Comments by Pablo Aliskevicius (Top 134 by date)

Striked out the irrelevant. Still hope to help.

Are you writing a compiler?
If not, google VTABLE and read the first 5 results. At least 2 should match your curiosity.
If you're writing a compiler, you'll need a big pile of books.

My guess? The hard disk.

Which of your parameters is a string?
What is the C++ function signature?

I get one record even when SQL Profiler is not running.

Look for this in that page:

empid ordermonth qty runqty
----- ----------------------- ---- -------
1 2006-07-01 00:00:00.000 121 121
1 2006-08-01 00:00:00.000 247 368
1 2006-09-01 00:00:00.000 255 623
1 2006-10-01 00:00:00.000 143 766
1 2006-11-01 00:00:00.000 318 1084
...
2 2006-07-01 00:00:00.000 50 50
2 2006-08-01 00:00:00.000 94 144
2 2006-09-01 00:00:00.000 137 281
2 2006-10-01 00:00:00.000 248 529
2 2006-11-01 00:00:00.000 237 766

Column 'runqty' seems to be exactly what you want.

The one about the islands. It's not exactly the same thing you've got (one date per record instead of two), but it's pretty similar.
Best wishes,
Pablo.

Did you look at this?

http://www.codeproject.com/Articles/23198/C-String-Toolkit-StrTk-Tokenizer

This should have been an answer.

What happens if you try this?

value=To<size_t,char,my_traits_type>(my_string);

What happens if you try startColorstr='#FFf0f0f0' instead of startColorstr='@from'? (Maybe the @from is not initialized to the background color, depending on browser settings).

Are you compiling in release mode? If so, try:
SHFILEINFO sfi = { 0 }; // zero memory of struct, otherwise may hold garbage.

Reason for my vote of 5 \n Better late than never...
Today I stumbled upon this article, it's great: clear, concise, and to the point.
Thanks for sharing!

I did something similar once:
* The Broker converts the request into a Job, and puts the Job in a queue. The Broker notifies the client that the Job is in the Queue.
* The Workers (number depends on number of CPUs) pull jobs from the queue, and execute them.
* The Client has a way to query the state of 'his/her' requests.
'Job' is an interface, with one method (Execute()).
In this way, the number of threads was minimized (all Agents disappear; since Job is polimorphic, any worker can do any Job).
In your case, you would also need a Log, that holds the Job state (Queued/Executing/[Succeeded|Failed]). Querying the job state may be done through the Broker, or separately. In my case, the client code may send a request, the client machine may disconnect, and much later, the client reconnects and asks for status.
Pushing notifications to the client is a different matter, but may be done in the context of the execution of the Job.

Just my two bits,
Pablo.

ie 8 can behave as if it was ie7, 6 or even earlier. Before standards, each browser had its own set of 'quirks'. Google 'HTML DOM Quirks', you'll find a lot of stuff there.

Try hitting F12, see if you are in standards mode or quirks mode. A bit of a long shot, but I've heard that happened to someone I know.

Google 'click once'.

It may be an issue with Jurassic itself.
Did you try to post one at their site?
http://jurassic.codeplex.com/workitem/list/basic
They have an 'issue tracker', and you can report bugs there.

Did you read the error message? They usually point you in the right direction.

Did you try putting the server in COM+?

I'd suggest to send everything but the pictures themselves to the client when loading, then using setTimeout() or setInterval() and AJAX to get the pictures themselves. Not exactly the realm of 'basic HTML' any more.

Thanks for your comment; and more than that, thanks for the link!

Reason for my vote of 5 \n Thanks!

Most text editors offer an 'encoding' option for files. In Visual Studio, you can see it using File/Save As: there is a combo with encodings there. You can do that for all your Javascript files.
For your ASPX-generated content, you can add a globalization element to your web.config. See http://msdn.microsoft.com/en-us/library/hy4kkhe0%28v=vs.71%29.aspx; if your machine.config specifies something different than UTF-8, you should specify UTF8 there.

Did you try some kind of popup blocker in IE?

Maybe not iOS, but Linux?

What happens if your string variables have actual values?

CloseHandle()?

Did you mean to call dtDetalle.Columns.Clear() sometime?

Consider adding 'InstallShield' to the tags of your question, and removing 'C#'.

Then do only the bulk insert part.

My guess is that there is some kind of race condition: ReadToEnd() may get called before ffmpeg writes to the standard output. But, it's just a guess...
I've upvoted the question in the hope of luring someone who knows better.
Best wishes,
Pablo.

Consider adding 'InstallShield' to the tags of your question.

What is the output if you run the command on a console window?
What happens if you call ReadToEnd() again after WaitForExit()?

If your question is about homework, label it as such.

Thanks!

Awk is open source. Download it, and read.

When it's ready, I'll gladly download it!

Did you try #import?

What error?

'objectarx'?

What is spDoc?
If it is a smart pointer, it will release itself when it goes out of scope.
If is is a raw DOMDocument40 pointer, call ->Release() on it one line above the line where it goes out of scope.

No, I don't.

I can't, so far I've never played with lilo.conf, only with less sensitive files (like wikipedia_lens.py).
Best wishes,
Pablo.

Sounds like an ANSI/Unicode issue. I'd suggest you add a few TRACE statements at the beginning of your unmanaged methods.

It all depends on what your custom action does. If it does not modify the system in any way (e.g., if it checks whether a given .ini file is present, and then updates MSI properties based on its contents) you can schedule it to run whenever you want.
Personally, I steer clear from any custom actions that have side effects: I use custom actions to update properties only, and regular actions based on those properties for everything else.

You can use ORCA to see the actual sequences in your MSI file. See http://blogs.msdn.com/b/astebner/archive/2004/07/12/180792.aspx

Hope this helps,
Pablo.

Is CSmartFile::~CSmartFile() imported from a VC6 DLL?
If so, the reason may be that name mangling (the '@@' part) is different in VC2010.

'Scriptomatic' does it. It's a free download!
http://www.microsoft.com/en-us/download/details.aspx?id=12028
http://technet.microsoft.com/en-us/library/ff730935.aspx

Why build it? WMI is already in all of your Windows servers.

Thanks!

Run the script and see what happens. Hint: I wouldn't be surprised if three records are inserted into one table, even if the word 'insert' appears only once in the script.

Not XP: if you're using TDD (which is an essential part of XP), you should write tests before you write actual code.
Not SCRUM: if you're using Scrum, you need a team. A team of one cannot realize many of the benefits of scrum.
Your description sounds exactly like Waterfall. That's a very respected methodology, even if it is not exactly 'Agile'; most of the successful production code that exists on the planet was written using Waterfall.
Best wishes,
Pablo.

Hilarious!
ROTFL!
You show a wonderful sense of humour. I mean, 'what is the best methodology for a team of 1?'. You've got my 5!

Then, I'm lost. Try the links in solution #2.
Good luck,
Pablo.

With DSN -> use the examples in the first link
Without DSN -> http://webcheatsheet.com/asp/database_connection_to_MySQL.php
Good luck,
Pablo.

Read this (top to bottom):
http://www.easysoft.com/developer/languages/c/odbc_tutorial.html
Particularly pay attention to the connection stuff.
Best wishes,
Pablo.

There are three options:
1. Use ADO.
2. Use OLE DB.
3. Use ODBC.

In each case, you create and open a connection object: to restore a database, you should connect to MASTER.
Then you build a string with the desired command.
At last, you execute the command.

Search MSDN for examples, there should be a few still lying around.

Hope this helps,
Pablo.

Did you try at
http://www.connectionstrings.com/mysql
?

Yes, the 'proc' folder shows a pseudo-folder for each process, and those folders have a file named 'commandline' or similar.
Those are text 'files' that have the full command line for the process.

Best wishes,
Pablo.

You create a process by running a program.
You can see the list of processes in the 'proc' system pseudo-folder (I think it's /dev/proc, I'm on a Windows machine at the moment).
Google for 'linux commands', spend a few hours reading, in the end you'll get the hang of it.
Best wishes,
Pablo.

You may.
I know you can manually assigning a work item to anyone, and add the name of the shelve set in the item history (with a 'please review' comment).
I don't know of an automatic way to do that, but manual may be OK if most people on your team follow the rules.
Hope this helps,
Pablo.

My 5!
Also:

cb.BackColor = Color.Transparent; // to show the toolstrip background.


As far as I know, there isn't. The wow6432node provides a 'virtual registry' for 32 bit processes, and they can't see the 64 bit areas. Even if you could DLL injection it would not help: the injected DLL would also be 'tricked into believing' that the wow6432node is the actual registry.
So, I don't see any solution but to place the information where the third-party application would look for it.

Hope this helps,
Pablo.

Actually, if you want to protect data in a file, you can (and should) encrypt the whole file to begin with. That's the standard way to protect data.

Delay loading may do the trick.

Did you try posting this question at the Oracle Technology Network?
Try starting here:
https://forums.oracle.com/forums/search.jspa?threadID=&q=.net+mono&objID=c84&dateRange=last90days&userID=&numResults=15

Do you have a reference to
<script src="/Scripts/jquery-1.6.4.min.js" type="text/javascript"></script>
?

NO YOU CANT!
If the subquery returns AT MOST one record, it may return zero records, which would generate an error.
If you know (in client code) that one and only one record will be returned, you also know the name of the city.
Best wishes,
Pablo.

If the subquery returns multiple rows, or zero rows, you're in a pickle. Using max() lets SQL Server know that there will be one and only one result.
If there are zero rows, max() will return null, which stays null after concatenation, and the outer query will return zero rows.
Otherwise, LIKE '%' + (zero rows) + '%' will generate an error.

Include xyz.h in your client application.

A file with the .asm extension is generated.

vs2008: right-click the project in 'solution explorer', navigate the tree to [configuration properties],[c/c++],[output files], and from the 'Assembler Output' combo choose 'Assembly with source code'.

I saw something at CodePlex, meant for VS2010, but there should be something similar. Something to do with schemas that you copy to some folder.
Did you try there?

Can you post some code? Possibly you're running into an ASSERT in the chromium DLL.
Are you using the latest version of the CEF?
Did you check the CEF issue tracker for similar problems? It's at http://code.google.com/p/chromiumembedded/issues/list
Hope this helps,
Pablo.

In Chrome, you can right-click the suspicious image, and ask for 'analyze element' (or something like that). It shows you the markup and CSS for the current element.
Did you try it? Didn't it help?

I thing you should consider using a temporary table for that purpose.
If you're coming to Oracle from SQL Server, note that Oracle temporary tables are different from SQL Server temporary tables in meaningful ways:
1. The tables are inside your database (no TEMPDB).
2. The table records are temporary, but the table structure is permanent: thus, it can be used in functions and procedures.
The similarity is that records are per-session.
A long, long time ago I had a similar need, creation of tables in a stored procedure, for the purpose of implementing 'the poor man's partition' (partitioning for customers with the humble editions of Oracle, back then partitioning was supported in the Luxury Gold Enterprise Edition, or whatever it was called).
Hope this helps,
Pablo.

No, those two scenarios are the only ones I've stumbled upon.
It seems that you'll have to search the registry, and see what you find...
Best wishes,
Pablo.

Possibly your evaluation version was more advanced that your production version: e.g., your evaluation version was SP1 or R2, and your production version is plain old sql2008.

And?
Try googling 'deadlock SQL Server'. It's how it works, its how it's supposed to work.

Please define 'held'.

I'd suggest checking the return values of the allocation functions (GlobalAlloc, GlobalRealloc, ...) for NULL. Lately, memory allocation functions are returning NULL more than they used to, particularly on virtual machines. Something to do with timeouts.

Pablo Aliskevicius - 7 secs ago
I'm sorry, I can't give you a snippet, since I don't own that code.
Basically, you'll have to play and see what happens.
Best wishes,
Pablo.

Reason for my vote of 5
Great question!

Why not?

Let's say you move to the West side of the Atlantic. When the 'system time' is noon, the 'local time' may be 8:00 am.
Also, the 'system time' does not provide for daylight savings. The 'local time' is the one you see in the Windows taskbar.
Best wishes,
Pablo.

You're right. Seen it now...
Thanks for the hint!

I'm a little aprehensive about functions like _localtime64(), that return pointers that it's not clear how or when they should be freed.
Just my two bits...

Different output every time is expected. Some encryption algorithms introduce random noise in the encrypted string; this makes them harder to break. The whole issue of encryption, with concepts like 'evidence' and 'enthropy' (which have, in the context of encryption, different meanings than their usual ones) has filled dozens of books.

Reason for my vote of 5
This can happen to anyone.

You say:
"super simple framework in C++ for a server and client"
IMHO, C++ is seldom super-simple; I use C++ mostly for heavy lifting. For RAD (Rapid Application Development), I'd suggest .NET.
Just my two bits...
Pablo.

The "Book" object is built (among others) of std::string members.
std::string may hold pointers to whatever, wherever (out of the Book instance), so when you save the Book, you're saving pointers.
Pointers are variables that hold an address in the computer's memory.
When you reload those pointers, you receive the same address again. But the contents of that memory may be now something else (if available).
See also solution #2.

Hope this helps,
Pablo.

Glad to hear.
Do you have a question?

"the target machine actively refused it" - means there is a target machine at 98.129.229.182. This is a good sign.
According to Wikipedia, Port 25 is SMTP (Simple Mail Transfer Protocol (SMTP)—used for e-mail routing between mail servers). This also looks good.
Sounds like a firewall issue. Are you using an in-house server? Can you check its firewall?

Off the top of my head, I can suggest two things:
1. Start your query with a double slash: XPath="//item.. insted of XPath="/item
2. Try debugging the XPATH part by part:

XPath="//item/description"
XPath="//item/description/InnerText()"
XPath="//item/description/[fn:contains(InnerText(),'IPL')]"
XPath="//item[fn:contains(/description/InnerText(),'IPL')]"
XPath="fn:contains(//item/description/InnerText(),'IPL')"

I'm a little rusty about how fn:contains works, and which namespaces you should import.
When you get it to work, please share the query!

Hope this helps,
Pablo.

Is the .mdf file, on disk, actually read only? This could be so if the file was read only when you packed your installer. For instance, if your .mdf file is under source control, and you checked it in before packing your installer.

Please clarify about "not working": is the event firing at all?
The function storeMouse, as posted, has a typo: does this typo avoid compilation?

You have onFocus and onBlur events in the client window, and you can handle them in JavaScript to set a property that would get submitted when your button is clicked. Would that help?

From your description, I assume that your UI is some kind of wizard, with the 'progress screen' near the end, and a BackgroundWorker in the outer frame, and that the 'manual update' runs synchronously in the UI thread.
I'd suggest putting another background worker instance in the form where the 'manual update' runs, and run also the 'manual update' in a background thread - of course, only if I guessed right about your architecture.
Hope this helps,

Did you test hr? Something like "if (FAILED(_hr)) _com_issue_error(_hr); "
This should help you identify the line where some action fails.

if (hRes==S_OK)...

Is this condition actually true?

This looks like a homework question. If so, please say so.

Which line is 222?
Are you using GCC?
What do you mean by <class dest_type="">? Shouldn't the default be a class name?

The shortest way would be exposing HTTPS endpoints only, no HTTP.
Would this be suitable?

You should get a string with "General\0Profile0\0\0". So, pSecNames + 9 should point to "Profile0".

Any reason not to let the user choose what to expand and what to collapse? That's what most people, IMHO, expect.

Reason for my vote of 5
Works like a charm: the pain is gone!

If the image is on paper, scissors may do the trick.

Reason for my vote of 5
Nice tip, Nir!

ReadFile will read the data in the file; you have to create CreateFile again, with different flags (GENERIC_READ only).
Try to play with the MSDN examples for ReadFile.
If what you want is data about the file (size on disk, date, ...) there are other API functions that may be of assistance.

See: http://msdn.microsoft.com/en-us/library/windows/desktop/aa365467(v=vs.85).aspx
and GetFileInformationByHandleEx (http://msdn.microsoft.com/en-us/library/windows/desktop/aa364953(v=vs.85).aspx)
Best wishes,
Pablo.

Is this a homework question? If so, you should mention that.
If not, it would be nice of you to mention how you happened to stumble upon this particular lines of code.

Are you sure you posted this question in the right place?

cout<<pin[i][j] % 2;

Is this what you wanted? The question is not clear.

Dear Abhinay,
Based on your question, I assume that you are new to C/C++. First and foremost, welcome on board!
Second, since you're on CodeProject since 2008, I assume you're not new to programming. In your profile, you don't say much about yourself, so anyone trying to help you will have to guess - I guess you have a background in either .NET or Java, or possibly Visual Basic.
If this is true, I can recommend a site where you can read about C++, you can find there a good book that is meant for programmers trying to learn C++ after knowing another language: http://www.mindviewinc.com/Books/downloads.html, the book names are "Thinking in C++, 2nd edition, Volume 1" and "Thinking in C++, Volume 2: Practical Programming". I have recommended those books to colleages coming to C++ from C and from C#, and they have found them interesting.
Of course, if my guesses are wrong and you have a background in C++, you should totally ignore this post.
Again, good luck,

Actually, the structure DLGTEMPLATE defines cx and cy in dialog units, so the pixel/dialog unit conversion is not the issue.
See:
http://msdn.microsoft.com/en-us/library/windows/desktop/ms645394(v=vs.85).aspx
Some manipulation of the window size in WM_INITDIALOG may help.
JM2B,

Ok, then I'd suggest you to write your own version of shared_ptr, writing something to the console in both the constructor and the destructor (for instance, the inner raw pointer and the reference count). Also, write to the console whenever your list is trying to release a pointer, and see whether they're released twice.

I also see that you figured a solution, if it works - keep it.

Hope this helps,

A dumb pointer it the opposite of a smart pointer: char * is a dumb pointer. Raw pointers are the same thing.
The reason why you're not using STL collections is relevant: if I was new to C++, I wouldn't try to improve on Stepanov's work. I don't want to get into much detail about why this is very relevant.

Good luck,

Try this. It may have some psichological influence ;)