|
Charlie. The other answers are good answers, but I'm going to throw one more piece of confusion into the mix. There's a form of WPF that runs in a browser called an XBAP, and it is possible to host a WPF control in Silverlight. See here[^] for an example of a desktop/web control.
|
|
|
|
|
I've been fighting this one for a couple of days. I have 2 nics, and 1 computer, the 2 nics are connected to each other (same or separate subnets).
In my C# tool I am essentially trying to send packets from one interface to the other. Seems pretty basic... The issue that I am having is the internal route tables seem to be routing, instead of allowing me to force what I need out the interface. I am using .NET sockets, and thought BINDing to the specific interface and then using the DONTROUTE socket option would do what I need. Should this work? The other option is likely to use raw sockets (spoof), but apparently winXP SP2 has canned that possibility.
|
|
|
|
|
What errors are you getting? Is any data getting through? Did you remember to use a crossover cable between the cards?
"A Journey of a Thousand Rest Stops Begins with a Single Movement"
|
|
|
|
|
I have a .Net 2.0 C# Windows Service that reads messages from a message queue (xml), transforms them using XslCompiledTransform (does some other work using XmlTextReader and MemoryStreams) and then finally inserts data into a sql server database (In a distributed COM+ transaction).
The problem is that sometimes the service crashes with the .NET Runtime 2.0 Error as source and EventId 1000.
Faulting application application.exe, version 4.2.0.0, stamp 48a93225, faulting module kernel32.dll, version 5.2.3790.4062, stamp 462643a7, debug? 0, fault address 0x0000000000027d8d.
(The machine is a has a multi-core CPU)
My initial thought was that the service tried to read or write to a stream while the system was moving data between the ram and the swap. But after some more testing, the problem occurs even with ~750-1000mb free ram)
Any ideas on how to locate and fix this problem?
|
|
|
|
|
when i try to open remote web site in vs2005 from file-->open web site at write path like
http://pt.tvtc.gov.sa the message comming
The Web server does not appear to have FrontPage Server Extensions installed
but the site can open in ftp.
please how can i solve that problem?
Ala Qunaibi
|
|
|
|
|
Install the FrontPage Server Extensions on the website? It might just work
Regards,
Thomas Stockwell
Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning.
Visit my Blog
|
|
|
|
|
Hello;
I have created ActiveXcomponent in c# which is used to downloads the file from internet multithrededly I am using background worker to accomplish this task. I am using javascript to call this dll . Everything works fine for some time but after that internet explorer crashes and give the error message
EventType: clr20r3 P1:iexplore.exe P2 :8.0.6001.18241 P3:48ae8fee P4:mscorlib P5:2.0.0.0 P6:471ebc5b P7:3404 P8:21c P9:system.io.ioexception
I dont know from where this error is coming (I am using .NET 2.0)
Note:When I call the same method from above dll using windows form I didnt get any exception..
I will be very glad if you guide me
Thanks in advance
Regards
Mahesh Dubey
|
|
|
|
|
I do not believe it is possible to write an ActiveXComponent in C# becuase C# is natively .NET and ActiveX is not. That in itself could be causing the problem.
Regards,
Thomas Stockwell
Programming today is a race between software engineers striving to build bigger and better idiot-proof programs, and the Universe trying to produce bigger and better idiots. So far, the Universe is winning.
Visit my Blog
|
|
|
|
|
Could not load file or assembly 'Microsoft.Vbe.Interop, Version=11.0.0.0, Culture=neutral, PublicKeyToken=71e9bce111e9429c' or one of its dependencies
I am getting this error when i run my application. Please provide the rectification for it.
|
|
|
|
|
When you deployed your application to the machine in question, did you deploy ALL the .DLL's in the bin\Release folder??
This may seem like a stupid question, but is Word installed on the machine?? Is it the same, or later, version of Word you have on your development machine?
|
|
|
|
|
Yes, I took the references of all dlls. And Microsoft word is also installed in the system.
|
|
|
|
|
Is the same version of Word installed? Different versions have different IDs.
|
|
|
|
|
Better yet, is Word installed on the machine?
Any suggestions, ideas, or 'constructive criticism' are always welcome.
"There's no such thing as a stupid question, only stupid people." - Mr. Garrison
|
|
|
|
|
I have a .net application that works with a sql server 2000 database. Both the program and the database must be on the client machine . So I need to run the sql server setup on the client machine. What I want is not to let the user access the database by enterprise manager or query analyzer. How can I do it ?
|
|
|
|
|
Define the login (or use existing) for your program (either windows or SQL Server authenticated) and disable other logins. If you use SQL Server authenticated logins, set their passwords complex enough.
The need to optimize rises from a bad design.
My articles[ ^]
|
|
|
|
|
The IndexOf(Of Integer)() seems to be a bit sluggish. It should be possible for the operation to run quite quickly, but performing IndexOf() on a 10,000 item array seems to take about 44us (measured over 1000 repeats); a simple version written in C, loaded as a DLL, takes about 20us. Searching for an integer within an array would seem like a common enough operation there should be some built-in function to do it quickly. Is there any way to do a fast search without needing to have a DLL separate from the EXE?
|
|
|
|
|
Hi,
I am inclined to doubt your measurements a bit; they seem to indicate an elapsed time of
a few tens of milliseconds, which is quite possible to measure in the wrong way (see e.g.
my timers article).
I would suggest you time a ten times bigger job, and then repeat it say 5 times and watch
the numbers. Also make sure the start conditions are the same, i.e. array loaded or not loaded
in memory before starting the timing, code JIT-compiled, etc.
I trust Array.IndexOf() to be in the same league as a simple C implementation.
|
|
|
|
|
Luc Pattyn wrote: I would suggest you time a ten times bigger job, and then repeat it say 5 times and watch
the numbers.
I expanded the repetition count a hundredfold. I had been running the tests repeatedly before; I wouldn't expect any JIT code to be purged between reps. The difference is no longer quite as great, but it's still pretty significant. Repeating 100,000 times the same search of 10,000 elements takes about 4.5 seconds with IndexOf, versus 3 seconds for the DLL.
Probably better to use a different algorithm than worry about search time, since even a 33% performance improvement probably won't be enough (though if I could get a 33% improvement just by changing an 'IndexOf' to something else, I wouldn't turn it down). I wish I knew of better ways to search for things so I could avoid reinventing the wheel.
My intention was to create a replacement for Dictionary which would add methods to purge items during an enumeration, and to perform a shallow clone. The implementation holds an array of hash values; it works well for small dictionaries, but gets a little boggy as things get larger. The majority of data sets will probably be ten items or less, but some may reach as high as 10,000. It seems that even the simple integer search becomes problematical for the larger data sets.
The dictionary is actually intended as part of a hierarchical collection class which needs to support intersection, difference, and union operations. I don't know if you're aware of any such thing already existing. My present implementation isn't horrible, but it gets a little bogged down sometimes. Something faster would be nice.
|
|
|
|
|
I don't know whether they will help with your particular problem but two collection libraries that you could investigate are PowerCollections[^] and C5[^]. The former is easier to use and better documented but the latter is more powerful. I've only used the former in anger and only sparingly. PowerCollections is less signifiacnt in the .NET 3.5 world though.
Kevin
|
|
|
|
|
Kevin McFarlane wrote: I don't know whether they will help with your particular problem but two collection libraries that you could investigate are PowerCollections[^] and C5[^].
I'll take a look at those. Right now I'm coding in VB.net; it seems from a quick glance that those libraries are in C#. How would one integrate the C# libraries into a VB.net executable using either VS2005 or VBexpress2005?
BTW, on a related note: right now I'm using both VBexpress2005 and VS2005 so that I can have different editor colors and taskbar icons for my VS projects from my VBE projects; this makes it easier to make sure I don't go editing files in the wrong project. Is there any way to set things like taskbar icon or editor colors on a per-project basis? I'm thinking I might be better off just using VS2005 for everything.
PS--I find it curious that the VisualBasic.Collection is in many ways superior to the other collection types. Its performance seems to be generally comparable to Dictionary, but it allows access by index and deletion during enumeration. Though it lacks some features that would be essential for emulating a Dictionary (e.g. enumeration of key-value pairs) and it has some other annoying weirdnesses (e.g. indices starting at 1 rather than zero, reversed arguments on the .Add method, etc.) it seems like the underlying data structure is pretty good. I wonder why Microsoft never implemented any class which shared the features, but not the limitations?
|
|
|
|
|
supercat9 wrote: it seems from a quick glance that those libraries are in C#. How would one integrate the C# libraries into a VB.net executable
Just add a reference to the DLL in the usual way.
Kevin
|
|
|
|
|
Hi,
1.
Array.IndexOf() returns the first match, which makes it essentially a linear search.
What you want is a "find any" that scales well, let's say a hash table.
As you know, when searching becomes expensive, you must invest in sorting, and the best way
for sorting often is "sort while growing" i.e. keep the list sorted at all times.
2.
.NET has a SortedList class, which is actually a Dictionary. It keeps its keys in sorted order
at all times.
3.
Wouldn't it be better to inherit from Dictionary and add the functionality you want?
I guess you can replace the Add and Remove methods, so you are aware of those operations,
then create your own IEnumerable implementation which could use indices and "return yield".
4.
It may be unwise to come up with an IEnumerable that behaves quite differently from regular
IEnumerables, those that do fail when the collection gets modified at all.
5.
You might use Reflector (or look into the Rotor source) to see how they went about it
in the first place.
|
|
|
|
|
Mark Salsbery
Microsoft MVP - Visual C++
|
|
|
|
|
Luc Pattyn wrote: 1. Array.IndexOf() returns the first match, which makes it essentially a linear search.
What you want is a "find any" that scales well, let's say a hash table.
There are a number of approaches that can be used. I figured I would try the IndexOf approach to see how far it would scale before the linear term became too dominant. The answer is, not quite far enough.
As noted, the most common number of items is probably below ten. Presently I start with a 16-element array and double the number of elements every time it overflows. If I start with a hash table, then the small dictionaries are going to be stored very inefficiently. Perhaps the thing to do would be to start with an array, and switch to a 16-way hash table when the number of elements exceeds 256, a 256-way hash table if the number of elements exceeds 4096, and a 4096-way hash table if it exceeds 65,536. Regenerating the hash tables at each boundary case wouldn't be free, but such the vast majority of elements would be regenerated at most once.
Luc Pattyn wrote: 3.
3. Wouldn't it be better to inherit from Dictionary and add the functionality you want? I guess you can replace the Add and Remove methods, so you are aware of those operations, then create your own IEnumerable implementation which could use indices and "return yield".
The Dictionary class does not expose its inner workings. While it would be possible to implement something like the 'DeleteForAll' method that's available on a List, such a method would have to queue up a list of keys and then process deletion using the list. This would require a key lookup for every entry to be deleted, whereas a well-constructed DeleteForAll primitive shouldn't require any key lookups.
Luc Pattyn wrote: 4. It may be unwise to come up with an IEnumerable that behaves quite differently from regular
IEnumerables, those that do fail when the collection gets modified at all.
So you're saying that working sensibly when a collection is changed during enumeration would be a breaking change?
Luc Pattyn wrote: 5. You might use Reflector (or look into the Rotor source) to see how they went about it
in the first place.
I'm not familiar with those things.
BTW, after posting my earlier message, I was thinking that for my hierarchical collection it might be better to build in whatever sort of dictionary implementation I'm using rather than using a dictionary-style object, since copy-on-write cloning would probably work wonders for performance and I don't think such a paradigm is apt to be useful with anything resembling a standard Dictionary (a Dictionary could implement copy-on-write shallow cloning, but in my scenario what's needed is copy-on-write deep cloning; that would require that when an object held in the dictionary is modified, it notify the dictionary so that the dictionary can perform the necessary hard cloning operation).
|
|
|
|
|
supercat9 wrote: working sensibly when a collection is changed during enumeration would be a breaking change?
yep, your SpecialDictionary.GetEnumerator() would behave quite different from
List.GetEnumerator() and others. I'm not saying better or worse, just different.
Which would tend to be confusing to readers/users of your code.
supercat9 wrote: I'm not familiar with those things.
The ultimate dictionary[^] will tell you more on Reflector, a very useful programming tool that generates pseudo-source-code for existing .NET assemblies.
|
|
|
|