Re: [Mono-dev] Updated CentOS/RHEL packages?
Hi Gentlemen, I share your need for RHEL packages and look forward to trying out your repo. If I can help, I will. Many thanks, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Adding privatekey support to X509Store
Hi Pablo, Your interest in improving Mono's CryptoServiceProvider offerings is very exciting. Your experience seems to be leading you towards an OpenSSL binding of some kind. May I suggest that you at least review other cryptographic providers like Mozilla's Network Security Services (NSS) or the Free Software Foundation's GnuTLS as well. I lack your experience so will defer to your better judgment but I known that Linux Standard Base 4.0 [1], Red Hat and Fedora are actively trying to reduce the number their use of OpenSSL and converge on NSS [2]. [1] http://ldn.linuxfoundation.org/article/lsb-40-the-cryptography-strategy [2] http://fedoraproject.org/wiki/FedoraCryptoConsolidation Cheers, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Latest packages
Hi, Has anyone build gtk-sharp-2 and the others (art-sharp, glib-sharp, gnome-sharp, etc) for RHEL/CentOS using the new (2.6.7) /opt/novell/mono based packages? I am the opposite of a gentoo-ish build everything from source guy. I like packages and I happen to be in the unfortunate position of needing Red Hat and loving mono. 1. Is there any chance that Novell will build these packages as well? If you only did it for the long term supported versions that would be fantastic. 2. If this is a question of motivation, could this perhaps be a sponsored effort (note I am not in a position to make promises, but I want to know if it is possible)? 3. If the above is not possible, I imagine I have to use http://www.mono-project.com/Parallel_Mono_Environments to get started. The only problem is that I am not looking to distribute my software as source. Is there a make dist as opposed to make install that will build packages using the prefix given to the ./configure command? Yes, I am that clueless. I apologize. Thank you for your help, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] C# Binding for elektra
Hi, Is someone planning to write a .Net binding to Elektra (1)? Like GConf, Elektra is a place to put configuration data. It is different in that it would be accessible even when you want to run without X and Gtk (and the many other GConf dependencies). I need this for some buildbot testing of my app. Assuming no one is looking at this, is the general recommendation to 1. Do it by hand? 2. Use SWIG to do everything? 3. Use SWIG some special way? 4. Other? Are there any tutorials/howtos or other references to write a good binding? Thanks, Vlad (1) http://elektra.g4ii.com/Main_Page ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Mono Regex not working the same as MS .NET
oh man, am i ever embarrassed... you are right on. :blush: with that change, the code works as expected. Thank you! --Dan Gentry I would suspect the \r\n part of the pattern. On UNIX systems new line does not include the \r. --- Adar Wesley You could also try using System.Environment.NewLine. I don't know if MoMA checks for it, but it is a good thing like Path.Combine() for cross platform code. Cheers, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Beyond C#
2009/7/26 Scott Thomas: That's my piece. I think we should do it. Respond! I think this is a good idea. It's about time we embrace and extend (TM) C#. -- Seo Sanghyeon Ditto. Can't we just get this in through ECMA and Microsoft would just have to go along? ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Gtk# 2.x release status and thoughts on 3.0
With the gtk+ 3.0 release planned in the next year, we have therefore decided to hold the 2.x release series of Gtk# at the 2.12 API version, and we plan to switch trunk from its current 2.14 target over to 3.0 when the gtk+ team gets closer to releasing. Is there any chance of migrating to a .Net 2.0 code base? Generics alone should be reason enough for moving to a 2.0 only codebase. Depending on the status/licensing of the ParallelFx stuff and the co/contra variant mess, it might even make sense to discuss the possibility of a newer dependency like 4.0. Can you make a call for all Mono consumers that want Gtk# to stay on the 1.1 profile so we can see what the obstacles are? Thanks, Vlad smime.p7s Description: S/MIME cryptographic signature ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Installers
From: Mike Kestner Feedback appreciated. In particular I'd like to hear from those of you using the current Gtk# for .net installer or those considering win32 ports in the future. Which would you prefer, one installer or two? If one, should it remain the Gtk# for .net installer for historical purposes, even containing the infusion of mono tastiness, or do we rename it to Mono Libraries for .Net? Hi Mike, I am not familiar with packaging on Windows yet, but I am willing to make a fool of myself. My understanding is that there is a Windows Installer XML toolset (WiX) which is open source and is the nice way to package things. It lets developers build MSIs and MSMs. Since it is really easy for me to volunteer you to do some work, I will suggest that you should help Mono consumers (developers like me) package their apps. The problem I will want to solve is actually how to package my application --not how my customers install Gtk#. Obviously, I would love for my users to have all my dependencies or to go get them themselves, but that is not reality. Can you provide a WiX template that we could follow. The tooling for automatically generating everything with pretty highlighting and such can come later ;). If you show how banshee or some other app that runs on windows is packaged using WiX, I would guess that I would not care how and where Mono.Addins is put. I could add a line or whatever to some xml file for my package that goes to get it. Generally, I would want to be able to pick and choose as finely as possible how much gets into my app. I realize I could use the linker but that is overkill. You could always have packages that lump a bunch of stuff together. I think it makes more sense for me to do the lumping together rather than you (the Mono/Gtk# packager). Thanks, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Red Hat/Fedora packaging
The following was taken from
https://fedoraproject.org/wiki/Packaging/Mono
File Locations
Mono packages should install assemblies to %{_libdir} rather than
/usr/lib or %{_datadir}. In most cases the preference is for
%{_libdir}/PACKAGENAME. We use %{_libdir} because we do not consider
mono packages to be noarch.
The main reason for this is that mono can ahead-of-time compile its
assemblies into ELF shared objects. These AOTs have to exist in the same
directory as their DLL/EXE counterparts otherwise mono cannot use them.
Even if we, as packagers, choose not to create the AOT files when we
build the mono rpms, the system administrator can choose to create them
after install. Since there's no way to place the mono assemblies into an
arch independent directory and the AOTs into arch dependent directories,
the whole thing has to go into an arch dependent tree.
It really bothers me that Red Hat/Fedora considers CIL packages not to
be noarch. I would like to propose using the config file to allow
relocating the aot-ed file to some arch dependent location. The problem
they have is that the aot-ed code must be side-by-side with the pre
compiled assembly.
Could we use Foo.dll.config to have sections per architecture that tell
the runtime where to load the aot-ed file from?
I don't actually use aot-ed code. I just want to get my package (exe)
to be noarch. Using Red Hat is not my choice so please don't suggest
that I use another distribution. Please help me get these packaging
guidelines changed.
Thanks,
Vlad
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Red Hat/Fedora packaging
Hi, 1. If today's assemblies are noarch and tomorrow's are not, why not conform to noarch guidelines today and conform to arch dependent guidelines when the time comes? 2. The problem is broader than Mono's assemblies. My original question is based on ndesk-dbus. It is a fine managed only implementation (not only for Mono). 3. (Out of my depth) Does the CLI not stipulate that variables can be resized and moved around depending on architecture implying the CIL to be architecture agnostic? The noarch difficulty: * According to the FHS, if CIL files are architecture independent they should live under /usr/share/ instead of under /usr/lib. Even if today Mono's assemblies are noarch, we have always reserved the right to make them arch specific if needed. Miguel. If this is not write once run anywhere but write once package everywhere, I will need to go sharpen my pitch fork and get oil for some torches. I am in essence being forced to make my package architecture dependant. Since you have already dropped support for anything other than SuSE it may be OK for you, but my software needs to run on Red Hat. I am looking for a solution to the aot exception*. I will let the other distribution packagers complain about how Mono packages its own software. Aot-ed binary must be placed side-by-side with dll/exe assembly. This is a packaging problem if one wants the CIL to be noarch. Best regards, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Red Hat/Fedora packaging
I never liked my idea so if there is a better way, then by all means lets go
with that.
A better approach would be to remove the requirement that the .so be in the
same directory as the CIL, and instead probe architecture-specific directories
for the AOT'd code, e.g. ${_libdir}/mono/aot/[???]/[assembly-name].so. (Or see
Perl for additional ideas.) This would allow AOT'd code to be kept in a
separate directory structure, thus satisfying Fedora's requirements.
The problem here is that this would be a code change, and the requirements for
this feature are lacking (and no one has really complained about the situation
before). So, we have ideas, but no real impetus to implement the ideas (and
find out what limitations are present and need further work).
So the best way to get the packaging guidelines fixed is to remove the scenario
Fedora is concerned about. Do we have anyone interested in prototyping and
implementing this? :-)
- Jon
Thank you for addressing the problem. I consider this a must have. Red
Hat/Fedora has a large rich user community. I am glad that the animosity is
not bidirectional. I think this will help all software developers packaging
their own software.
Vlad
Digressions follow:
Hopefully, the next step will be everyone agreeing to use the rpm from
rpm5.org. Kudos to OpenSuSE on using LZMA, but when will the ftp site offer
Mono and other packages LZMA-ed instead of just gzip-ed or bzip2-ed?
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] inherited event
Hi,
I was wondering why events have to be in local scope to be called. If
the event is in a base class, shouldn't I be able to invoke it?
This is not a bug in Mono as the behavior is the same on MS.Net, but I
was hoping for some insight.
Thanks,
Vlad
namespace TestEvents
{
public abstract class Foo
{
protected event EventHandler OnBar;
public event EventHandler Bar
{
add
{
lock (OnBar)
{
this.OnBar += value;
}
}
remove
{
lock (OnBar)
{
this.OnBar -= value;
}
}
}
}
public class ConcreteFoo : Foo
{
public void DoStuff()
{
if (null!=this.OnBar) //Why can't I do this?
{
this.OnBar(this, new EventArgs());
}
}
}
class Program
{
static void Main(string[] args)
{
Foo foo = new ConcreteFoo();
foo.Bar += new EventHandler(foo_Bar);
//...
((ConcreteFoo)foo).DoStuff();
}
static void foo_Bar(object sender, EventArgs e)
{
Console.WriteLine(FOOBAR happened);
}
}
}
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] System.Threading.Thread.CurrentPrincipal.Identity.Name
Should this return something on Linux (OpenSuse 11.0)?
Thanks,
Vlad
using System;
class Eg
{
static void Main()
{
Console.WriteLine(Environment.UserName);
//This works
Console.WriteLine(System.Threading.Thread.CurrentPrincipal.Identity.Name
); //this returns empty string
}
}
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] About System.Net.Security.SslStream
I would like to reiterate my request for a binding to nss. I realize demand is not high, I just want every vote counted :). IMO if you want your vote to count then you should open a bug report* (severity Enhancement) otherwise it will just stay noise on the mailing-list. As you know I filed the enhancement https://bugzilla.novell.com/show_bug.cgi?id=444291 , but I don't think that noise is all that useless. Fedora is going to try to standardize on NSS. http://fedoraproject.org/wiki/FedoraCryptoConsolidation NSS will be part of LSB 4.0. In other words, yes Linux is great and diverse, but there will be a standard. http://ldn.linuxfoundation.org/node/5576 and probably elsewhere... I just don't want mono to be overlooked when it comes to writing standard secure software on Linux. Apparently, Red Hat and Sun have a lot of clout and can make their nss the Linux standard. It would not be farfetched for (Open)SuSE, Debian/*Buntu, etc to also migrate to using NSS. If I were on the appropriate mailing lists, I would certainly advocate for it there as well. I miss the Linux Hater. He would have my back... :) Vlad smime.p7s Description: S/MIME cryptographic signature ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] About System.Net.Security.SslStream
Sebastien, On Tue, 2008-11-11 at 15:12 +0800, ?? wrote: Hello! I'm using mono 1.9.2 and trying to write a small program using SslStream. The program runs as server side program. I found if I write the program as serverCertificate = X509Certificate.CreateFromCertFile(certificate); SslStream sslStream = new SslStream( client.GetStream() ); sslStream.AuthenticateAsServer(serverCertificate, false, SslProtocols.Tls, true); I will get a Server certificate Private Key unavailable error. This depends on which type of certificate file you're providing the X509Certificate ctor and the framework version you're using (I assume 2.9 since you're using SslStream). A typical .cer does not have a private key inside it so it wont work. Using a PKCS#12 file, containing a certificate and a private key, will work. The main difference is not the framework but the fact that Windows CryptoAPI defines relations between certificates and keys (when persisted) and allow code to find a private key when you supply a certificate (even if its not loaded from a certificate store). No such facility exists under Linux. I would like to reiterate my request for a binding to nss. I Believe that such a binding would provide the capability sought here. I realize demand is not high, I just want every vote counted :). Cheers, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] FIPS 140 cryptography
The Java community has JSS. Would asking for a MonoSS be asking too much? It depends from who you're asking ;-) Think of me as an ISV that wants to produce applications for the US government. The entire federal government including DoD is mandated to use FIPS compliant crypto libraries. Network Security Services for Java (JSS) is provided by Mozilla. So yes Mozilla *could* do something like this (not Mono-specific but for all .NET users) just like they provide the API for Java. OK. Today the picture is a little different. Red Hat produced Python bindings https://fedoraproject.org/wiki/Features/PythonNSS . These will be in fedora 10. And someone is producing Perl bindings http://search.cpan.org/~claesjac/Crypt-NSS-0.03/lib/Crypt/NSS.pm Making Mono and SuSE second class citizens in this realm. However I don't think this (NSS) should ever become a direct(*) Mono goal(**). Mono itself has already too many things to complete to afford a duplicate effort (since we already offer the same features). I am going to show my inexperience so please don't get upset when you tell me it is not so simple. Is there not a way to automatically generate a bunch of C# stubs based on the C headers? Besides NSS there are other FIPS140 certified libraries that could be wrapped to give the same end result. However I don't know any available on Linux that have .NET binding. If you are thinking of OpenSSL it is not as attractive (neither as current nor Level 2 certified) as NSS. If you are thinking of CryptoAPI it is a Windows only technology. I like NSS because it is cross platform (on top of NSPR) and certified to the teeth. Btw, will Crimson come out of hibernation? Thanks, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Regex and Unicode
Hello list members,
Since .Net regular expressions don't do POSIX, I replace POSIX character
classes by Unicode ones. See
http://www.regular-expressions.info/posixbrackets.html for the translation
table I used.
I ran into a problem with
POSIX Description ASCII Unicode
[:alpha:] Alphabetic characters [a-zA-Z]\p{L}
The problem occurs in .Net too so you might choose not to fix it and I would
understand. I hope however, you can fix it (and get MS to do the same).
The \p{L} regular expression throws an exception because the ampersand is
neither a character nor a close brace.
System.ArgumentException: parsing \p{L} - Incomplete \p{X} character
escape.
Parameter name: \p{L}
at System.Text.RegularExpressions.Syntax.Parser.ParseUnicodeCategory ()
[0x000a8] in
/tmp/monobuild/build/BUILD/mono-1.9.1/mcs/class/System/System.Text.RegularEx
pressions/parser.cs:796
Rest of exception clipped
using System;
using System.Text.RegularExpressions;
namespace test
{
class MainClass
{
public static void Main(string[] args)
{
Regex r = new Regex(@\p{L}); //running this code will
throw an exception
}
}
}
Since this might be by design, I am not using bugzilla. I hope someone can
tell me how this is supposed to work first. Also, I am not looking for a
workaround, as I have one. This is more of an FYI.
Thanks,
Vlad
smime.p7s
Description: S/MIME cryptographic signature
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Gdk and Cygwin
Hi,
I am in the unfortunate position of having to run Windows because my company
uses exchange and will not open IMAP or POP so I can read mail from
Thunderbird. So I run VMWare on Windows with an OpenSuse VM. I run CygWin
and use SSH to forward the display back to Windows/Cygwin. It works amazing
well as long as I close things regularly, because there are some memory
leaks somewhere.
I have a dual headed display because having two monitors allows me to spread
my mess which is the point of a big desktop.
Getting to the point:
Gdk.Screen s = Gdk.Screen.Default;
if (2 != s.NMonitors)
{
Console.Writeline(Where did it go?);
}
Given the complexity of the scenario, I doubt it is in the Mono layer, but I
figured I would report it here, because I don't belong to any other lists.
The number of monitors is but one... The size is the whole area.
Maximizing works as expected (stays on one monitor). I can span across
monitors. It mostly works, just the number of monitors is wrong.
Thanks,
Vlad
smime.p7s
Description: S/MIME cryptographic signature
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Mono.Security + SecureString
Hi, As you know, in .Net Framework 2.0 Microsoft added the SecureString class to keep passwords and other private data hidden. They did not add SecureString to the hashing or encryption/decryption providers to allow developers to take advantage of this new class. Mono does not use it in PKCS12 or anywhere else it could. It would be great if Mono took the lead and made touching private data a thing of the past. I could list a few places where password is accepted, but I am sure the security gurus know these classes way better than I do. I realize that this is an enhancement request, but security helps to sell technology. It would be nice to be able to say that Mono is more secure than .Net (or Java). Thanks, Vlad smime.p7s Description: S/MIME cryptographic signature ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] private members
Hi,
Given the following program, I have a question about Mono and .Net. This is
not a bug in Mono because I think the same thing happens on Windows, but I
was hoping someone could explain it to me. The line below with the comment
is surprising because it is accessing a private variable of another
instance. Why does it compile? Shouldn't it say that member bar is not
visible?
Thanks,
Vlad
using System;
namespace Test
{
public class Foo
{
private string bar = null;
public Foo(string bar)
{
this.bar = bar;
}
private Foo(Foo foo)
{
this.bar = foo.bar; //This line is surprising
}
public Foo Copy()
{
return new Foo(this);
}
}
class MainClass
{
public static void Main(string[] args)
{
Foo foo = new Foo(foobar);
Foo foobar = foo.Copy();
}
}
}
smime.p7s
Description: S/MIME cryptographic signature
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] [PATCH] Fix for RegEx parser.
Hi Arina, I've added a new set of tests for regular expressions and a correction to the parser.cs for passing the test PerlTest.Trial0990(). You mention a PerlTest along with regular expressions. This is very interesting. I was under the impression that Microsoft Regexes POSIX Regex Perl Regexes. How do I choose a regular expression language? Or am I limited to Microsoft expressions? Thanks, Vlad smime.p7s Description: S/MIME cryptographic signature ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Mono version numbering
On the other hand, just today I received an email with this text: It is C# based (some of the code uses .Net 2.0 features so it's not mono-ready code). The editor . I tried out the tool and aside from some typical path issues, the app ran great. Kudos to the WinForms developers! Is there any way to get MoMA to do some Cecil magic and fix path issues? 2 Points for evaluate, but you really get extra credit for remediate! Developers in the Windows world just don't realize how mature Mono already is. When Mono did generics I would already have called it 2.0. I would have made all the disclaimers needed, but people need to know that Mono works! We can shrink the list of disclaimers as progress is made. It is a question of degree of compatibility and IMO there is enough. When I can copy an exe and it just works, I may move my project over to Mono and add the missing functionality I need myself (the Mono dream). Thanks, Vlad _ From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Jerry van Leeuwen I think the only way to minimise confusion in the developer population that does not keep up-to-date with the details of development on Mono is to let at least the major version track full .NET compatibility. That is, do not move to v2.x.x until at least .NET 2.0 can be fully supported, do not move to v3.x.x until at least .NET 3.0 can be fully supported and maybe even do not move to v3.5.x until .NET 3.5 is supported. People that are aware of what is really going on under the hood will be able to make more detailed judgement calls and will not really care what version numbering you exactly use. It's the people that do not glance further than the version number that need to be accomodated. smime.p7s Description: S/MIME cryptographic signature ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] MonoSummit: Planning the Sessions
I am also hoping for webcasts of some kind. As a member of the Mono community, I feel a sense of pride whenever Mono gains respect on the Linux platform. As a consequence, I am hoping that more must have applications get developed with Mono. Whether it is games, tools or whatever, I truly love to hear that developers are having success with Mono. Java is the Honda of the market. It sells itself. Maybe we need a proselytizing campaign? To misquote Rodney Dangerfield How can we get some respect around here? A big concern of mine is penetration on RHEL. This is a very large user community and it seems to have diplomatic issues with Mono. Can this be resolved? * Moonlight/Boo/DLR I am cautiously curious about the Moonlets (the Moonlight desktop apps). Let's face it; sex sells and if it becomes easy to make sexy apps, everyone will use this technology. Will it become easy? Mono is becoming more and more efficient with memory. I would love to hear some plans to join with OpenMoko, GNOME (http://www.gnome.org/mobile/), etc on the embedded front. There is Java ME, where is our competing offering? Will working on the Mac spread core developers too thin? Should we stick to Gtk and get that right first? Vlad P.S. Thanks for all the hard work from Miguel, his team at Novell, and the rest of the community. -Original Message- [EMAIL PROTECTED] On Behalf Of Miguel de Icaza Hey folks, Am looking for ideas for topics that developers would be interested in hearing about at the Mono Summit. smime.p7s Description: S/MIME cryptographic signature ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] Replicating System.Web.Script.Serialization bugs
My apologies for being nosy, but if performance is what you preach, using
as is even better...
String s = obj as string;
If (string.IsNullOrEmpty(s))
{
return null;
}
3. Please commit the code below for performance reasons.
a. 'is' and casting are much cheaper.
b. don't need to create an instance, returning null should be OK, isn't?
+ /*
+* Take care of the special case whereas in JSON
an empty string () really means
+* an empty value
+* (see:
https://bugzilla.novell.com/show_bug.cgi?id=328836)
+*/
+ if ( (type.IsGenericType)
(type.GetGenericTypeDefinition() == typeof(Nullable)) )
+ {
+ if ( (obj is String)
(((String)obj).Length == 0) )
+ {
+ return null;
+ }
+ }
+
Vladimir Giszpenc
smime.p7s
Description: S/MIME cryptographic signature
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
Re: [Mono-dev] FIPS 140 cryptography
Sebastien, No. The short story is that it's a big, long and costly project and that there's been, so far, not many demand for it (wrt to other Mono features). I agree that there are more pressing needs, but it never hurts to voice your own so they are recognized. I am a big Mono proponent, but it is difficult to compete with Java on maturity of tools. While it's less elegant (imho) you can still use FIPS140 certified crypto in Mono by wrapping an existing toolkit (e.g. nss) in C# [1] and using this as a replacement [2] for Mono's crypto (split in both Mono.Security and mscorlib). The Java community has JSS. Would asking for a MonoSS be asking too much? That seems like a big project as well. This is not even what I am really looking for. I need SSH. I found the capability in SharpSSH, but in order for me to make certification labs happy, I need the encryption used in my SSH implementation to be FIPS 140 compliant. [1] which is exactly what MS is doing on Windows: the FIPS140 crypto comes from CryptoAPI (managed stuff isn't certified) and wrapped in the class library. [2] it's possible to remap cryptographic algorithms using machine.config (so all mono tools and *correctly* written applications/libraries will be using your own crypto). That is nice for Windows, but I am targeting Linux. I would love to point some config file at NSS. Is there any chance such a thing will happen? I am a little peon doing RD hoping to move the Army toward accepting Mono. Any help you can give me would be much appreciated! Many Thanks, Vladimir Giszpenc DSCI Contractor Supporting US Army CERDEC STCD IAD Tactical Network Protection Branch (732) 532-8959 ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Idea for gendarme
Hi,
Until recently, I have been able to stay away from threads. I assumed
the event handlers in the GUI would be threaded enough and all was
good in the world. Alas, my Gtk# application needs worker threads to
be responsive.
I use the anonymous delegates heavily though probably still not enough.
Gtk.Application.Invoke (delegate { --code to update GUI here-- });
For the most part it works great!.
I would like to share something I learned. Event handlers need to be
cleaned up. It is always a good idea to cleanup after one's self, but
with a great Garbage Collector one can easily forget.
If you have some code that adds an event handler in a set property,
you need to find a way to remove the event if it was previously added.
This probably seems obvious to all the senior C# gurus out there
(please don't give me lessons in style), but I was very impressed to
resolve this problem.
public SomeProp SomeProp
{
set
{
//DO NOT forget to remove the event previously added
if (null != someProp)
{
someProp.ResultsUpdated -=
new EventHandler(SomePropEventHandler);
}
someProp = value;
someProp.ResultsUpdated += new
EventHandler(SomePropEventHandler);
OnChanged();
}
}
Is there any way to have Gendarme verify that the events get cleaned
up somewhere? Good manners with regard to resources in my mind should
be something that could be analyzed to produce some kind of warning.
I still have some thread/resource issues and it is extremely difficult
to hunt down these bugs. I am unable to close my application after I
start my workers even if they run their course and finish. I probably
have more issues of the same type.
Thanks,
Vlad (a humble Gtk# noob)
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] more Gendarme rules
Hi,
In programming spelling counts and case usually matters as well. A
perfect example is in properties.
//Microsoft says we should no longer use m_ or hungarian notation
//note the lower case first letter
private Type propertyName;
public Type PropertyName
{
get
{
return this.PropertyName; //this will crash your app
}
set
{
return this.PropertyName = value; //this will crash your app
}
}
This is fairly easy to find, but should also be easy find by some code
analyzer. Gendarme rule or compiler warning... Either way, I await
the next MonoDevelop with baited breath.
Thanks,
Vlad
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Interlocked
Hi, This is probably a question for Microsoft, but since I am using Mono I will ask you fine folks. System.Threading.Interlocked.Read() addresses 64 bit atomicity. If I use the shortcut int, is it System.Int32 or System.Int64 depending on the architecture the code is compiled on? If that is the case, shouldn't I be able to use the Read() with ints as well as long? Or is reading any kind of int always an atomic operation? I have never even been near a 64 bit machine, but I feel I should write code that is portable. Thanks for your patience, Vlad ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Subversion on MonoDevelop
Hi, The subversion add-in has made my MonoDevelop experience nice and smooth. Yesterday, it hit a slight bump when my subversion server certificate expired. SVN from the command line was able to get a nice message: Error validating server certificate for 'https://myhost.myserver.com:443': - The certificate is not issued by a trusted authority. Use the fingerprint to validate the certificate manually! - The certificate hostname does not match. - The certificate has expired. Certificate information: - Hostname: myhost - Valid: from Tue, 23 May 2006 18:24:13 GMT until Wed, 23 May 2007 18:24:13 GMT - Issuer: IT, My Company, Eatontown, New Jersey, US - Fingerprint: 72:06:72:f4:68:7f:63:e1:f2:93:35:58:df:d4:d5:0f:a1:3b:68:0b (R)eject, accept (t)emporarily or accept (p)ermanently? Is there any way to make the GUI client capable of displaying a similar message? Thanks, Vlad PS While I am on the subject of Subversion, why is the diff window not editable? ___ Mono-devel-list mailing list Mono-devel-list@lists.ximian.com http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] XmlDocument.CreateElement
Hi,
I don't have Windows so I could not test this on the .Net CLR but the
behavior I am getting seems weird. If I call
xmlDocument.CreateElement(foo:bar), the foo gets removed if I don't pass
in the namespace as the second parameter. Is this by design?
See attachment for example...
Thanks,
Vlad
using System.Xml;
public class Test
{
static void Main()
{
XmlDocument doc= new XmlDocument();
XmlElement header = doc.CreateElement(foobar);
header.SetAttribute (xmlns:foo, http://foo.bar.org/XMLSchema;);
//The name of this element has a namespace in it, but apparently
that is not enough
//because it is removed in the saved file.
XmlElement foobar = doc.CreateElement (foo:bar);
XmlElement ok = doc.CreateElement (foo:ok,
http://foo.bar.org/XMLSchema;);
header.AppendChild(foobar);
header.AppendChild(ok);
doc.AppendChild(header);
doc.Save(foobar.xml);
}
}
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
[Mono-dev] Mono enhancements
To the Gendarme gurus,
I don't know if the GC is smart enough to Dispose of resources for us, but I
will assume that it does not hurt to help it. There are two patterns that
Gendarme could look for.
class Foo
{
Foo()
{
IDisposable resource = new Resource();
throw new System.Exception();
resource.Dispose();
}
}
This should either be remedied with a try finally or with the using construct
which encapsulates try finally (I vote for the latter as a suggested fix).
Another related rule should be to check if resource.Dispose() is never called
anywhere in scope.
An unrelated pattern which is not too important to me:
class Foo
{
Foo()
{
System.Console.WriteLine(string.Format(hello {9}, world));
}
}
Note that I missed the zero and typed nine.
Is it possible to promote Gendarme rules to compiler errors/warnings or would
that somehow break compatibility with csc? Is it worth it if it means better
code is produced? I realize that it won't matter as much once there is a nice
Gendarme Plugin for MonoDevelop but even then, we could ask for severe flaws to
be compiler warnings at least.
Many thanks,
Vlad Giszpenc
___
Mono-devel-list mailing list
Mono-devel-list@lists.ximian.com
http://lists.ximian.com/mailman/listinfo/mono-devel-list
