Is Entropy polling advisable?

[Txv2]

(I use the "stable" 5.86a version of Eraser). Is the "background entropy polling" option useful, because it provides an extra layer of pseudorandom noise, or unnecessary, because it can make the wipingprocess slower and XP unstable, when I use it for a pseudorandom free space wipe? I have read there where some problems with it in the past, is this why it's disabled by default?

 

[Joel]

It's not really important since the default already includes background entropy gathering. Things like cursor/mouse position are already gathered. The slow one adds even more (a few hundred times more) information by calling NtQuerySystemInformation (IIRC) and that gathers alot of information. Also disk statistics are queried in v5, though that is of questionable value since it may slow down performance.

Joel

 

[Txv2]

OK, so a regular free space pseudo-1pass wipe, with Eraser's default settings, should create enough entropy and is therefore sufficient.


Thanks

 

[Overwriter]

I have read there where some problems with it in the past, is this why it's disabled by default?

I find that when I enable background entropy polling it locks my computer.

OK, so a regular free space pseudo-1pass wipe, with Eraser's default settings, should create enough entropy and is therefore sufficient.

I personally think its ok for purposes of deletion. However I am looking forward to the implementation of a CSPRNG to enable plausible deniability for Truecrypt users.

 

[Joel]

ISAAC is a CSPRNG already, just that people have yet to prove or disprove it.

Joel