On 07/29/2014 05:55 PM, Tres Finocchiaro wrote:
This isn't the biggest hurdle here IMO.  The core engine is the problematic part.

I'm going to dive a bit more into the kickstarter idea...  

When you ask the public to fund something large, you have to make promises that meet public demand.

Despite what your opinion is about the importance of developing the "core" (infrastructure, etc), I can assure you more people will be excited when undo is working (and I mean properly, not what we have now) and more people will be interested when they can export MP3s than they will be about a smoother running core.

I think you misunderstand my intent a bit.

It's not so much about "what is important" - it's about what we can do ourselves, what we can't do ourselves, and - if we do manage to fund and hire a developer - what that paid developer's work hours are best spent towards.

UI features, functionality, etc. are of course important - but those are not things that we'd need to hire a developer for. If we pay for a developer, if we pay for X hours/days of development time, then the best way to utilize that paid time is to spend it towards things which we're not capable of implementing ourselves.

We also need to target our userbase.  If our users want to make Skrillex "growls", we make that possible.  The "growls" and "wobbles" get buy in from people who stole Ableton and FL Studio and are sick of breaking the law.  Our campaign video represents an aggregation of the requests that are both popular, possible and make sense in our budget and timeline.

It's already plenty possible to make "Skrillex growls". I don't think that's anything we need to specifically support or encourage. Plus... people who pirate FL studio and just want to make "growls" aren't very likely to be loyal supporters of LMMS and I kind of doubt that chasing fads and pandering to the "skrillex crowd" would bring in that much support. In other words, I kind of doubt if those who are comfortable using pirated FL studio are likely to contribute money to improve an open source DAW...


The improved core would however be necessary in making LMMS into a true professional quality DAW. The limitations of the current engine will catch up with us eventually, and at some point we will very likely hit a wall where we can no longer keep patching up the holes... with the improved core, we could offer reliable live performance, which would bring a whole new market to the reach of LMMS.

And even if you don't perform live, being able to play your songs real time is pretty important aspect of the workflow. Even now, I'm constantly facing problems with most of my projects where my projects grow big enough that I can no longer play them back properly in LMMS, I have to export to be able to listen through. That's not very good workflow and it will only get worse the more functionality we pile up on LMMS.

So it's not just about infrastructure... it's that the core rendering engine, the very heart of LMMS, is dysfunctional and needs fixing. And it's not just a "nice to have" thing... When I said the stakes were to "keep LMMS alive", I was only half joking. I very much fear that unless we manage to somehow solve this issue, LMMS faces eventual extinction as the codebase becomes unmanageable and eventually bitrots when no one is left who understands how all of it works...