gives the other operating systems another mark so that they can leapfrog Windows. Competition is a good thing, so long as it doesn’t try to shut everyone else out of the game…
All Vista does is prettier graphics and everyone knows how bad Microsoft is at that. If you want an advanced OS get a Mac.
Open Source fanboys always will tell you that X or Y feature implemented by MS has been in [insert your favorite Open Source Software/Toolkit/OS/Browser] for years. They may be right, but fail to see that a minimal change in something as used as Windows affects millions of developers and breaks apps. I do not have exact numbers, but I do not think that GTK usage is even near to GDI/GDI+.
And for the record I have developed apps for Qt and GTK and let me tell you that they suck big time, it is the API inconsistencies, the lack of “good” documentation (and don’t tell me the classic “the community supports you” crap) and the dependency hell that sends me back to windows.
I prefer a somehow limited toolkit, that it is consistent and well documented.