More DirectScan with 2k/XP

[T-top87GN]

OK guys. I have been trying for a few weeks now to find a way to give a DOS based program direct access to the hardware ports on a Win2k/XP based system.

Good news is I have found a few. And they work great. Except with DirectScan.

I have found a few kernel level drivers that gives programs full 0 mode access to the hardware ports. I think that DirectScan has bigger problems however.

It looks like the DIrectScan problem may be more than just "getting access to the correct hardware ranges."

Being a DirectScan user this gives me a few options:

1) Get the source from Kent to modify it for use with newer Os's. (Tried that)

2) Reverse engineer the exe to try and get the source code. -- I've done this with other exe's before, but at least I knew what they were compiled with when I started!

3) "Listen" on the parallel port and write my own program that can decode the signal, like DirectScan does.


Unfortunately, it looks like the best way to do things is the latter. Seeing as how my GN is down now (prolly will be for a while) I figured that I would post my idea here to see if I can garner any help.

The program to be written would be GPL'd and source code available to anyone. I can't do this alone however. I'm going to need help from any of you that can help.

Please post here or e-mail me or PM or something if you are interested.

This is going to be a HUGE undertaking. It may, or may not ever get done. But I am sure gonna try.

 

[T-top87GN]

No takers. That's ok.

I guess I'll be on my own!

 

[Metal]

How about dual booting the laptop or booting it with an ms dos bootdisk?
Hell you could even get a cheap Pentium Thinkpad for about 100 bucks.
Keep in mind I don't have direct scan but seems to me you could "make' a system run dos easier than the "HUGE undertaking" your contemplating.

 

[Scott231]

Originally posted by Bandit
No takers. That's ok.
I guess I'll be on my own!

Hey! I e-mailed you butthead!

 

[T-top87GN]

Originally posted by Metal
How about dual booting the laptop or booting it with an ms dos bootdisk?
Hell you could even get a cheap Pentium Thinkpad for about 100 bucks.
Keep in mind I don't have direct scan but seems to me you could "make' a system run dos easier than the "HUGE undertaking" your contemplating.

I have a low end laptop that runs 98 so I can run DS. I also have a dual booting XP and Linux laptop that I use as my main.

My point is I want to run DS in windows. Period. Hell, I may even add a nice GUI to it. Let it plot out graphs for me, who knows.

The main obstacle is interpretting the data stream. Once I get that down, the programming part should be the easy part. SHOULD be, we'll see

Yeah Scott, I got your e-mail. I'll get with you on this soon.

 

[EluZioN]

You've got mail

 

[salvageV6]

Get that data stream.

I want to try some stuff with Labview software and a data logger from Fluke. Have to mix Direct Scan stuff in with the external data logger stuff on a nice front panel display.

Not sure what help you need but I have some time.

 

[QuickWrench]

Fiiiiiinnnallly....

I hunted Kent down years ago for the DS source (no go...). Looks like things haven't changed much.

DEFINITELY keep me in mind on this. I'm a C++/C# developer. I kind of gave up trying, since I figured that all I'd really be able to come up with in the long run is WinDirectScan. Just a better GUI, wrapped around the existing functionality of Kent's 'black box'.

I haven't really pursued the decompiler route just yet, but FYI...it's compiled with an older version of Borland C++. Version 1.52 must have been compiled with a newer version, since the executable doesn't have the obvious Borland version # embedded like the previous versions did. Open version 1.51 (if you have it) in Notepad. You'll see.

I've written all kinds of extended viewers to look at .csv's exported from DS. 2D/3D Graphing, trending, etc.

All we really need at this point is an ATL COM object (or whatever) to provide an API for the parallel port I/O. I've got a kick-@ss GUI already done, I just haven't convinced myself to go through all of the BS to do figure out the DS I/O. But...now that some other people are on board, it may not be a big deal.

'later.

 

[kenmosher]

Just a word of caution ... you might want to read Kent's software license.

It may not allow for reverse engineering ... this is the stuff that some of the lawyer types may pursue.

 

[Dean]

Just what I was thinking, Ken.

Jeremy, I would think you need to get Kent's blessings on this. If BTW, I knew squatola about this kinda stuff, I would be happy to help... once Kent gave the thumbs up.

 

[salvageV6]

Once the stream is known you dump the old code. Hope you don't need permission to get rid of it.

You are just using the hardware to write some good new code and updated software to make a better display and log more data.

How somone finds the data stream could be simply monitoring the parallel port on their computer, at least that's the excuse I would use.

 

[QuickWrench]

Once the stream is known you dump the old code. Hope you don't need permission to get rid of it.

You are just using the hardware to write some good new code and updated software to make a better display and log more data.

How somone finds the data stream could be simply monitoring the parallel port on their computer, at least that's the excuse I would use.

Yup...agreed.

 

[T-top87GN]

Well as long as I don't decompile his executable we should be ok. I don't see any law that tells me I can't listen on the LPT port of my own machine.

It's all a matter of decoding the data streaming in on the LPT port. Once we get that figured out, writing some Windows based software to interpret the stream should be no problem.

I just need more frickin time to get this done! heheh Always sooo busy.