Verbose Output -v62 does not include data

[hex-dreamer]

Very interesting project - thanks Quentin!
Been using CW and find the -v62 useful for including the data from every read of every sector -u logfile.txt
On a damaged disk this is useful to help build up a damaged sector from multiple reads - I cut and paste from the logfile.
I tried gw2dmk beta version just now, and the -v62 does not include the actual sector data in the -u logfile.txt - at least on my system. It does produce the dmk file so I know it reads the disk ok.
Using GW4 with FW1.6 with gw v1.21 installed, Ubuntu 18.04 64bit Python 3.8.
Just let me know if you need me to run some tests etc or further info.
Thanks again for a very interesting project.

[qbarnes]

You're right!

For some reason when I ported that code from cw2dmk, I turned off that feature at the time. I'm not sure why yet:
https://github.com/qbarnes/gw2dmk/blob/v0.0.1.beta.1/gwdecode.c#L858

I'll go back and check my notes to see if I had a reason. I suspect it was because the way the verbosity level works from cw2dmk with some levels not be a true addition of its previous levels. I think I was thinking of just removing this level because of the weird level hackery.

In the mean time, you could always use level 7. You just get the histogram too.

I've got a bunch of stuff in my pipeline at the moment, so I'm not going to be turning out a new release soon. If you're comfortable building C code, you could uncomment the line, build, and give it a try and tell me how it goes.

Why would you want to use this level instead of level 7 (or level 5) which gives you the hex version of each byte?

Do you have a Catweasel too by chance? Mine's currently broken, so I can't do a compare and contrast.

[hex-dreamer]

Oh wow - many thanks for your very fast reply. It is appreciated.
The truth is I never seem to be very good at compiling code from other people's projects - probably about a 50% success rate - I always seem to get errors thrown! But I shall certainly try :)

I used level 6 just because the output gives the byte chars which makes it simple for me to cut and paste into the final disk image , as opposed to listing the hex values (which I think level 7 does)
I mean I could write a small script to convert each back hex code to a byte, so it is not impossible to parse out the data and work with that.

Yes! I do have a catweasel Mk4 pci card so am happy to help testing if you need it. Unfortunately I only have the one otherwise I could ship you one. What is broken on yours? It might be worth looking at / replacing say the electrolytic capacitors or Oscillator crystals depending on your issue.

[qbarnes]

I think I can up your percentage! I make it as trivial as I can because a lot of my gw2dmk users aren't devs. Assuming you're on Linux, just clone the repo, cd into the directory, and type make. If any tools failed to find, check BUILDING.md. Depending on your distro, it has a one or two line command to install the packages you need, but given you're a dev, they're likely to already be installed. The build should work and then you'll find build/gw2dmk to play with.

Then change //msg(MSG_RAW, "%c", val); to msg(MSG_RAW, "%c", val); and type make again. The build/gw2dmk will now be updated with your change.

which makes it simple for me to cut and paste into the final disk image , as opposed to listing the hex values (which I think level 7 does)

Don't you have the problem that non-ASCII characters aren't captured in an L6 log? Do you just not care about them?

Yes! I do have a catweasel Mk4 pci card so am happy to help testing if you need it.

Yay! If you can, on a given disk, run cw2dmk 4 times collecting a level 4, 5, 6, and 7 log to files, and then send me the 4 files and the DMK. That would help tremendously.

What is broken on yours? It might be worth looking at / replacing say the electrolytic capacitors or Oscillator crystals depending on your issue.

I'm getting overtones from the clock circuitry causing nearly a 4X oversampling problem. I've replaced the crystal. No change. I've worked with Jens (the creator of the card). He gave me some ideas for modifying the circuitry by adding a couple of 15pF caps to the oscillator. I have the caps, but just I haven't tried that yet.

[hex-dreamer]

I think I can up your percentage!
Perfect - I will give all that a go :)

Don't you have the problem that non-ASCII characters ...
I should say I cut and paste using a hex editor (or a script to chunk the data out to the final disk image)
that way none ascii characters are not lost.

If you can, on a given disk, run cw2dmk 4 times ...
Sure - will do that.!
Again many thanks for your time and replies. They are appreciated.

As for the problem with the card, it does sound like a cap problem not filtering. But I am only amateur and guessing! Best of luck with it.

[hex-dreamer]

Well well. Compiling the modified code worked :)
-v62 output restored. Thanks very much!
At least for the quick test I did on simple ibm 3.5inch 1440 disk it did what I expected
(at least at a brief glance - it is late here now in UK - I have to review in more detail later)

I had trouble (permission denied error messages) when trying to clone the repo as outlined in BUILDING.md
. So I downloaded zip file and unzipped to a folder.
Also during the "make" I had many Warning Fatal error messages ( maybe because it is making from the zip rather than clone of a repo?)
However a build folder was created with .o and executables in it for Linux which is more than I expected I would achieve today.
So thanks again very much for your time and excellent work.

Will get to the other stuff - hopefully later this week.

[hex-dreamer]

cw2dmk logs.zip
Attached requested catw files - all from same disk.

[qbarnes]

I had trouble (permission denied error messages) when trying to clone the repo

I have no idea why you had this problem. I suspect you'd also have the exact problem attempting to also clone any other public repo in github. Do you clone github repos often? Maybe you didn't have your github registered ssh key loaded and you tried to clone with git@github.com:qbarnes/gw2dmk? Maybe try https://github.com/qbarnes/gw2dmk.git instead?

Also during the "make" I had many Warning Fatal error messages ( maybe because it is making from the zip rather than clone of a repo?)

Nope, no problem from zipped source. When I adding man page support recently, I forgot to update the directions for installing the packages for generating the PDFs. BUILDING.md has been updated. Thank you for reporting that problem!

cw2dmk.logs.zip

Snagged! Thank you! This will help a lot with my testing!

[hex-dreamer]

Just to let you know this error log with CRC error data is fantastic.

I find GW2DMK reads disks slightly differently and finds data that gw, fluxengine, catweasel don't find on problem disks.

The fact that the CRC error data goes in the log really helps with recovery of text. Even if the sector is bad, often some text in it is good enough for an author who vaguely knows the original manuscript to complete.

Just to say again - many thanks for showing me how to compile with -v62 option as I process the text log for Bad CRC Data sectors to help complete manuscripts on damaged disks.

I also have a small favour - I have compiled -v62 for linux, but don't have success compiling a .exe. for Windows. Any chance you could run it for me. gwdecode.c - just has to be unhashed at line 858

I know you don't want to release the -v62 enabled version into the wild, so could I download .exe from say a google drive if you get chance to compile it?

[qbarnes]

I'll see about kicking out a new release in the next few days with that change. I need to review some other changes I've made to see if I want to pull them in too as well.

Why couldn't you get the MS Windows version to build for you? I'd like to fix my BUILDING.md instructions to address that issue. Could you fill me in on your platform, the approach you used, and what step and how it broke for you?

[hex-dreamer]

Many thanks for your reply. I am building in Fedora 42 and will post results again here for feedback. Usually the problem is simply me rather than the machine:)

[hex-dreamer]

Your instructions were perfect - I built on Fedora Workstation 42.
It turned out my problem was the Fedora 42 distro I first used had "rpm-ostree" rather than "dnf"
Once I got the dnf version everything went smoothly.
Thanks again for your time on this issue which, as normal, was my fault :)