About FreeADSP
FreeADSP, acronym for "Free Audio-oriented Digital Signal Processor" is a free, modular, audio-oriented, real-time, cross-platform DSP software system aimed at becoming a reliable replacement to pedal boards and stomp boxes for music players, as well as being suitable for any other kind of real-time sound processing.
Latest news
2007-05-18: Yet another recipe
Ingredients (what happened):
- 2 do-nothing development releases
- 1 tiny, yet significant, citation as a "fledgling" project by a notorious audio FLOSS figure
- a pair of links (1, 2) on two major websites for Linux audio applications
- 1 new developer
- 1 new web admin
- a lot of thinking and talking to other people involved in other FLOSS projects
- pretty much personal study of known programming issues
- tons of time lost at the university with some horrible marks
Directions (how it happened):
- Start writing code. In one month publish the first development release, and after only two weeks publish the second too. Put a website on the Internet and improve it to look less terrifible. Keep in mind that this is to give possibly interested people, who you are already talking to, a general vision of what you're doing.
- Realize that you can do a lot better implementing more interesting stuff like proper multithreading. Also realize that in order to do this and to rework the same code less times, you may want to get back to the design phase and write a general overview of what you're doing and some guidelines on how you think to accomplish such tasks.
- Try to find some help in either people you know or not. "Hire" the developer and the web admin and soon find out that they won't help at all. "Fire" them but keep on going on.
- Discover that some people noticed what you're doing, altough noone contacted you intentionally to talk about it.
- "If Mohammed will not go to the mountain, the mountain must come to Mohammed": gather opinions from the audio FLOSS developer community about some cool ideas you have, which almost everyone can take benefit of. Understand that too many of them think that flamewaring is a more interesting sport than constructive discussion.
- Contact the ones who seem more interested, and soon get one of (or both) the fatal answers: "Cool, we can work on it! But take a look at our TODO list - which has nothing to do with what you were talking about -, maybe you want to cooperate with us!" (and let go what you were doing) and "We have no time to work on it even we find this very interesting - euphemism for 'crucial' - but we look forward to hearing from you when you've done, and maybe we can integrate - aka 'steal' - it."
- Understand that you can count only on your own forces, and even if you would like to send to hell everything regarding FLOSS, notice that an idea is right or wrong reagardless the people behind it, so don't hesitate and go on.
- While going into deep about design decisions to take, discover with extreme disappointment that a lot of mainstream applications and libraries out there are well known to act against any kind of logic, and even the C language itself has many design flaws. On the other hand notice that some less known pieces of code behave better. Ask yourself why and find no answers not involing marketing rules and/or political decisions. Decide to take your own way and clearly show how things ought to be done starting an ambitious parallel project.
- When you've done all of this, you'll probably feel ready to rock. Be patient! You still have to face the first problems with the ridiculous Italian university system. Be warned: only the bravest and the strongest got out of this dangerous and endless jungle. You have been warned!
Dig in! (what now?): Serve this meal in a renewed (also if not yet completed) website. Then work on the parallel project before restarting coding on this one. Take into account suggestions and requests, accurately descerning truly interested people from the rest. And most important: use your little time the best possible way!
posted by Stefano D'Angelo, 2007-05-18 00:00:00. Comments (35623).