Easy md5sum checking

[asqwerth]

that's what I usually use

I do as well, looking only at the last 4 digits.

Same here, which is why I'm quite happy to see md5sums displayed on distro download pages and immediately accessible to my eyes.

[lucky9]

I'm an old stick-in-the-mud type here I guess. The MD5sum goes into the same folder as the .iso. And I check all digits even though I'll readily admit that a bad .iso file has always shown a bad MD5sum in the first digit (usually). Certainly in the first few in my experience. But I always verify my burns also. So I'm certain that I have a good copy of what was on-line.

Back when I used to run a certain Sid-based distro the devs there would blame bad burns even with a verified MD5sum and a verified burn. I decided those people didn't know what they were talking about. Mind you it could have been an optical drive that got dirty after the fact of making the .iso. But they had no idea of how modern DVD burners worked (obviously) when they claimed a bad burn after the MD5sum was checked and the burn was verified. You do need to keep a drive clean, or at least clean it if anything is happening that's weird (as a first step in troubleshooting).

[TJ Hoye]

The "Check md5sum" Thunar Custom Action in the article that Danum linked to: http://pclosmag.com/html/issues/201309/page06.html doesn't require any copying or pasting, you just right-click on a .md5sum or .md5 file.

checkmd5sumaction1.png

It md5sums the iso file specified in the .md5sum or .md5 file and gives a "OK" or "FAILED" indication

checkmd5sumaction2.png

I substituted 'yad' for 'zenity' in the command, and set the file pattern to "*.md5;*.md5sum" instead of just "*.md5sum".

It does assume that the .iso and .md5sum (or .md5) files are in the same directory.

Thank you for this. It's what I was looking-for.

The Arnote article shows how to implement command-line functions of md5sum, both of them, with guis.
kmathern has adapted that for his use with MX or Mepis.
Thanks also to danum for giving us the Arnote reference.

Jerry3904 and asqwerth got this thread rolling using gtkhash implementation which I think falls short
of the Arnote approach for all the reasons previously presented.

The Arnote approach uses md5sum behind the scenes in two separate guis rather than gtkhash in its implementation,
if I understand it correctly.

[richb]

I admit I do not check mdsums unless I have a problem. I have not yet. That said I did implement the "Danum" solution for KDE.

[malspa]

I always run md5sum and check each digit, even though it would be a shocker to see only one digit be off. I'm not sure why I'm so meticulous about it, but that's how I am.

[richb]

I always run md5sum and check each digit, even though it would be a shocker to see only one digit be off. I'm not sure why I'm so meticulous about it, but that's how I am.

Your policy is better than mine.

[uncle mark]

I rarely (never) have a need to copy or paste the output results of md5sum when verifying the hash code of a file because I don't check the entire 32-character string; I only eyeball the first 4 or the last 4 characters. If those match the published results, I assume the file is good. Why? Because I assume that the md5sum will be completely different if only one byte of information in the file is missing or corrupted. Or am I way off base with that assumption?

I'm with Joany. Life's too short.

[asqwerth]

.....
Thank you for this. It's what I was looking-for.

The Arnote article shows how to implement command-line functions of md5sum, both of them, with guis.
kmathern has adapted that for his use with MX or Mepis.
Thanks also to danum for giving us the Arnote reference.

Jerry3904 and asqwerth got this thread rolling using gtkhash implementation which I think falls short
of the Arnote approach for all the reasons previously presented.

The Arnote approach uses md5sum behind the scenes in two separate guis rather than gtkhash in its implementation,
if I understand it correctly.

I have to thank Danum and kmathern as well. The suggested method in the pclosmag link does work very well. That is something new I learnt.

Some distros may provide sha1/sha256 sums instead of md5 (granted they often provide all 3). So you'll need separate custom thunar actions for each type of hash sum file (which is one reason why I like gtkhash's 'one-click to compute all sums' approach).

For me, I still like having the hashsums displayed on the download page since I visually compare just a part of the whole sum. Proper .md5 files can be uploaded for those who want it, but for MX download page, please keep the md5sum up there!

[TJ Hoye]

I have to thank Danum and kmathern as well. The suggested method in the pclosmag link does work very well. That is something new I learnt.

Me, too.

FWIW, I hardly ever check md5sums myself.
Considering all the work to set up some guis, I'll probably still use the command-line if I ever do a checksum.
However, it would be nice if someone made a Debian package that brings the Arnote material in via Synaptic.
I'd use that gui. Otherwise, being lazy, I'd use the command-line check.

I hope this has been a useful learning experience for all concerned.
It has been for me as well.

[Adrian]

I don't know if this was mentioned somewhere in the post, but MX-14 has a way to test the md5sum of the initrd, linuxfs, vmlinuz (the most important files on the ISO) when you boot F4 -> Check md4sum, I found that convenient.