[Ffmpeg-devel] [RFC] av_random...
Wed Jan 10 18:21:12 CET 2007
On Jan 10, 2007, at 2:53 AM, Reimar Doeffinger wrote:
> On Tue, Jan 09, 2007 at 05:03:20PM -0600, Ryan Martell wrote:
>> On Jan 9, 2007, at 4:47 PM, Michael Niedermayer wrote:
>>> On Tue, Jan 09, 2007 at 04:03:21PM -0600, Ryan Martell wrote:
>>>> For the record, the timing units I sent before were only timing the
>>>> generate_untempered function, since it's only called every
>>>> AV_RANDOM_N times av_random is called.
>>>> Final question: typically, I use time(NULL) to initialize the seeds
>>>> of a random number generator. Since ffmpeg says don't use time
>>>> (NULL), what's the best pseodo-random seed generator to use?
>>> 0 should do
>> Not really; then it won't be even random- the GUID of everyone (in
>> asf.c for my mms stuff, that started this all), will always be the
>> same for every user out there. The seed must be initialized with
>> something that's at least pseudo random.
> In how far is that a problem? Using truly random numbers is both a
> privacy concern and a problem for regression test, so IMO they
> should be
> avoided by any means necessary, if really necessary maybe adding an
> option to enable them...
Well, for me, if I had three or four versions of ffplay connecting to
the same mms server, the mss server could get confused; I don't know
though, since the mms stuff is a closed protocol. I suspect that the
GUID is used internally for "unique-ness", but not sure why it would
need that when it also has ip ports & destinations (which are
inherently going to be unique, unless there is some weird translation
going on, or multiple ip interfaces on the machine).
I thought about using the mms url and hashing that for the key, but
since several apps could be using the same url, that would go back to
the same point above.
url connection port would be good for my usage, but I don't know how
to get that out of the URLContext (doesn't appear to be available).
I guess I could just hardcode something in there, and know that every
time through it would be getting the same value, and see if the
Windows WM9 Server breaks....
More information about the ffmpeg-devel