I'm not sure if this is a possible bug in the .NET Framework, the framework
documentation, or the C# compiler. Perhaps someone can comment.
I'm using Reflection.Emit to create a nested type. Assume I'm trying to
model the following classes:
public class Outer
{
public class Inner
{
It's been a while since I played in this space, but now it's all coming back
to me... :-)
I was able to reproduce your configuration (assuming you're calling
Context.RegisterDynamicProperty from within IsContextOK on your context
attribute). It works fine for me. The only thing you have to do i
From: "Ethan J. Brown" <[EMAIL PROTECTED]>
>
> The one or two examples I've seen always create a new context no
> matter what, when a specific required ContextProperty (which
> implements IContextProperty and IContributeObjectSink) is not present
> in the list of properties for a context. After l
So I've been messing around with custom attributes and contexts
in .Net quite a bit lately, delving into the Rotor source as
necessary when I needed documentation, etc. To give a basic rundown
for those who haven't jumped in to these undoc'd waters, when your
attributed class is constructed, the
And regarding easy extensibility: Please tell me who implemented the
MSMQ
transport for queued components? Microsoft, true? And for Remoting? Some
independent developer.
Where can I find that MSMQ queued component for remoting implementation?
TIA,
Richard A. Hein
Level Platforms Inc.
Softw
An interesting scenario has arisen with the previously-confusing security
paradigm. All is working now in terms of the remote authentication server
authenticating each request, and we've implemented a custom HttpModule to do
some of the "upstream" work. Before i forget - thanks to K. Brown for the
The last major project that I did where I was the data layer developer had
seven backend databases that I could get to, a couple of "services" from
which I got some data, and one other non-database data source. Some of the
databases and all of the non-database data stores were owned by third-part
FOLLOWUP:
After testing this on another machine another issue come up which would
explain why some people can't duplicate my original problem.
When I try to download the exe from the web on this machine IIS actually
denies access to it. But somehow the app. still runs! Is it somehow still
mana
Going to Windows Update and updating IE6 solved the problem.
Thanks.
You can read messages from the Advanced DOTNET archive, unsubscribe from Advanced
DOTNET, or
subscribe to other DevelopMentor lists at http://discuss.develop.com.
Yes, I've read all of the relevant posts that I could find. None of them
addressed my specific problem.
The problem is NOT how to access parameters passed to the web-launched .exe
in the query string. That works fine. The problem is that when you pass
parameters using the query string like tha
On Tue, 3 Sep 2002 13:41:43 -0500, Cavnar-Johnson, John wrote:
>Do you really consider it a bug that Microsoft implemented Internet
>Explorer to follow the HTTP 1.1 spec?
>
No I consider it a bug that the framework doesn't play well with IE and
it's limit of two connections and thereby frequen
By looking at the book in my hand I would say Ingo prefers .NET remoting.
JJ
> Disclaimer: Clemens and I have known each other personally for some time
and
> we have a "deep, but friendly and agreed-upon disagreement about the value
> of .NET Remoting [and DCOM in .NET]". It's up to you to figure
Wups! I forgot to address the CallContext issue - you say that
ServicedComponents support CallContext although the docs say they don't.
Actually, the support is spotty at best - I have seen calls into
ServicedComponent-to-ServicedComponent calls lose the CallContext and I know
for a fact that Serv
I'm not sure I entirely follow what you're saying. I was talking about
physical layers. Were you referring to them as logical layers?
If so, I'm opposed to a separate "data layer", where said data layer
consists of a set of objects. I think that the mid-tier (i.e. what the
presentation layer talk
Disclaimer: Clemens and I have known each other personally for some time and
we have a "deep, but friendly and agreed-upon disagreement about the value
of .NET Remoting [and DCOM in .NET]". It's up to you to figure out who
prefers which model.
I generally think in highest terms of Clemens but we
I'm going to propose that, if we want to continue this talk, we take it over
to DOTNET-ADVOCACY. I think we'd all agree it's been an interesting and
mentally invograting thread :-), but we're long past responding to Jesse's
original question and are well into the mode of trying to convince each
o
Ian,
First there's a general point I'd make. One thing that really makes
me uneasy about not only this thread but other threads I've
seen on the .NET lists about the implementation inheritance
issue this: Reading between the lines of some posts
I get the impression there's an underlying attitud
Hi Ian
I've split my reply to your post into two posts because I think
the length of the messages was getting too long. This post addresses
the question of exactly how inheritance might or might not
break code, my other post is more generally about whether
sealing by default is bad or not.
First
I agree with your comments, however, I must take exception to your
characterization of the data tier as merely being sprocs. If you have a
single database as the respository of all of the data, sprocs is a way to
go. If you are taking data from multiple databases and there are
differences in cod
1) ASP.NET doesn't use the SoapFormatter, but uses the
System.Xml.Serialization framework and you should stick to that.
2) It looks like you need to manipulate and map the stream in the
AfterSerialization/BeforeDeserialization stages and do a wholesale
replacement of the body in the outbound chain
I have to note that you've done what I encounter in many
classically-trained OO programmers: presented the ideas that
encapsulation, inheritance, and polymorphism are Good Things, without
saying why.
Note - I'm not saying that you don't have reasons. I'm saying that there
are lots of people out t
Hi Andrew,
I don't understand why you would want to do that. What is your problem
with ServicedComponents, what's your problem with DCOM transport ?
DCOM transport is the fastest thing you can get. On the same machine, no
technique known to man beats LRPC in terms of performance (calls execute
o
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Martin is right. There is a call to Dns.Resolve in this version of the
constructor. If you know the address, you really want to bypass
anything that takes a string, but rather use IPAddress directly.
You can also take the long way around and constru
Wish I could contribute more to this conversation, but for now...
> Does this make it clear why I think that putting such helpers into classes
> that are not related by inheritance is better? It gives you more
> flexibility. Inheritance paints you into corners. (And this is not the
> only case
I believe I have developed a workable solution for supporting distributed
transactions over .NET remoting. The purpose behind this post was to make
sure I didn't inadvertently duplicate someone else's work or take a
wicked-bad wrong approach.
Note: To anyone who is going to remind us that distrib
Just in case anyone else has made the mistake of using the constructor of
the TcpClient class to connect to a Server where the IP Address is already
known, don't do it. The constructor is for Hostnames and will send a UDP
packet to try and resolve the IP Address. This will hurt connection
perform
> When I think of how I would do this I'd think in XML only. If you can
> clearly describe the shape of the "bad" XML and the shape of the XML
you
> want why not transform one stream into the other using XSLT? The
> SoapExtension framework is good for that sort of thing.
>
> Not sure that helps
> In many (maybe most?) enterprise systems, tiers are built by different
> developers. E.g. HTML/ASP/ASP.NET is written by one guy, business
logic
> (VB
> com(+), VC/ATL com(+), C#. VB.NET) by another, and database structure
and
> stored procs - by yet another.
>
> Yet, for scalable systems, when
> My question is, provided that to many request objects DON'T fill up
the
> available memory, isn't it the responsibility of the ThreadPool class
to
> manage the number of thread being used, perhaps, even less then the
> maximum allowed, since that to could cause problems. In other words,
if
> fi
When I think of how I would do this I'd think in XML only. If you can
clearly describe the shape of the "bad" XML and the shape of the XML you
want why not transform one stream into the other using XSLT? The
SoapExtension framework is good for that sort of thing.
Not sure that helps but I can't
I haven't had any problem opening connections with the Managed Povider for
Oracle. There are a couple of things I noticed in your post.
m_Connection = "Provider=MSDAORA.1;User ID=" + userId + ";Password=" +
password +";Data Source=ORCL";
You no longer need to provide the provider in the conne
I am not really sure what your message is about but it does strike a chord with
me if you are refering to alternate ways to invoke web service calls. Some
time back I posted a request asking about ways to do web services without an
http call. I was interested in queuing up requests that could th
Did you try this? It might work I'm not sure with the latest .Net SP's?
http://discuss.develop.com/archives/wa.exe?A2=ind0202B&L=DOTNET&P=R30516&I=-
3
-Original Message-
From: Dave Adair [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, September 03, 2002 5:42 PM
To: [EMAIL PROTECTED]
Subject:
I've just tried this and get a dialogue box which is empty apart from 3
buttons.
Do I need to do something else to get it working?
Cheers
Simon
-Original Message-
From: Dominic Pease [mailto:[EMAIL PROTECTED]]
Sent: 05 September 2002 06:18
To: [EMAIL PROTECTED]
Subject: Re: [ADVANCED-DOT
>Which reminds me of a question I've really been anxious to ask for quite
>some time:
>
>In many (maybe most?) enterprise systems, tiers are built by different
>developers. E.g. HTML/ASP/ASP.NET is written by one guy, business logic (VB
>com(+), VC/ATL com(+), C#. VB.NET) by another, and database
First, from the online help for FolderNameEditor:
"This type supports the .NET Framework infrastructure and is not intended to
be used directly from your code."
So I think you are on dodgy ground here. And the
FolderNameEditor.FolderBrowser class has the same comment applied. However,
we can i
Erick Thompson <[EMAIL PROTECTED]> wrote:
>
> Ian's comment is that if a function can be done using the public
> methods of an object, then the function should be in a static
> utility class. This could work for the darken method mentioned
> in this thread, but what about something like adding a n
You need to call FlushFinalBlock on EncodeCrypto stream before closing
it.
Just as a side note:
It could be better if you derive key from the password using random
salt, and set encryptor's initialization vector to some random data too.
You can simply store IV and Salt to the stream just before
Did you search the DOTNET archives? This has been heavily discussed here (or on
DOTNET-WINFORMS). I believe the search page (available at the bottom of every email)
now offers the ability to search across the different lists. Hun Boon and Chris Sells
both posted some working code to deal with t
public class FolderChooser :
System.Windows.Forms.Design.FolderNameEditor
{
public static string GetFolder()
{
FolderBrowser fb = new FolderBrowser();
fb.ShowDialog();
return fb.DirectoryPath;
}
}
-Original Message-
Doesn't implementing an interface really only give you the polymorphic side
of inheritance? and how would you model 4 deep with each derivation adding
more and more methods.
-Original Message-
From: Erick Thompson [mailto:[EMAIL PROTECTED]]
Sent: Thursday, 5 September 2002 12:02
To: [EMAI
41 matches
Mail list logo
