Re: [LAU] [ANN] LADI Session Handler - Preview 2

Nedko Arnaudov wrote: Hi all, I must admit I was skeptical about Ladish, after the disappointing progress of LASH and the jackdbus debate on LAD. By accident I got an conversation with Nedko on IRC and being in an 'tweak-mode' I was so 'stupid' to try ladish, not knowing what to expect really cause the website couldn't make that clear to me at that point. But also with in mind that an Session Handler is what Linux needs if you want to work 'the modular way'. Since I met Linux it has been an debate, whether the modular approach on Linux is good or bad. It became obvious to me that technical spoken there where advantages for sure, and after I did see the JACK design presentation on jackaudio.org I found it a intelligent concept, which could especially be useful in an open source community, where it seems to be harder for people to spend a lot of time on a big 'all-in-one-app' or to work together like you can do within a company. So an approach where developer A works on an 'one-task-one-tool-app' for task Y and developer B is working on a 'one-task-one-tool-app' for task X, and being able to connect one tool with another, looked like it was the best way to gain success when working as an community. But, all though striving hard to proof wrong, the disadvantage of the modular approach was hard to tackle. Launching all apps and settings manually again and again. Especially till approximately one year ago, when I was less capable of configuring my system for RT properly and when the RT kernels where not as good as they are now, and JACK was shutting down more often then the recent JACK2, it was really bad often. Just launched four apps with their settings, made some sounds and then "jack shuts down unexpectedly", end of party, restart party. Being busy for 15minuts to get it all back again... With an session handler like Ladish, I feel the modular approach gets another chance. It could fulfill the potential of the modular approach, the-one-task-one-tool, the unix philosophy. It makes working on Linux beneficial again in certain areas compared to other OS. I'm not saying an 'all-in-one-app' with plugins is not good. No I like that to, but with the LinuxAudio infrastructure at the moment in mind, it is a 'must' to be able to work both ways. Now (there are no LV2 plugins of Hydrogen, ams, etc. soon) but also in the future. After some testing (I decided to compile jack2 from git and ladi branch with --dbus and --classic so I was able to use both jackdbus and jackd. ) I must say that I'm enthusiastic about Ladish. It does make working the modular way on Linux far more pleasurable already. It launches apps automatically with in a studio and restores the connections. As a 'tester' I found the support excellent via IRC (#ladi at freenode). I had some Debian related issues and they where fixed within an hour. Fixes where quick not only for ladish but also for the laditools and a2jmidid (I build them all from git). Also the devs are open to suggestions and critique. I really can recommend everyone with a bit of Linux experience to try it out! I think Ladish is especially useful for the 'Light' music musicians (pop, rock, dance etc.). It is not necessarily the best session handling approach for everyone though. I can imagine that projects who work with a very stable set of tools (openoctave for example) , could equally benefit from some kind of scripting solution, like the openoctave project is using already. But when I think about the discussion on LAU about 'fast composing tools' by Atte and Ken, I really think Ladish could be great for that kind of projects. It could even be a alternative to Renoise, when launching tools like Aldrin/ Epichord, Lv2rack, fst/vsthost, jack-mixer together in a 'studio' in Ladish (just an example). Why not sit around the table with some people who have the same goals and make those apps work perfectly together with each other and are doing the things you want in your workfow (the openoctave approach)? The same is true for the discussion about 'Ableton live alternative on Linux' imo. I'm not saying Ladish is already there. There has been a firm debate about the good and the bad of jackdbus on LAD (which I don't want to do over again in this thread ;) ). It doesn't seems to be totally clear what are actually the benefits or disadvantages of jackdbus, what are myths about jackdbus and what not... So I think another benefit of having more testers, users and developers involved in Ladish, is that you can influence the development of Ladish a bit and that the community can point to, and solve some problems if they are there (btw the developers are not necessarily sticking with jackdbus for ever, it seems to be the best choice atm and we have to see how it develops). Another point later in it's development is the implementation of level 1 and higher in other apps (http://ladish.org/wiki/levels). How could this be done as easy and good as possible and how can apps implement Ladish support? At least I go further with testing and enjoying the features of Ladish for my workflow. I hope more people (users and developers) will join to test, stimulate and criticize it. If Ladish succeed it will be a major breakthrough for LinuxAudio and especially the users imho.