I'm interested in using F# for numerical work, but my platforms are not windows based. Mono still has a significant performance penalty for programs that generate a significant amount of short-lived objects (as would be typical for functional languages).
Silverlight is available on OSX. I had seen some reference indicating that assemblies compiled in the usual way could not be referenced, but not clear on the details. I'm not interested in UIs, but wondering whether could use the VM bundled with silverlight effectively for execution?
I would want to be able to reference a large library of numerical models I already have in java (cross-compiled via IKVM to .NET assemblies) and a new codebase written in F#. My hope would be that the silverlight VM on OSX has good performance and can reference external assemblies and native libraries.
Is this doable?
Technically speaking, Silverlight assemblies are similar as normal CLR assemblies, with the exception that they reference different version of runtime (and different version of system libraries such as mscorlib). Silverlight can run only Silverlight assemblies, so if you're compiling F# code you need to instruct the F# compiler to reference Silverlight (here are Visual Studio templates from Luke Hoban and here is a recent sample Silverlight app in F# by Brian McNamara).
Now, regarding the non-F# assemblies, I'm afraid this may be a problem. In principle you don't need to recompile them - there are tools to change the version (to turn CLR assembly into Silverlight assembly). See for example this article. In practice, Silverlight has many limitations (lots of methods missing, you are not allowed to do some Reflection tricks for security reasons etc.). So, I'm afraid that if you simply convert the assembly to Silverlight, it won't really work, but you may still try that... (but be careful - if a referenced method is missing, you won't find this out until Silverlight tries to call it at runtime).
Finally, there is also a problem with communicating with the application running in Silverlight as Silverlight apps have pretty limited capabilities. However, Silverlight 4 RC should allow you to read/write local files when running in the Out-of-browser mode (which may be good enough).
In summary I think that there are a lot of issues that may make it impossible to use Silverlight for this. I would maybe consider doing some more testing on Mono and sending feedback to them (if you find some case where the performance is clearly poor) - From my experience, they can be quite effective in responding to the feedback from users and I have the feeling that F# may be quite interesting thing for Mono team.
Related
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 10 years ago.
I am trying to get my head round the new Windows 8 Runtime that is used to create Metro style apps. I know you can use it with XAML and it is based on .NET so C# and VB.NET can be used to write the apps, but then it seems to have something to do with HTML, CSS, DOM, and JavaScript.
Can someone explain what it is in a few paragraphs, in terms that a .NET UI programmer can understand? (I am missing something “key” that is necessary to understand it.)
We all know that WPF, Silverlight, Windows Forms, etc. will keep working under Windows 8 (and Windows 10) on at least on Intel systems, so please don't tell me that...
At the lowest level, WinRT is an object model defined on ABI level. It uses COM as a base (so every WinRT object implements IUnknown and does refcounting), and builds from there. It does add quite a lot of new concepts in comparison to COM of old, most of which come directly from .NET - for example, WinRT object model has delegates, and events are done .NET-style (with delegates and add/remove subscriber methods, one per event) rather than the old COM model of event sources and sinks. Of other notable things, WinRT also has parametrized ("generic") interfaces.
One other big change is that all WinRT components have metadata available for them, just like .NET assemblies. In COM you kinda sorta had that with typelibs, but not every COM component had them. For WinRT, the metadata is contained in .winmd files - look inside "C:\Program Files (x86)\Windows Kits\8.0\Windows Metadata\" in Developer Preview. If you poke around, you'll see that they are actually CLI assemblies with no code, just metadata tables. You can open them with ILDASM, in fact. Note, this doesn't mean that WinRT itself is managed - it simply reuses the file format.
Then there are a number of libraries implemented in terms of that object model - defining WinRT interfaces and classes. Again, look at "Windows Metadata" folder mentioned above to see what's there; or just fire up Object Browser in VS and select "Windows 8.0" in the framework selector, to see what's covered. There's a lot there, and it doesn't deal with UI alone - you also get namespaces such as Windows.Data.Json, or Windows.Graphics.Printing, or Windows.Networking.Sockets.
Then you get several libraries, which are specifically dealing with UI - mostly these would be various namespaces under Windows.UI or Windows.UI.Xaml. A lot of them are very similar to WPF/Silverlight namespaces - e.g. Windows.UI.Xaml.Controls is closely matching System.Windows.Controls; ditto for Windows.UI.Xaml.Documents etc.
Now, .NET has the ability to directly reference WinRT components as if they were .NET assemblies. This works differently from COM Interop - you don't need any intermediate artifacts such as interop assemblies, you just /r a .winmd file, and all types and their members in its metadata become visible to you as if they were .NET objects. Note that WinRT libraries themselves are fully native (and so native C++ programs that use WinRT do not require CLR at all) - the magic to expose all that stuff as managed is inside the CLR itself, and is fairly low level. If you ildasm a .NET program that references a .winmd, you'll see that it actually looks like an extern assembly reference - there's no sleight of hand trickery such as type embedding there.
It's not a blunt mapping, either - CLR tries to adapt WinRT types to their equivalents, where possible. So e.g. GUIDs, dates and URIs become System.Guid, System.DateTime and System.Uri, respectively; WinRT collection interfaces such as IIterable<T> and IVector<T> become IEnumerable<T> and IList<T>; and so on. This goes both ways - if you have a .NET object that implements IEnumerable<T>, and pass it back to WinRT, it'll see it as IIterable<T>.
Ultimately, what this means is that your .NET Metro apps get access to a subset of the existing standard .NET libraries, and also to (native) WinRT libraries, some of which - particularly Windows.UI - look very similar to Silverlight, API-wise. You still have XAML to define your UI, and you still deal with the same basic concepts as in Silverlight - data bindings, resources, styles, templates etc. In many cases, it is possible to port a Silverlight app simply by using the new namespaces, and tweaking a few places in code where the API was adjusted.
WinRT itself doesn't have anything to do with HTML and CSS, and it bears relation to JavaScript only in a sense that it is also exposed there, similar to how it is done for .NET. You don't need to deal with HTML/CSS/JS when you use WinRT UI libraries in your .NET Metro app (well, I guess, if you really want to, you can host a WebView control...). All your .NET and Silverlight skills remain very much relevant in this programming model.
From the Build keynote:
They're providing common APIs to both HTML/CSS/JavaScript apps and C#/XAML apps. C# and XAML will be used, but it won't be WPF or Silverlight exactly.
The key idea is that now there is two development tracks - the Desktop and Metro.
The desktop is where the old apps live.
The new class of applications, Metro applications, can be built in a number of ways, including by VB.NET, C# or C++. These three language options can use XAML for building the UI. The alternative is to use JavaScript/HTML5/CSS for the development of both the UI and application code.
Some important points:
Windows 8 feels sort of like an upscaled mobile phone OS.
In Metro, there are no overlapping top-level windows, just as there are none on a mobile phone. If you want an MDI style application, you need to stay on the desktop.
Metro style apps are automatically suspended when not visible. This was done to prolong battery life. This means it won't make sense for many existing desktop apps, which perform background processing even while the user is not interacting with them, to be ported to Metro.
The ARM version of Windows 8 will not support desktop applications. So if you want to write an app and you want it to work on any version of Windows then it has to be a Metro app.
There's modified version of the architecture that'll surely help you understand where exactly the things lies. One of the Telerik ninjas had chat with the CLR team and modified the picture:
Here you can see where the CLR stands. The .NET framework now has two profiles
1- .NET Metro profile (CLR that deal with Metro application)
2- .NET Client profile (CLR runtime for C# and VB.NET applications)
I hope this gives you a clearer picture. Read the full article in A bad picture is worth a thousand long discussions..
Lots of detail from Microsoft here.
The Windows Runtime is exposed using API metadata (.winmd files). This is the same format used by the .NET framework (Ecma-335). The underlying binary contract makes it easy for you to access the Windows Runtime APIs directly in the development language of your choice. The shape and structure of the Windows Runtime APIs can be understood by both static languages such as C# and dynamic languages such as JavaScript. IntelliSense is available in JavaScript, C#, Visual Basic, and C++.
In short, Windows Runtime is a new set of libraries exposing Windows functionality and available to JavaScript/C#/VB/C++. Each language has been made to understand and be able to call them directly rather than having to go through some thunking layer.
Silverlight and WPF are flavors of XAML that run on the CLR. Among other functionality, Windows Runtime exposes a version of XAML very similar to Silverlight, but does so in a native way, not via the CLR. It can be accessed from the CLR, but also from C++.
It seems like these two technologies, already similar, are on a path to merge into a single technology. There are a lot more WPF-like controls in the Silverlight toolbox, and WPF now has Silverlight's VisaulStateManager. At this point, it's probably fair to say that Silverlight has even surpassed WPF in terms of the number of themes available.
How long until these two technologies become one? How long until the difference between a rich client app and a rich browser app is a simple compile-time setting?
EDIT
Let me clarify my question. I realize that any browser application needs to run in a "sandbox" for security reasons, and I also understand the desire to keep the browser plugin as small as possible, but there are several minor differences between the two technologies that could probably be massaged out without compromising either of these goals. For example, there could be a lot more overlap between UI controls and themes. Today, you can't just use a Silverlight theme in a WPF app, but how much of a leap would it be for Microsoft to make this possible?
I don't think they'll ever merge into one product. Microsoft has intentionally left a lot out of Silverlight to keep its footprint small. And then there's a plethora of security issues Silverlight must abide by when running in a browser. And of course they've designed it so it'll run on a PC or a Mac (unfortunately the same can't be said for the .NET Framework).
I am happy to see resources shared between WPF and Silverlight though. They were supposed to be similar from the beginning. As a result, it's relatively easy to port a Silverlight project into WPF. On the flip side, it's not quite as easy going from WPF to Silverlight simply because WPF has always had more features, but that's just the nature of the beast.
UPDATE:
So your revised question is interesting. It would be cool if Microsoft could make it possible for you to basically flip a switch to change the behavior of your app between Silverlight-like functionality and WPF. They would be facing a great deal of challenges though, not only with security but with the fundamental behaviors of some of the lightweight Silverlight controls vs the feature-rich WPF controls. These differences could potentially complicate things for the developer even further.
For example, in WPF there's a built-in multiple undo & redo system in the textboxes. In Silverlight there is no such thing so I actually had to write my own. In order for the developer to account for things like that they'd have to do build a lot of feature-checks into the application.
With all that said, I suppose a compile-time switch as you described might be feasible. But I still think it's unlikely Microsoft will create this kind of capability any time soon.
XAML and databinding may become closer between the two
but the rest of the framework will probably never be the same.
For once, you can not automate an Office application using Silverlight.
And that may never happen, unless MS decides to open a bridge of some sort
between the plug-in and the .NET Framework.
Security, consistence, and great UI are main driving forces in Silverlight.
If you can do something in Windows that you cannot do in Mac than
consistence is lost.
Silverlight installations has its own libraries built for specific operating systems. We as developers use what microsoft gives us that can run on those systems. WPF is full trust to the windows operating system which uses specific windows api calls.. So i'd say never will they merge.
Yes we will see that WPF and Silverlight becomes more and more alike, if the acutely will be merge? maybe It's not impossible but what we will see is just that what you stated that they will be more alike. So in the future you will not implement WPF OR Silverlight you will just implement XMAL.
Given the list of new features announced in Silverlight 4, when is WPF still required?
WPF still has a lot fuller 3D rendering capabilities.
Also even though they announced enhancements to the commanding and binding capabilities, that's no guarantee that it will have the full ability that WPF currently has or will have in the next version
Edit: After playing around with Silverlight 4 beta, it looks like pretty much all of the new features (webcam, file access, full screen keyboard, COM interop, etc) only work when the application is elevated trust, and elevated trust can only be enable when running out of browser. Some of this may change by RTW, but for now, WPF still looks like the only way to do these things inside of a browser window
I've been using mostly XBAPs for browser deployments while waiting for a Silverlight version that includes WPF's full templating and data binding capabilities. Perhaps Silverlight 4 will do it.
Several things that Silverlight 4 definitely won't be able to do are:
Allow you to seamlessly integrate your WinForms and MFC user interface components with your application
Efficiently work with data file structures originally defined in C/C++ (with WPF you can simply copy the code across and easily replicate the original C++ code using unsafe and StructLayoutAttribute)
Include unmanaged C++ code in your application
I have not actually downloaded the Silverlight 4 beta yet, so this list is necessarily incomplete.
The next version of Visual Studio will not be written in Silverlight ;) I can't wait to write an Audio utility in WPF---and Silverlight must be sand-boxed away from the incredible Windows Audio subsystem. However, I can see the day when WPF will be called Silverlight (or perhaps the other way around).
When you don't want to depend on a browser to do your work, when you need full and fast access to the file system without prior confirmation, when doing interop with unsafe bits of code. These are the few I can think of, but they might be other reasons as well, depending on what features Silverlight 4 will really have and how well will they really work.
The applications we (in our company) write today use nothing that is not supported by Silverlight except for local TCP/IP connections without limitations.
This is the only reason we are using WPF.
If we could use Silverlight instead (desktop mode) we'd be able to give our customers
the choice of working on Mac (and linux) systems as well instead of being forced (by us?)
to use a recent version of Windows (you'd be surprised how many companies still use old (partially) unsupported versions).
Even in WPF we try to limit access to the system, we create our own sandbox and
try not to go outside of it. (Never require admin rights, never access stuff you don't have to)
So for what we do, Silverlight + real tcp/ip support would be more than enough...
but there is absolutely still room for WPF and I'm pretty sure it's here to stay.
It all depends on what you're building.
Silverlight will always favor small payloads over functionality. Strategically, Microsoft will attempt to out feature Adobe's offerings using Silverlight, while WPF will be competing against other heavy platforms. If you look at WPF 4 you'll notice a push towards heavier weight features. I try to keep in mind that WPF started out as Avalon which was intended to permanently displace Win32. I won't be surprised if one day Windows XX is "native" WPF and Win32 will be emulated.
If you want to do direct database access through ADO.NET that's not something you'll be able to do in Silverlight 4. Access to devices will still be limited although you will now have some printing support and webcam/microphone. I haven't heard anything about 64 bit support so if you want to offer a 64 bit version of your app you'll need WPF. I also think even if Silverlight 4 makes better use of the GPU, it still won't be to the level that WPF utilizes it.
There are also a ton of other assemblies in the .NET framework that Silverlight doesn't include so if you need any of those you'll need to go WPF.
Just like HTML5 will reduce the need for a plugin like Flash or Silverlight in some cases, Silverlight will reduce the need for WPF but there are still plenty of cases where you'll need it.
WPF is still required if you need to do extensive Rich Text (FlowDocument) editing. Silverlight 4 has the new RichTextBox, but it is limited to very basic content, and doesn't provide the full set of features that are available in FlowDocuments in WPF. A FlowDocument can effectively do what the WinForms RichTextBox can do; but if you only need the basics, a RichTextBox in Silveright may do what you need.
The bottom line is that silver light is limiting. So, if you are going to use a less capable technology you should have a compelling reason for doing so. The only one I can come up with, and it is a big one in some circumstances, is that Silverlight is more platform indifferent. I just can't imagine anyone wanting to bring the browser and all of it's headaches into a solution. There must be a good reason.
Silverlight Out of Browser technology and WPF ClickOnce on the surface have similarities. Easy and simple deployment, the ability to specify the level of trust access to the underlying host, etc.
What are the key issues I need to consider when choosing one over the other?
To put a finer point on it, I'll be deploying LOB apps on a corporate network running only windows computers.
The big one is cross platform compatibility. If you need you app to run on a mac as well as windows (not sure if Silverlight is supported in Linux yet) then use Silverligt. If you want to make an assumption that all your users will be in a windows machine then go WPF.
Obviously WPF has a much richer toolkit than silverlight so it may well be that silverlight just isn't an option. If I was just building for windows though I know my job would be easier in WPF.
Given that you are targetting a private infrastructure running Windows, two points worth thinking about
Wpf has a richer control tree, whereas Silverlight is a reduced set for compact size
Wpf requires .Net framework installed locally, whereas Silverlight has its own platform independent browser-based runtime
While your target platform will likely have the latest .Net framework installed, rendering this last point moot, keep in mind any updates to the framework [ie .Net4.0 and any future updates] may require a restart of the machine - which is a major pain point for businesses that demand constant-on stateful desktops [ie anything in finance, like banks and trading].
As with all problems, your requirements, not the technology, should inform your solution. :)
You mentioned trust access to the host which I think rules out Silverlight unless you want to run SL4 (beta).
We recently went through a lot of discussion about file system access. Silverlight 3 runs in a partial trust sandbox more or less. You can't maintain a pointer to files in the files system outside of your application's isolated storage. This was an issue for us as we wanted the user to be able to use the application to reference odds and ends on your file system. That said you can allow the users to load and save files from anywhere on the system but you just get/or push the file stream and (to the best of my knowledge) don't have access to the folder or file path information.
Silverlight 4 (in beta) has support for your application running in full trust mode. I haven't played with this yet however and can't speak to how well it works.
In talking with a lot of people who work with both Silverlight and WPF, even those who are excited about Silverlight and push for it strongly, I hear a lot of the say fairly emphatically that if you are going to be developing exclusively for a full-trust Windows environment, WPF is hands-down the obvious choice.
That's not to say that Silverlight is an inferior product or that there aren't times will Silverlight will be the clear winner. But when you say "I'll be deploying LOB apps on a corporate network running only windows computers," it sounds like WPF is the clear winnder.
You could decide to go down the Silverlight route in anticipation of all of the great new OOB feature os SL4. I've even heard rumors that SL and WPF will eventually merge, so it may not even really matter, right? Well, I think what you'll find if you go with Silverlight is that some of the advanced features that you thought were there weren't there in the way you expected. For example, SL4 will be able to run in "Elevated Trust" (not full trust) and you might find this limiting at a frustrating point in the project where a lot of your code base is already in Silverlight.
Certainly keep your eyes on Silverlight, but for your current business case, WPF will likely be the best fit.
I am fairly new with silverlight and I really find it cool. I have a question about how it runs the code client-side tho..
Say for example, I have a site that calculates a certain amount based on user inputted amounts. This of course I would love to do client-side. The catch though, is that the formula used for the calculation is proprietary and a trade secret. If I put this formula client-side using SL, will it be safe? Or can it be reflected?
If you want to keep algorithms secret, don't push it to the client side. No form of obfuscation or protection is ever perfect.
Also, when you have calculations on the client side, you should always check the results on the server, rather than just assuming they're correct. Assume that the client is compromised.
Silverlight pushes the XAP file to the client. The XAP file is simply a zip file containing your .NET assemblies, which can then be unzipped and reflected against. The company I work for (PreEmptive Solutions) markets Dotfuscator, which can obfuscate Silverlight assemblies. Right now you have to unzip the xap, obfuscate and zip them back in, but we're working on improving the workflow.
Just a note to Dotfuscator users: If you create a Dotfuscator project, you must use the "User Defined Assembly Load Path" property in the "Settings" tab to browse to the Silverlight libs. The paths you need are:
\Program Files\Microsoft SDKs\Silverlight\v2.0\Reference Assemblies
\Program Files\Microsoft SDKs\Silverlight\v2.0\Libraries\Client
or on 64 bit operating systems:
\Program Files (x86)\Microsoft SDKs\Silverlight\v2.0\Reference Assemblies
\Program Files (x86)\Microsoft SDKs\Silverlight\v2.0\Libraries\Client
Don't fall into a trap of think hiding the algorithm will protect it. Once you put it on the web somebody will figure it out no matter what you do. With enough sample data anybody with some math skills should be able to figure out your algorithm.
All you can do is make it harder. If this algorithm is is something proprietary that you have bought then it will need to be server side. Putting the algorithm on the client side is essentially publishing it and you could be liable.
IntelliLock and .NET Reactor (my preferred tool) obfuscates my assemblies nicely.
While obfuscation is not a fool-proof method, it makes it that much more difficult for somebody to see your code. One has to really jump though convoluted hoops to get to your final code if the layers of obfuscation are good. Crypto Obfuscator is one obfuscator which supports obfuscation of Silverlight assemblies.
Another cool tool is CodeFort. It has free edition. See it in action at http://www.codefort.org
CodeFort .NET & Silverlight Obfuscator
CodeFort is an advanced obfuscator and protection tool for Microsoft .NET and Silverlight applications.
BAML and XAML obfuscator - obfuscate 100% of your code
CodeFort is the first tool ever to be able to obfuscate identifiers inside the XAML and BAML code which is used in Silverlight and WPF applications. This makes it for the first time possible to obfuscate 100% of your code.
Powerful protection against attackers
Coupling the XAML/BAML obfuscation with powerful protection features such as Reference Scrambling and Anti-Tampering CodeFort is a state-of-the-art obfuscating tool.
I must COMPLETELY agree with Marcus. Even obstruficated .NET assembly is still easy to read for a good programmer.
My solution would be WCF service for the calculation. Just push all the data there and give an answer. If your formula is top secret and not obvious (like ax+by+c*z) then even is somebody would get access to service, then it would be hard for him to get it.
There are many companies that support obfuscating Silverlight 2.0 applications. DeepSea Obfuscator has a nicely integrated experience, Dotfuscator also work and soon, the free Eazfuscator will also support it.