Difference between revisions of "Talk:Securely wipe disk"

From ArchWiki
Jump to: navigation, search
(Removed timeout recommendation for shred: re: thx for the links)
(Restructured Data Remanenced and removed bullshit hype: new section)
Line 22: Line 22:
 
:::::I added prompt to my script now. Thanks for the links!
 
:::::I added prompt to my script now. Thanks for the links!
 
::::: - Andy Crowd 19:10, 7 January 2015 (UTC)
 
::::: - Andy Crowd 19:10, 7 January 2015 (UTC)
 +
 +
== Restructured Data Remanenced and removed bullshit hype ==
 +
 +
Hey guys, its always bothered me that everyone freaks out about data recovery, so I did a little(a lot) research to see what exactly is and is not true. TL;DR: It's virtually impossible.<br>
 +
I cut down and re-factored the section so its more readable. I'm basically treating this section as an explanation behind securely wiping data, which it kinda already was, since the rest of the article goes over the ''how'' of doing it.
 +
[[User:Swashy|Swashy]] ([[User talk:Swashy|talk]]) 20:28, 7 January 2015 (UTC)

Revision as of 20:28, 7 January 2015

wipe script to test

My wipe disk scripts body is completely done, any one can add own wiping patterns there now or uncomment some of existing. - Andy Crowd 17:02, 9 December 2014 (UTC)

Hi, I have had no opportunity to test it yet, but maybe others do/did. I'm moving your comment linking to the script to a new talk item with this edit, so that we can remove the closed discussion (Talk:Securely wipe disk#Wrong Description.3F) which mainly focussed on the existing article content. --Indigo (talk) 13:03, 13 December 2014 (UTC)

Removed timeout recommendation for shred

I removed the recommendation to invoke the shred command with timeout 3. I don't believe this will benefit anyone. If you mistype the shred target it should be plainly obvious that something went wrong. Three seconds is an arbitrary and useless amount of time, and you'd probably instantly recognize that you messed it up, especially given that shred is called with the verbose option. Additionally, if you accidentally call shred on the wrong device or partition the data is usually instantly compromised, and the three second timeout will do nothing to help you. Furthermore, aliasing timeout impedes calling shred on large files. —This unsigned comment is by Freebullets (talk) 4 January 2015 7:37. Please sign your posts with ~~~~!

Thanks for explaining, fine with me. For completeness: the edit you reference is [1] and the content removed got moved here with [2]. It may be that the removed alias part was meant to make it escape when you use shred to wipe individual files rather than a device. In this case I see it may be helpful to timeout, but it is not the primary target of this article. Let's wait for other opinions just in case.
One comment to your editing: For editing a subsection you can use the edit button next to its header instead of the edit button top of page, which is used to edit the article intro. The advantage is you get the automatic reference of the subsection in the history. Moreover there is less risk of edit conflicts while you do the edit. --Indigo (talk) 08:36, 4 January 2015 (UTC)
I was also thinking about this and I am agree about of removing "timeout" but it was actually something new to me with the "timeout" command. Might be used to pre-shred or for limited shred of a part of something...
-- Andy Crowd 19:36, 4 January 2015 (UTC)
I too agree with the removal for the mentioned reasons. If we really want something like that in the article, a simpler alternative would be to use shred's native -s option with a low value in the alias, if as Indigo says it's meant to be used for single files. When shredding entire devices is needed, the user would be forced to comment or override the alias. Another alternative would be to use functions like [3], [4] or [5]. If using a confirmation function, it could use du to show the size of the target. Or, again, we could make a one-liner that first checks that the size of the target is less than something, and then it starts shredding, otherwise it exits/asks for confirmation... and if I don't close this page now I'll probably keep dumping ideas, so I'll move on to the next discussion :) -- Kynikos (talk) 03:47, 5 January 2015 (UTC)
Spot on links. Would be something for Andy Crowd to use in his delete test script linked in above talk, perhaps. I think it is pretty certain the removed [6] bash trick we talk about was meant for files, because in this case you don't need #. You need that for a device per default; another reason the trick made no sense in this article. IF indeed someone uses such an alias for /root .. hiding history from her/himself, oh well .. "good luck Arch install". --Indigo (talk) 14:25, 6 January 2015 (UTC)
I added prompt to my script now. Thanks for the links!
- Andy Crowd 19:10, 7 January 2015 (UTC)

Restructured Data Remanenced and removed bullshit hype

Hey guys, its always bothered me that everyone freaks out about data recovery, so I did a little(a lot) research to see what exactly is and is not true. TL;DR: It's virtually impossible.
I cut down and re-factored the section so its more readable. I'm basically treating this section as an explanation behind securely wiping data, which it kinda already was, since the rest of the article goes over the how of doing it. Swashy (talk) 20:28, 7 January 2015 (UTC)