Empty document on SDI App - C / C++ / MFC Discussion Boards

Roger Allen6-Jun-02 4:37

6-Jun-02 4:37

I used a DSP56001 board many years ago. (Like when the 186 was the top dog). It was good and easy to use at that time, considering I hadn't done my programming degree at that time. (that makes it 12 years ago OMG | :OMG:

)

Sorry if I can't be more help.

Roger Allen
Sonork 100.10016

If I had a quote, it would be a very good one.

Re: DSP Boards

Brian Azzopardi6-Jun-02 4:52

Brian Azzopardi

6-Jun-02 4:52

10x Roger.

Any idea anyone where I can get benchmarks/performance figures. Ideally I'd like to know whether the bang/buck of a DSP compared to a 2Ghz AMD is better, etc and any pitfalls. I couldn't find any perf data on the Texas Instruments site.

close view but leave document open

Gary Kirkham6-Jun-02 4:28

Gary Kirkham

6-Jun-02 4:28

I have am writting an MDI Doc/View app and have a situation where I would like to be able to close a view (there is only one view and it is derived from CScrollView) and leave the document open...then, sometime later, create a new view and attach it to the still open document.

How do I do this?

Re: close view but leave document open

Tomasz Sowinski6-Jun-02 4:35

Tomasz Sowinski

6-Jun-02 4:35

To avoid deeper hacking, I'd create a 'dummy' frame/view and hide it. So technically a view would exist - but user would be unaware about that.

Tomasz Sowinski -- http://www.shooltz.com

- It's for protection
- Protection from what? Zee Germans?

Re: close view but leave document open

Gary Kirkham6-Jun-02 4:49

Gary Kirkham

6-Jun-02 4:49

Thanks for replying...I thought about that, but there can be many documents/views being opened and closed and if they are not really closed (only inactive), the inactive views continue to receive OnUpdate messages. I am worried about a possible performance hit...It seems that it would be cleaner to close the view, but leave the document open.

Re: close view but leave document open

Tomasz Sowinski6-Jun-02 4:52

Tomasz Sowinski

6-Jun-02 4:52

Gary Kirkham wrote:
I am worried about a possible performance hit...

Well, the dummy view would be just a class derived from CView which ignores OnUpdate calls.

You may also try to dabble with undocumented CDocument::m_bAutoDelete. Probably if you set it to FALSE closing last view will not destroy document. I've never played this trick personally, however.

Tomasz Sowinski -- http://www.shooltz.com

- It's for protection
- Protection from what? Zee Germans?

Re: close view but leave document open

Gary Kirkham6-Jun-02 15:37

Gary Kirkham

6-Jun-02 15:37

Never mind I figured it out...you gave me part of the answer (m_bAutoDelete)

Thanks

Cool | :cool:

Empty document on SDI App

6-Jun-02 3:51

At the begining of Application SDI, I want a Empty Document but when I write this code :
// no empty document window on startup
if (cmdInfo.m_nShellCommand == CCommandLineInfo::FileNew) {
cmdInfo.m_nShellCommand = CCommandLineInfo::FileNothing;
}
on App::InitInstance(), it doesn't work ...
Please Help me !

Re: Empty document on SDI App

Tomasz Sowinski6-Jun-02 4:10

Tomasz Sowinski

6-Jun-02 4:10

Check this link:

http://www.microsoft.com/msj/defaulttop.asp?page=/msj/archive/s204d.htm

Tomasz Sowinski -- http://www.shooltz.com

- It's for protection
- Protection from what? Zee Germans?

Synchronization problem (how do I make it read-only?)

6-Jun-02 3:11

Hey all,

I have a shared resource in my application. I want to allow all running threads to READ from that resource simultaneously, but want them all to block when the resource is being WRITTEN to.

I know there's a standard technique to do that, but I can't remember it, and I have no idea what keywords I should use in the search engines (I do recall that this problem has a formal name).

Thanks!

Re: Synchronization problem (how do I make it read-only?)

6-Jun-02 3:19

MRSW lock (Multi-reader, single writer). You can find a decent implementation at http://www.quality.nu/dotnetguy/2001/21/10.aspx.

Code to change data from Big Endian to Little Endian

JMajors986-Jun-02 3:19

JMajors98

6-Jun-02 3:19

Hello,
I am trying to read a binary file written on a big endian machine(SGI) and read it on my little endian(PC). Does anyone have code or any thoughts on how to do the byte conversion? I would appreciate and help I can get. Thank you for your time.

Josh

Dead | X|

Re: Code to change data from Big Endian to Little Endian

6-Jun-02 3:16

I think you should be able to use ntoh* functions to do the switch, since network order is big endian.

Re: Code to change data from Big Endian to Little Endian

Joaquín M López Muñoz6-Jun-02 3:46

Joaquín M López Muñoz

6-Jun-02 3:46

This should work:

template <typename T> T change_endian(T in)
{
  T out;
  char *pin=reinterpret_cast<char *>(&in);
  char *pout=reinterpret_cast<char *>(&out)+sizeof(T);
  for(size_t n=0;n<sizeof(T);++n)*--pout=*pin++;
  return out;
}

Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo

using SAFEARRAY

RaphaelBr6-Jun-02 2:24

RaphaelBr

6-Jun-02 2:24

Hi, I am using a OCX component, this component return to me a SAFEARRAY Pointer, I am sending this array using a CAsyncSocket component .. like this:

SAFEARRAY **psa;
psa = Data->pparray;
m_UDPSocket.Send((*psa)->pvData, (*psa)-rgsabound[0].cElements);

And I get in the other side like this:

int hr;
SAFEARRAY psa;
BYTE pdata[MAX_BUFFER];

psa.pvData = pdata;
psa.cDims = 1;
psa.cbElements = 1;
psa.cLocks = 1;
psa.fFeatures = FADF_FIXEDSIZE | FADF_AUTO | FADF_EMBEDDED | FADF_STATIC
psa.rgsabound[0].lLbound = 0;
psa.rgsabound[0].cElements = hr;

it´s work well ... but I need to send 2 bits together, I need to represent the number 1 (01), or 2 (10), or 3(00), I want to represent what kind of data I am sending using the Socket. How can I link the array with this 2 bits ?

Thanks a lot

georeferenced tiff files

jpg6-Jun-02 2:16

jpg

6-Jun-02 2:16

For mapping software, tiff files can contain positioning information, such as coordinates of the lower left corner, width and height in meters, ... Sometimes this
information is included in a header in the tiff file itself, sometimes it is contained in a separate .tfw file. Has someone some clues for getting this information ?

JPG

Re: georeferenced tiff files

C^hris L^osinger6-Jun-02 2:55

C^hris L^osinger

6-Jun-02 2:55

do a google search for LibGeoTiff

-c

Cheap oil. It's worth it!

ISEffects - effects for images

stl iterator

_Magnus_6-Jun-02 2:10

_Magnus_

6-Jun-02 2:10

A simple question.

Say you have.

iterator itBegin = whatever.begin();
iterator itEnd = whatever.end();

and you after that adds more stuff to 'whatever'
will 'itEnd' point to the real end or will it point
to what was the end when you got it?

/Magnus

Re: stl iterator

C^hris L^osinger6-Jun-02 2:57

C^hris L^osinger

6-Jun-02 2:57

in general, any operation that changes the size of a container will invalidate all iterators. you will have to call whatever.end() after changing the container, if you want to know where the end is.

-c

Cheap oil. It's worth it!

ISEffects - effects for images

Re: stl iterator

Joaquín M López Muñoz6-Jun-02 2:56

Joaquín M López Muñoz

6-Jun-02 2:56

Depends. On some containers, like vector, any iterator becomes invalid (or is not guaranteed to remain valid) after adding/removing data. For lists, on the other hand, iterators remain valid (unless you erase the very object the iterator is pointing to, of course.) Check the docs for more info on each particular container, this kind of things are thouroughly specified.

Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo

Re: stl iterator

_Magnus_6-Jun-02 3:29

_Magnus_

6-Jun-02 3:29

In this case i was using a map and i was thinking about doing something like this

iterator itBegin=map.begin();
iterator itEnd=map.end()
for(it=itBegin;it!=itEnd;++it)
{
AddToMap();
}

to add to the map but not loop trough the new added items.

Joaquín M López Muñoz wrote:
Check the docs for more info on each particular container

That is the one thing i hate about STL, the docs are completely useless.

/Magnus

Re: stl iterator

John M. Drescher6-Jun-02 6:28

John M. Drescher

6-Jun-02 6:28

I believe that this won't work. For the following reason: most implementations of map store items in sorted order so adding items
to the map will probably be between begin and end. Unless you are adding things that are bigger than what is in the map.

John

Re: stl iterator

Mike Nordell6-Jun-02 10:22

Mike Nordell

6-Jun-02 10:22

From the docs for map::insert

inserting an element invalidates no iterators, and
removing an element invalidates only those iterators
that point at the removed element.

If your AddToMap is adding to this map, you will get those elements back also. What are you really trying to accomplish here? It seems you want to double the size of the map, generating new keys and values based on the previous ones. I don't think I've ever seen such an algorithm before.

But to solve your problem you could create another map object that was a copy of this map and then add to the old map based on this new one. Like:

std::map<whatever> map2(map.begin(), map.end());<br />
iterator itBegin=map2.begin();<br />
iterator itEnd=map2.end();

use these iterators to insert stuff into the original map object.

It's probably not what you really want, but it should do what you explained you want. Smile | :)

Re: stl iterator

John M. Drescher6-Jun-02 10:21

John M. Drescher

6-Jun-02 10:21

Mike Nordell wrote:
What are you really trying to accomplish here? It seems you want to double the size of the map, generating new keys and values based on the previous ones. I don't think I've ever seen such an algorithm before.

Yes. This looked very strange to me. Also map will not allow you to enter the same key in the map more than once so I hope there is some translation..

John

SysListView32 (Desktop)

[xan]6-Jun-02 2:01

[xan]

6-Jun-02 2:01

How can I put my window transparently on desktop (just like Explorer does) using SysListView32 ?

Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages.