Hello David
Well, in first place, thank you for the good explanation.
I see there are a lot of issues for this test. For clarify, I only
needed something to write into a text file the order and sequence the
liquidsoap perform the songs, only for switches and playlists. Of
course, this not involve real time transitions, inputs from harbor, etc.
Based on your excellent explanation, I will try to made an external perl
script to "simulate" how LS will perform the playlists and switches, and
write down in a text file. The perl script will accept two parameters,
like LS script path and the amount of time to simulate, ie 24hs.
Then the perl script will read for playlists and switches into the LS
script and then simulate, based on the previous calculated song
duration, the playlists and switches for 24hs, and write this into a
test file.
Thanks again David!
Regards
Normando
El 15/12/2010 06:49 a.m., David Baelde escribió:
> Hi Normando,
>
> In a sense you can already do this, and in another sense you will
> never be able to do it... Liquidsoap is really something interactive,
> this is what gives you so much liberty, but everything has a price: We
> let you write any predicate in a switch, but as a result we can't do
> anything else than running this predicate. In particular, we can't
> guess when the predicate will be true or not. There are many more
> examples like that.
>
> So, we have to run the stream to know what it does. But this is good
> enough in many situations. It is possible to run your stream really
> fast: first remove the synchronization on your clocks (in old
> liquidsoap, set("root.sync",false), and in new ones, wrap outputs in
> clock(sync=false,...)) and second, change your outputs for
> output.dummy so that you avoid encoding. Now you have a stream that
> goes as fast as your CPU can, and that you be plenty enough for
> testing. A few pitfalls: obviously, it won't work if your script
> relies on a soundcard input or output, or an external stream
> (input.harbor/http).
>
> Finally, a problem that might matter for you: several components in
> the script (including external scripts) may rely on the "real" time
> (i.e. the time given by your computer's clock) and that time won't be
> accelerated when we accelerate liquidsoap's clocks. I don't think
> there is a unique answer to that difficulty, but solutions could
> involve computing a simulated time based on the date of liquidsoap's
> startup and the current time in liquidsoap's clock (which you can get
> using get_clock_status).
>
> On Tue, Dec 14, 2010 at 11:02 PM, Normando Hall<[email protected]> wrote:
>> 10:00:00 - song1 (duration 2:01)
>> 10:02:01 - song3 (duration...)
> For getting the actual output in the form that you want, simply use a
> couple on_metadata. You can use a global "debug" variable in your
> script to enable disable easily all those mechanisms.
>
> I don't really know how to conclude: All this probably looks like too
> much pain for a simple test. In particular, it would be nice to have a
> simple solution to the problem of accelerating the "real" time, or at
> least tools to deal with it. On the other hand, I'm not sure how
> useful it'd be to work on improving testing because it'll never work
> for bigger systems, and on simple cases like the one you mention it
> has only a limited interest: hopefully, you'll quickly come to trust
> what you write using playlists and switches (transitions are another
> story, but require a different testing approach in my opinion).
>
> Cheers,
------------------------------------------------------------------------------
Lotusphere 2011
Register now for Lotusphere 2011 and learn how
to connect the dots, take your collaborative environment
to the next level, and enter the era of Social Business.
http://p.sf.net/sfu/lotusphere-d2d
_______________________________________________
Savonet-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/savonet-users
