On Friday 06 June 2003 03:01 am, you wrote:

> GUI about them.  This is necessary after loading a file, because the
> GUI reinitialises its Studio configuration when the file is loaded
> (i.e. using information found in the file, such as the patch maps
> and so on) and then needs to synchronise the resulting information
> with the actual devices that the Studio knows about.
>
> Yes, this is a rather fundamental issue actually.  If new ALSA
> clients are created while Rosegarden is running, it should pick them
> up and offer you devices corresponding to them -- but it can't tell
> the difference between this situation and the situation in which
> you've just loaded a file with fewer devices in it than existed
> previously, because both of them are handled through the same
> syncDevices mechanism.  In both cases the GUI simply sees the excess
> devices as "new".
>

Chris, I have some observations about how this works for me with what I am 
trying to do.  The studio handling not smooth yet; I'm not yet sure if it is 
still "in progress" or if I don't understand the design, or maybe the design 
just doesn't fit my mode of operation very well.  So, as typical with open 
source projects, you can take what I have to say or leave it. 8-)

Basically, in the old days when I loaded a sequence into a sequencer program 
using all hardware synths, all the outboard gear used the same MIDI port and 
channel as it did the last time, unless I rerouted cables or programmed in 
new MIDI channels on the gear, neither of which hardly ever happened.  
Generally, the sequencer had no idea what was on the other end of those 
cables. If something didn't make noise or perform it's function, it was 
usually because it wasn't turned on or something like that.  Although this 
situation made reconfiguring your (hardware) studio a lot of work, it also 
provided a very stable configuration and, generally, the sequence that you 
were working on yesterday would work just as well today because the 
configuration hadn't changed.

With rosegarden, of course, we have software synths and all sorts of 
flexibility to route MIDI or audio around "virtually".  This is a great 
feature in general, but right now it feels like the stability of the hardware 
setup, which was one (the only?) of it's strengths, is lost.  Right now, 
every time I load a sequence into rosegarden I have to go reconfigure my 
"virtual studio", because the dynamic configuration hardly ever comes up the 
same way twice.  Often, the names of the devices (like ZynAddSubFX) are 
intact, but they are associated with the wrong device.  What happens if I 
haven't yet started a synth that I had running the last time?  I would prefer 
that rosegarden just point to a nonexistant port, waiting for me to start 
that other program.  This is similar to a "oops, forgot to turn it on" error 
which, IMHO, is much simpler than "oops, I forgot to turn it on and now my 
sequencer has rerouted that channel somewhere else so I have to reroute that 
back after I start the synth".  This turns out to be more work overall than 
using a full hardware setup.

Also, because my system has a complex MIDI setup, the Rosegarden Studio is 
always maximally complex, even for a sequence that only uses one port.  I 
suppose this wouldn't be too bad, except that the scans take so long that I 
have to think twice before opening the "studio" setup box.  In the hardware 
situation, you just have MIDI ports going nowhere and the system ignores the 
ones you aren't using.  Again, it would be nice to have the flexibility of 
the "virtual studio" but the simplicity of the old days.

> And of course in many cases after loading a file you _do_ want any
> additional devices not listed in the file to continue to be
> present -- for example, the default autoload.rg loaded at startup
> may have fewer devices in it than are actually present, but you
> still want the additional ones to be visible.  Or if you borrow a
> .rg file from someone else, you don't necessarily want it to get
> rid of most of the devices in your setup that your friend doesn't
> have.  I guess this is where we need the option to merge studios
> between files.

This is a feature that I would like to disable most of the time.  I can 
definitely appreciate the coolness of sharing a sequence with somebody else 
and having it work, but I doubt this feature is used very often, and there 
are a zillion things that can break the feature even when rosegarden does 
it's part of the job correctly.  The way it works now, whenever a ALSA device 
appears or disappears, rosegarden makes a huge deal about it with a lengthy 
scan, even though the change is probably temporary and maybe doesn't even 
concern rosegarden in the first place.  I would rather use a mode where 
rosegarden never scanned for devices, and only used the ones that were 
explicitly aconnected to it.  Again, this is similar to a hardware setup 
where I can move a MIDI cable without having some component in the rack freak 
out about it.

Well, thanks for reading this far. 8-)  Please realize that I wouldn't even 
care how Rosegarden handled it's studio if I didn't think that Rosegarden 
might one day be the app that finally kills the need for Windoze sequencers, 
at least for me.  I am currently working on a very complex project, and would 
like to use Rosegarden if at all possible.  If you like this kind of 
discussion, I will continue to raise these kinds of issues.  If, on the other 
hard, you find me annoying, then I will quietly deal with what I have. 8-)

John



-------------------------------------------------------
This SF.net email is sponsored by:  Etnus, makers of TotalView, The best
thread debugger on the planet. Designed with thread debugging features
you've never dreamed of, try TotalView 6 free at www.etnus.com.
_______________________________________________
Rosegarden-devel mailing list
[EMAIL PROTECTED] - use the link below to unsubscribe
https://lists.sourceforge.net/lists/listinfo/rosegarden-devel

Reply via email to