URL:
http://gna.org/bugs/?21300
Summary: Network compression can send packet which receiver
interprets incorrectly
Project: Freeciv
Submitted by: jtn
Submitted on: Sun Nov 24 14:41:49 2013
Category: None
Follow-up Comment #1, bug #21300 (project freeciv):
A related problem is that compressed data of size 49149 or 49150 would I think
overflow the packet length field, causing a Trying to put %d into 16 bits
warning on recent branches, and different comedy on the receiver (on any
branch). This is
URL:
http://gna.org/bugs/?21301
Summary: Possible network stall with short packets at start
of connection
Project: Freeciv
Submitted by: jtn
Submitted on: Sun Nov 24 15:51:14 2013
Category: None
Severity:
URL:
http://gna.org/patch/?4313
Summary: Diagnose too many calls to conn_compression_thaw()
Project: Freeciv
Submitted by: jtn
Submitted on: Sun Nov 24 15:52:39 2013
Category: None
Priority: 5 - Normal
URL:
http://gna.org/patch/?4314
Summary: Network compression for big map updates
Project: Freeciv
Submitted by: jtn
Submitted on: Sun Nov 24 15:56:42 2013
Category: None
Priority: 5 - Normal
Follow-up Comment #1, patch #4314 (project freeciv):
(I do intend to commit this to S2_4, but after 2.4.1 so that we get time to
test it.)
___
Reply to this item at:
http://gna.org/patch/?4314
URL:
http://gna.org/patch/?4315
Summary: Avoid sending compressed packet that would be bigger
than uncompressed one
Project: Freeciv
Submitted by: jtn
Submitted on: Sun Nov 24 15:58:52 2013
Category: None
URL:
http://gna.org/patch/?4316
Summary: Add assertions for invariants and assumptions of
network compression code
Project: Freeciv
Submitted by: jtn
Submitted on: Sun Nov 24 16:00:33 2013
Category: None
Update of patch #4313 (project freeciv):
Planned Release: 2.5.0,2.6.0 = 2.4.2,2.5.0,2.6.0
___
Reply to this item at:
http://gna.org/patch/?4313
___
Message sent
Update of bug #21297 (project freeciv):
Status:None = In Progress
Assigned to:None = jtn
Planned Release: = 2.4.2,2.5.0,2.6.0
Follow-up Comment #3, bug #21297 (project freeciv):
I was able to trigger this on Ubuntu with FREECIV_COMPRESSION_LEVEL=0, even
when taking extra care to not emit more than MAX_LEN_BUFFER.
Reducing the queue max to MAX_LEN_BUFFER/2 or even /4 doesn't seriously reduce
the level of compression
Update of bug #21300 (project freeciv):
Status: In Progress = Ready For Test
___
Additional Item Attachment:
File name: trunk-S2_5-S2_4-S2_3-net-compress-jumbo.patch Size:1 KB
Update of patch #4313 (project freeciv):
Status: In Progress = Ready For Test
___
Additional Item Attachment:
File name: trunk-S2_5-S2_4-net-compress-thaw-assertion.patch Size:0 KB
Update of bug #21297 (project freeciv):
Status: In Progress = Ready For Test
___
Follow-up Comment #4:
(Also noticed in passing that conn_compression_flush() can send
a compressed packet that's
Update of patch #4314 (project freeciv):
Status: In Progress = Ready For Test
___
Follow-up Comment #2:
As noted elsewhere, a big map revealed in its entirely is reduced to 20% of
the previous
Update of bug #21301 (project freeciv):
Status: In Progress = Ready For Test
___
Additional Item Attachment:
File name: trunk-S2_5-net-accept-small-packets.patch Size:0 KB
Update of patch #4315 (project freeciv):
Status: In Progress = Ready For Test
___
Reply to this item at:
http://gna.org/patch/?4315
___
Message sent
Update of patch #4316 (project freeciv):
Status: In Progress = Ready For Test
___
Additional Item Attachment:
File name: trunk-S2_5-net-compress-asserts.patch Size:2 KB
Follow-up Comment #9, bug #18532 (project freeciv):
when the map is revealed [...] there's no attempt to enable
network buffering or compression for the large amount of data
that is likely to be sent. Perhaps adding some would help.
OK, there is buffering via buffer_shared_vision().
Follow-up Comment #3, bug #21272 (project freeciv):
Just had my client crash for a fourth time with this bug.
I was looking into a team mates city again.
I was able to look into that city both before the crash and again after I
logged back on. Nothing had changed to the city in the meantime.
20 matches
Mail list logo