Re: [Jack-Devel] jack session questions

On 2 July 2012 09:27, hermann <[hidden]> wrote:
> Am Montag, den 02.07.2012, 09:12 +0100 schrieb James Morris:
>> On 30 June 2012 10:15, James Morris <[hidden]> wrote:
>> > Hi,
>> >
>> > Just wondering if there are any behaviour recommendations?
>> >
>> > Any recommendations to disable quit/close/open/save/etc when the
>> > client is under jack session management? (Can a client know if it's
>> > under jack session management - does that even mean anything outside
>> > of the callback?)
>>
>> Guess not then.
>>
>
> Hi James

Hi,

>
> In guitarix we check if the client is called with a given UUID, and
> store the result. Depend on, if the UUID is empty or not, we avoid self
> connections and store/read presets from the default or the jack_session
> path. So yes, the app know if it is under session control or not.
>

Ok thanks, I hadn't thought of looking at it that way.


>> > Should clients assume their executable is installed in a path in
>> > $PATH? If my client can work out it was started in a relative path
>> > should it then work out the full path and pass that in reply to jack
>> > session? This would allow a using a different version to the installed
>> > version of some software to interact with jack-session (ie for
>> > development testing of jack session).
>>
>> Some do, some don't.
>>
> Well, for local testing purpose out of my source path, I have added the
> path, otherwise, we give only the (exe)client name and assume that the
> app is installed in $PATH.
>>

amSynth uses BinReloc by Hongli Lai but I thought it looked a bit
overkill so wanted to know what is expected behaviour. I guess it's
one of those things that people would do if there weren't so many
portability problems.

Thanks.

James.

>> >
>> > Cheers,
>> > James.
>
> greets
> hermann
>

1341220854.3017_0.ltw:2,a <CAPLoHyrWU+8GzqBoVg3esnsfv4jNMan-aB72Lca4BYb4yB8JyA at mail dot gmail dot com>