Talk:.NET Framework/Archive 2

BCLTeam's WebLog link relocation
I propose to move the BCLTeam's WebLog link under Other to Microsoft External links since the BCL team are all Microsoft employees. Comments anyone?

--CSharp-Online.NET Editor 20:54, 7 February 2007 (UTC)

codeproject.com link removal
I propose to remove codeproject.com from the Other links. It has 110 links on Wikipedia raising the issues of advertising, promotion, spamming, and just plain over-linking. Comments anyone?

--CSharp-Online.NET Editor 20:51, 7 February 2007 (UTC)

Okay, I posted the above for comments. Got none. Deleted the link. And User:Futurix restores it saying: "Stop the personal vendetta". I was trying to follow procedure and I get slammed again. What's up?

--CSharp-Online.NET Editor 09:35, 14 February 2007 (UTC)


 * It appears that your proposal to remove codeproject.com is a result of you having your link removed. That may not be the case, however it appears to be a bad faith suggestion. Your coments here and edit summary here. Just because it has 110 links, does not necessarily mean it was spammed. there are some links on Wikipedia that are in the thousands, however they are WP:RS and were placed in good faith over a long period of time by seasoned editors. I assume you're here to improve Wikipedia, not to promote your site. based on your contributions it does seem you may have an adgenda other than to improve the encyclopedia. I'll make a friendly suggestion that you stop canvasing for removal of a compeating site--Hu12 10:23, 14 February 2007 (UTC)

isn't it just peachy when someone supposed to be writing an encyclopedia can't spell agenda? — [ Unsigned .]
 * Just as peachy as someone who can't use capital letters or sign their comments... :) -Localzuk(talk) 19:06, 22 July 2007 (UTC)

new question
What is the native language Windows Vista speaks? Is every call to the Win32 API translated to .NET or does .NET wrap its calls to Win32 API? What's the system's base layer? I can't imagine that .NET can ever be a system's native language because it's too far away from the machine. Isn't this the crucial problem with .NET (as well as with Java)? -RumBug 217.50.135.142 16:22, 7 January 2007 (UTC)


 * As far as I know, the kernel is all natively compiled (from C/ASM I think). I've heard rumours that some parts of the userland code are written in .NET. --Stestagg 16:22, 15 August 2007 (UTC)


 * As far as I understand it Vista contains no .NET code other than the framework which comes installed with it. Vista is C with C++. RobChafer 00:07, 4 October 2007 (UTC)

Drop .net vs com
I think we should drop the .net vs com section. It has very little content, and it isn't even a valid comparison really. COM was not replaced by .net wholesale, but by several technoles, including web services, remoting, assemblies, WCF/Indigo, etc. If it isn't dropped, it needs to be radically expanded to encompass explaining all of these other technologies, which I don't think is appropriate for this article. Gaijin42 18:22, 23 June 2006 (UTC)


 * I agree - it should be dropped. Leotohill 19:18, 23 June 2006 (UTC)


 * I strongly agree, too. And in fact, COM was never dropped, only extended. You still have to know your IUnknown if you want to code certain things in native WinAPI. After all, I don't see how .NET can be compared to COM in any aspect. COM is a quite refined technique of object programming while .NET is an over-sized virtual machine designed for people who don't have the time to know the operating system (or who hope that .NET will are really be platform-independent). RumBug 217.50.135.142 16:35, 7 January 2007 (UTC)


 * I agree, .NET relies heavily on COM for its implimentation (on Windows anyway). RobChafer 00:03, 4 October 2007 (UTC)

Criticism (to be continued)
A little comment from an average home pc user novice when it comes to .net. I read this article trying to find out if .NET was something I wanted to download for my home pc or not. Thats not easy to decide upon at all! It should have a few sentences in the heading or beginning of the articles pointing out the real purpose of .NET in a short and easy to understand language. As far as I could tell, it's more of a programming tool than anything else. But why is it downloadable for every single home pc user on windows update? What exactly will it do for the non programming novice home user? Is it a JAVA substitute of some kind? Or Java competitor?

Thats valuable information for everyone who's trying to decide to download this or not, much more so than the technical in depth aspects. --129.241.125.61 02:51, 20 January 2007 (UTC)

I don't know why the fact that .NET is tied to Windows platform must not be put in the criticism section, considering that this is a critic some people do to the framework : see here for example : and here : "I believe the truth is that Microsoft is schizophrenic on the issue. It is serious about standardising C# and the CLI, and it’s edging towards a Unix implementation and perhaps a commercial OSX release, who knows. On the other hand, there will certainly be other Microsoft insiders who will want to keep .Net Windows only."

"The .Net Framework is outstanding technology, but Microsoft must let it grow beyond Windows if it is to fulfil its long-term potential." Hervegirod 10:58, 15 July 2006 (UTC)


 * That's not criticism... that's editorialising about what Microsoft should do. -/- Warren 14:59, 15 July 2006 (UTC)
 * OK, here's another : "Unfortunately, too much of the .Net Framework is tied to the Windows API to make anything that is closely tied to .Net really cross-platform compatable. Yes, Mono can compile and run straight C# code, but is C# a good language without the .Net Framework? I really do not think so.", and here : "Nonetheless, analysts contend that .NET is too closely tied to Windows. "There's nothing in .NET that makes it more attractive to legacy integration, other than Web services, than the old COM technology," says Gartner's Driver.", or here : "Being tied to the Windows platform for runtime deployment is a quite severe limitation that prevents adoption of .NET in some quarters." or here: "it's still all too easy to write a .NET app that is tied to Windows and won't run properly on Mono".
 * although "criticism" may not be the best description for this issue, I think it's appropriate to make the point in the article. It's like the complaint that the IPOD doesn't handle other DRM schemes, or that Blue_ray devices don't play hd-dvd, and vice-versa.  It's not like it's a fault of the product, its just a business decision, and it's ok to point out that the restriction limits some uses of the product that the end-user would like to have.  Leotohill 15:08, 16 July 2006 (UTC)
 * I added a "criticism" on this matter, what do you think ? Hervegirod 09:27, 23 July 2006 (UTC)
 * I think it's ok, but I'd try to rephrase "too much tied to" into something a little more formal. Leotohill 15:57, 23 July 2006 (UTC)\


 * I don't see the Windows-only aspect of .NET as something that needs critizing Before .NET, Microsoft developers had a huge problem consisting of about 36 different Windows API versions, and developers had to constantly test which version when calling any Windows library.  It was a nightmare, and .NET solved it.  I think that was extremely important for Microsoft, and the need for it vastly overshadowed a desire to support other non-Microsoft platforms.  Of course, that is just my speculation as an outsider (I don't work for Microsoft), and it's probably not helpful in the article.  Also, I've done enough cross-platform development in Java to know that that is not painless.  Microsoft succeeded in making all Windows alike and achieving multiple language support.  Multiple OS support would be nice, but to me, it's lack does not detract from the success of what has already been done via .NET. Harborsparrow 16:54, 8 October 2006 (UTC)

This is a great article
I don't know how to nominate something as "a good article," but this one deserves it. Aside from it being in depth, covering a lot of true info, and all of that, it's also conscice for the amount of material it covers. In fact it's taken Microsoft a book to explain the information on this page. Great job, people - all of you deserve a pat on the back. DigitalEnthusiast 21:56, 19 December 2006 (UTC)


 * It's a good article but it doesn't provide all information which is necessary. I still don't know if .NET is of any importance to the core (or the kernel) of Windows. Is it something like Java and therefore exterior to the system, or is it something like COM which is a programming technique? There is still too much Microsoft proapaganda in this article. I have a high opininion of some of the things Microsoft did (like COM and IDispatch) but I do miss a lot of information about .NET. -RumBug 217.50.135.142 16:40, 7 January 2007 (UTC)
 * Since Windows does not require .NET to run, it cannot be considered core or kernel. It (the .NET framework) is an optional component of the operating system.  Whether it is "part of" the operating system or not is arguable, but IMO it's not a very interesting argument.  The article provides a good comparison of the .NET framework to Java and the J2EE:  in architecture and intent they are very similar.
 * If you can point out something specific that sounds like MS propaganda, perhaps it will be addressed.
 * Leotohill 03:01, 18 January 2007 (UTC)


 * Section "Significance" is basically marketing-speak rephrase of "Design goals and principal features" section. Futurix 10:37, 18 January 2007 (UTC)
 * I may agree w/ you. I also notice, as in many articles though, that the usage of references and sources is very limited. --  Szvest   -  Wiki me up ®  10:58, 18 January 2007 (UTC)
 * FYI guys, the articles has been peer reviewed months ago (see Peer review/.NET Framework/archive1) and the issue of references and citations has been mentioned and little has been done so far. --  Szvest   -  Wiki me up ®  11:02, 18 January 2007 (UTC)
 * I see some phrases that might arguably be NPOV or marketing-speak, but so far I'd argue that they are acceptable. For example, the "most significant" claim may seem objectionable to some, but I think that it is valid.  The great majority of MS's developer's tools now involve the .NET framework, and MS's own products use it. It would be helpful to have some citations.
 * I think that "significance" provides more detail, allowing the introductory paragraphs to be shorter.
 * Leotohill 01:18, 19 January 2007 (UTC)
 * Leotohill 01:18, 19 January 2007 (UTC)

Actions needed
I've just posted the article at the Article Creation and Improvement Drive. Please add your votes there. I've also created the to-do list here Talk:.NET Framework/to do which you can update and edit. --  Szvest   -  Wiki me up ®  11:13, 18 January 2007 (UTC)

Our major complaint about .NET: Deployment
Our major complaint about .NET is that it is not as simple to deploy as traditional applications, because the user must have Administrator privileges to install the runtime. This is a problem in controlled environments. On the other hand, non-.NET applications can be installed entirely in the user space. I imagine this difficulty in deployment must be a factor for other developers, too, and if so it should probably be discussed in the article. --Steven Fisher 22:04, 17 January 2007 (UTC)


 * .NET is a system-wide component (just like DirectX, Internet Explorer, MS XML, etc), so it requires write access to system folders. I don't see how this issue is worthy of discussion. Futurix 10:34, 18 January 2007 (UTC)


 * It's because even now, years later, Microsoft has not pushed this update out to users. This makes it a severe PITA to deploy an application requiring .NET Framework in locked down environments where an admin is not readily available. Compare this to non-.NET applications and how hard Microsoft has pushed to make C# seem like first class Windows applications, and you can see that it's largely unique among "first class" application development techniques. For instance, C++ applications require updated runtime libraries. No problem: Just include the runtime libraries as a private assembly. There's no need for admin access. In comparison, installing the first C# application on a computer will always require admin access. Of course it is worthy of discussion. --Steven Fisher 23:26, 31 January 2007 (UTC)


 * This is not a real problem - home users are always admins, and office users have admin who will install Framework for them. Not to mention that it is the same situation as Java, DirectX, Flash, and many others.
 * And finally - .NET 2.0/3.0 is included in Vista. Futurix 00:15, 1 February 2007 (UTC)


 * Judging from my experience, that can be a problem. It is possible to use a full Java developement environment (not only the runtime) just by copying the Files on disk (and it is not necessary for the files to be installed locally on the PC, providing that the files are accessible on the network) and calling java (for the runtime) or Javac (for compiler). I use Netbeans at work with such an "install" of Java and Netbeans itself (as I only have a JRE pre-installed on the PC), and it works fine. Our admins have installed .NET and Java (and of course not the last versions, I think 1.4 for Java and 1.1 for .NET), but only the runtime (and they won't install the full development kit, unless you have a special account and a specifically configured PC, which is not easy to obtain). So I think that it would not have been possible for me to do the same with .NET as I am doing with Java. Hervegirod 21:31, 1 February 2007 (UTC)


 * All right, Java is not a good example. But DirectX and Flash still are. Futurix 21:58, 1 February 2007 (UTC)


 * Perhaps relevant for this discussion is the latest addition to a related article. Obviously it is advantageous to not require admin rights, but I don't see this as a big problem either (let alone a "major complaint", any references?). Indeed home users are already admins, and often already installed .NET through updates, and office users have an admin who can install it for them. And finally Vista users get .NET combined with Windows. If this criticism is to be listed here, those points should be mentioned too, and it should be synchronized with the comparison article. Mfb52 23:00, 1 February 2007 (UTC)


 * It was exactly same thing, that same user pushed into different article. I removed it, as it is mostly POV and personal problem of Hervegirod.
 * As both me and Mfb52 wrote - this is not a problem for majority of users, both home and office. The only affected areas are spyware/adware (software that installs without user's consent) and various application that users try to install without administrator's consent - not something important or occurring often.
 * Even more important is fact that both Java and .NET can run without installation - unofficially (I'm not going to advertise those hacks in here though - some of them are commercial).
 * Futurix 01:48, 2 February 2007 (UTC)


 * Many environments do not have admin staff easily available. Schools are a big one; we've had enough problems with this that we're moving out of .NET. If this really isn't an issue for anyone else, fine. But if it is, I'd like to see it in the article, arguments that it shouldn't be an issue aside. 70.71.224.200 22:45, 9 May 2007 (UTC)

Patents
Recently, I put this in the lead of the article: ''Microsoft has applied for patents on some parts of the Framеwork. '' This was moved to the Criticism section. The following was added by Mfb52: ''An agreement was made however between Microsoft and Novell whereby Microsoft agreed to not sue Novell or its customers for patent infringement, allowing the implementation of these parts in the open-source Mono implementation of .NET. ''

A couple points. First, the fact that Microsoft has applied patents for some parts of .NET is not, by itself, criticism. Admittedly, the Cnet article I used as a reference has a critical spin--perhaps someone can find another news article on the topic. But still, Microsoft has deliberately chosen to apply for patents on .NET. This is an important fact about the technology and I believe it deserves mention in a prominent place in the article, not just in the criticism section. Many companies brag about their patents, as Steve Jobs did with the iPhone.

Second, the conclusion regarding the impact of the Novell-Microsoft agreement on Mono is not supported by the references given. The Novell press release doesn't mention Mono or .NET at all. The FAQ from the Mono project, significantly, alludes to the possibility that their may be a finding of infringement even with the Novell-Microsoft agreement. ("With the new Novell/Microsoft agreement, will the patent policy change? ...we will continue to follow the steps outlined in the previous topic if code that potentially infringes is found: finding prior art, finding different implementation techniques, or if none of those are possible, removing the code from Mono.") So this sentence needs to be revised or removed. Numberp 03:47, 4 February 2007 (UTC)
 * I try to add some facts about the Microsoft / Novell deal, in the Criticism section. Hervegirod 11:15, 4 February 2007 (UTC)


 * Herv, I think that your text (if it is all yours) doesn't accurately reflect the content of the references.  It seems to me that there are some parts of the framework (e.g., ADO.Net) that are covered by patent but not by the agreement.  Also, the second and third sentences are in contradiction:  "Microsoft agreed not to sue Novell..." vs. "this agreement extends to Mono but only for Novell developers and users."  In other words, it does not cover Novell the company.  The original press releases stated clearly that MS was agreeing not to sue individuals on these matters, but did not say that it wouldn't sue companies.  Finally, I think that it is easy to confuse the parts of the agreement that cover Novell's SuSe Linux with the parts that cover .NET. I think that on the former, MS did actually agree not to sue the company.


 * And finally finally, what's the difference if MS's product is covered by patent vs. copyright? Either way, MS could sue, right? I ask this question sincerely - I don't know.
 * Leotohill 14:28, 14 February 2007 (UTC)


 * Well, copyright just means you can't go and look at their source code and copy/steal it. You can still do a clean room implementation however. With sofware patents, almost any program you write could be breaking patents you never even heard of. (And most programs probably are, seeing how trivial most software patents are.) Luckily, patents are not very actively enforced, so you only get sued if you have some cash (or if they dislike you). As for this case, I think it's simply in Microsoft's best interest here not to sue Mono/Novell. But if that ever changes, you never know what happens next, although the agreement should prevent them from taking legal actions. - Mfb52 15:09, 15 March 2007 (UTC)

.net/.net framework - for users or developers
I guess since I'm not a developer it's ok that it's not clear if this set of programs is something a "normal" user of windows needs to have loaded on their computer.

I believe the answer is no I don't need it loaded as an average user but it would be nice if somehow this was stated early on in this article.

my 2cents —The preceding unsigned comment was added by 64.58.31.86 (talk) 18:07, 18 February 2007 (UTC).


 * The answer is that any program that is written to use the .NET Framework requires that the Framework is installed on any computer that runs the program. So your question could be rephrased as "does the average user of windows run any .NET application?".  That's hard for me to answer, since I don't have any numbers to go by, but my guess is that "many" Windows users do.  Furthermore, Windows Vista not only includes the Framework, I suspect that it requires it to run some of its commonly used built-in applications.  So, we could probably say that all users of Vista require the Framework.


 * I agree, the article should make this more clear, but more concisely than I have done. Leotohill 01:19, 20 February 2007 (UTC)

Application Architecture for .NETApplications and Services
I have been given this as a project to reseach (Application Architecture for .NETApplications and Services focusing on Security, Operational Management and Communications Policies) Please help me out any one.

Tshepiso Mogoswane mrmogoswane@yahoo.com —The preceding unsigned comment was added by 41.244.209.221 (talk) 08:53, 28 February 2007 (UTC).

Release date
.NET 1.0 - the release date is different in the text and in the table - it probably needs fixing.  mfx  Q&A 13:03, 1 March 2007 (UTC)

Serious POV problems
This whole article reads like a commercial from Microsoft.

Seriously, how about: "The creation of the .NET Framework reflects what has been learned over the years about the typical challenges that programmers face and useful approaches to those challenges."

I suggest the article keeps the POV tag, especially since others has criticised its neutrality on this talk page.

130.225.96.2 17:49, 20 March 2007 (UTC)


 * I agree that that particular sentence needs revision. I think that it is largely correct, but the phrasing is not right for an encyclopedia.  I wrestled with it for a few minutes without success - I'll try again later, unless somebody else changes it first.  Leotohill 18:48, 22 March 2007 (UTC)


 * I added POV tags on the two sections that may be need revision on this matter. Hervegirod 14:22, 25 March 2007 (UTC)


 * Why not add a modifying word somewhere to get it into a more NPOV ? for ex. "is an attempt to reflect..." ? Vijeth 12:29, 26 March 2007 (UTC)


 * Good point. I heard someone quip recently that .NET is the latest framework MS uses to write bugs.


 * Every time MS tries to improve something someone has to go and put it down, if the latest .NET apps don't run on your Mac then write about that, don't start making up crap. Personally I find .NET combines the RAD elements of Java with the speed of C++, and don't care if it won't work on some minority OS. LaudanumCoda


 * We put it down because its proprietary rubbish that displays no originality, reinventing the wheel a hundred times, and because it displays architectural imagination and pinash that the average first-year C programmer would grinningly give to their teacher, and because it mainly exists to cover up the whole DOS TSR mess that evolved into their current 'product' range. I think they are pretty good reasons for a put down, really.

Directory
Someone might want to find a place to mention that the files for each framework are found in \Microsoft.NET\Framework\v (ie, ...\v2.0.50727). Stevage 01:47, 27 March 2007 (UTC)

Mobile
It appears the .NET Framework doesn't support Mobile at all? I'm not speaking from practice, but rather wikipedia would lead one to believe such. I was trying to setup .NET_Mobile_Profile but I can't find anything. Is it the current Microsoft Mobile Technology? This article only had this to say on the subject:

" Changes since 1.0

* Built-in support for mobile ASP.NET controls. Previously available as an add-on for .NET Framework, now part of the framework. " Mathiastck 06:03, 24 May 2007 (UTC)

Java vs .NET (I'm totally confused)
When Sun started work on Java, their primary goal was platform independance. Write a piece of software once, publish it to the internet, then have it execute on the Java-VM of the target computer. It didn't matter if the target platform was a minicomputer, PC or Macintosh because it was all the same unified world when viewed from the internet.

Has Microsoft ever stated the purpose of their VM technology? As far as I can tell .NET only runs on Wintel (Windows on Intel) and Microsoft has never really gone out of their way to get it to run on any other platform. What is the payoff for the Microsoft? How will this contribute to their bottom line? --Neilrieck 10:18, 18 June 2007 (UTC)


 * The CLR engine is described by a ECMA/ISO specification and there are alternative implementations, including Microsoft's Shared Source Common Language Infrastructure (for XP, FreeBSD, and Mac OS X 10.2), DotGNU's Portable.NET, and Novell's Mono. See .NET Framework.  Rod (A. Smith) 17:23, 18 June 2007 (UTC)


 * I find comparissons between the two rather pointless - .NETs targetted at generic RAD rather than at web application development, so has different focusses:
 * - Faster loading and running
 * - Designed to support several programming languages providing fully-functional IDEs - not just quick 3rd party interops
 * - The default .NET setup is also much more simpler (lexically) than the Java equivalent (a ListBox is just a ListBox, not a JList within a scroll pane, etc).
 * LaudanumCoda


 * While it is true that Java can be used for web development (as in the case of a Java Applet or Servlet), Java can and does get used in stand-alone applications. But onto a different point: I've talked to many developers and come away with an answer to my original question: Java requires that you learn a new language (Java) while .NET allows you to use other computer languages. (although the people who had to convert from VB6 to VB.NET found fault with the "no need to learn" argument). Now some people have told me that C# (C-sharpe) is the only way to make full use of .NET and that it is very similar to Java when solutions are compared line-by-line.

--Neilrieck 22:06, 29 June 2007 (UTC)

Mention of "Rotor" Project
I am missing mention of Microsoft's "Rotor" project in the article. Rotor is a shared source implementation of .NET released by Microsoft themselfes that compiles on Windows and on FreeBSD-compatible systems.

http://research.microsoft.com/programs/europe/rotor/

The code can be freely downloaded (without signing an NDA) and you are allowed to use it for learning on how to create your own implementation of an ECMA-Standards compliant .NET compiler + runtime. To my knowledge, this has been available even before the Mono project was launched.

In this context, the following phrase in the article seems wrong:
 * Microsoft's reference implementation of .NET is closed source, whereas Sun's reference implementation of Java is becoming open source (including the class library, the compiler, the virtual machine, and the various tools associated with the Java Platform).

85.25.133.66 08:04, 9 July 2007 (UTC)

Windows Live ID
Does Net framework connect to "Live" and Windows Live ID, and how? David1776 12:58, 23 August 2007 (UTC)

cryptographic guarantee?
The article currently says, "Two assemblies with the same public key token are guaranteed to be identical."

I don't know what specific techniques are being used with assemblies, so that claim may be true.

However, cryptographic hashes typically don't guarantee that distinct inputs result in distinct hashes. Rather they have the property that it is computationally hard to find another input that results in the same hash as a given input. This implies that it is in some sense rare for distinct inputs to result in the same hash, but NOT that it is guaranteed not to happen. (In particular, if the space of tokens were finite and the space of assemblies were larger, then there would necessarily be collisions.) Is that a more accurate description of what is going on with these assembly tokens?

MarkAb 02:15, 1 September 2007 (UTC)


 * I was thinking that myself; to guarantee they're identical would take a cryptographic hash the same size as the file. I don't know the full details here either, but perhaps it should be changed to something like 'Creating two different files with the same hash is computationally infeasible.' Mark Grant 23:22, 1 November 2007 (UTC)

.NET Future features
Here are articles about upcoming features, indexed here for reference when someone will want to update the article: PLINQ, TPL. --soum talk 15:20, 13 September 2007 (UTC)