Closed GoogleCodeExporter closed 9 years ago
Oops.
Not a defect, nor Medium priority...
Original comment by doomagent13@yahoo.com
on 27 May 2008 at 4:53
Malformed patch @ line 221. If someone could redo this to work, it would be
great.
Original comment by Paradox460
on 1 Jun 2008 at 7:57
Attached patch cleans up the code formatting of the original patch, and removed
the
patching error. Also removed g_randomMaps, as it was an unused cvar.
That is however all I did, the code itself is not run-tested.
Since no instructions were posted, it looks like it does this:
a) if a map of name '*RANDOM*' appears in the maprotation, a map is randomly
chosen
from the maps listed in the file pointed to by g_randomMapList.
b) g_randomLayoutPercent: if more than one layout exists for a map, will only
load
one of them g_randomLayoutPercent percent of the time.
FIXME: length of the file g_randomMapList is unchecked, which could be pointed
to a
huge file, which could cause an inability to allocate the memory - with unknown
results.
Original comment by Rez...@gmail.com
on 3 Jun 2008 at 1:49
Attachments:
You got it about right, Rezyn. I was planning to add random map votes, but at
the time I made this, had not
yet done it. I originally posted the patch in a codebox on the forums, along
with instructions, but I forgot to
copy the instructions when I submitted the patch...
I have tested it, but only in .dylib format. I didnt test it in qvm format
because: 1) I dont think I have the right
server executable to run this qvm without in effect backporting it, and 2)
.dylibs yield much more useful
information if they crash.
While the file length is unchecked, I was picturing it just being a number of
mapnames separated by
whitespace. Such a file would almost certainly remain below even 1 kB. It
intentionally does NOT check if a
map is listed more than once, so that random chances can be adjusted to be
something other than
1/<number_of_maps>.
Original comment by doomagent13@yahoo.com
on 4 Jun 2008 at 12:01
Here is a fresh patch straight from "svn diff". I tested, and it cleanly
applies to revision 63...
Original comment by doomagent13@yahoo.com
on 4 Jun 2008 at 2:31
Attachments:
Fixed.
Released as of 1.6
Revision 64
Original comment by Paradox460
on 7 Jun 2008 at 5:09
Original issue reported on code.google.com by
doomagent13@yahoo.com
on 27 May 2008 at 4:50Attachments: