Mailing List Archive

http://bugzilla.mindrot.org/show_bug.cgi?id=896





------- Additional Comments From dtucker@zip.com.au 2005-02-11 13:43 -------
Created an attachment (id=818)
--> (http://bugzilla.mindrot.org/attachment.cgi?id=818&action=view)
allow maximum utilization of input buffer

I've changed my mind: allowing input buffer to be used to its absolute maximum
(eg attached patch) doesn't make much sense. 10MB+ of buffered input is a bit
much...



------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=896


dtucker@zip.com.au changed:

What |Removed |Added
----------------------------------------------------------------------------
Attachment #685 is|0 |1
obsolete| |
Attachment #818 is|0 |1
obsolete| |




------- Additional Comments From dtucker@zip.com.au 2005-02-11 13:55 -------
Created an attachment (id=819)
--> (http://bugzilla.mindrot.org/attachment.cgi?id=819&action=view)
Buffer only a reasonable number of packets' worth of input data

This patch allows buffering of up to 1MB of input data (equal to 32 default
sized or 64 max sized SSH packets).

Setting the limit too low may reduce the effectiveness of compression on highly
compressible data, and setting it high may result in excessive memory usage.
Compressing 1MB of zeros with "gzip -6" reduces to ~1kB so this seems a
reasonable compromise.



------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.