On Wed, 2012-03-28 at 14:11 +0200, Emanuel Rumpf wrote: > Am 28. März 2012 05:42 schrieb David Robillard <[email protected]>: > > On Tue, 2012-03-27 at 14:24 +0200, Emanuel Rumpf wrote: > > I am having a hard time imagining anything *less* likely to be adopted > > than trying to cram a *database* down everyone's throats to save some > > files! ;) > > > With database here, actually, I'm refering to a more or less simple text > format. > for example, recently I stumbled about: GNU recutils > (readable, but it is slow) [...] > Did you ever re-assign 200 symlinks ? > Compare that with a simple search-and-replace in a textfile, with an > editor of your choice.
In addition to not being archivable by any archive tool, and not transparent to file system tools, anything but normal files means you don't have normal app state in the session anymore, but need some kind of mechanism to ask for every single file. This means loading code gets weird and depends on the session manager, and the app's session format also depends on whether or not it was saved with the session manager. Apps are not going to go for that, period. It should be obvious by now that a prerequisite for a session manager that is actually going to be adopted is it doesn't ram a bunch of annoyance down implementer's throats. Any such annoyance needs a very compelling argument to counteract that. I don't see one here. A sufficiently "compelling" argument would be something important that works that could not work without all the additional junk. -dr _______________________________________________ Linux-audio-dev mailing list [email protected] http://lists.linuxaudio.org/listinfo/linux-audio-dev
