Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

Some news:

There is ongoing effort to RE the closed Nokia pulseaudio bits. So far we have https://gitorious.org/pulseaudio-nok...998671f6c8d72:

Having those as source code will not only ease porting fremantle to other devices, but will allow other distributions to be more easily ported to N900.

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

(I guess some crossposting is in order for THIS)
Welcoming Juho onboard FPTF team. N900/N9 first hand experience as PA responsible. Our chances to sort audio just increased by factor100
what a happy day! \o/

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

Nice! Very Nice!

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

a small update on those pulseaudio bits:

both music and record module seem to work, sourcecode is on https://gitorious.org/pulseaudio-nokia , if someone wants to test, I can provide pre-build .deb files

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

Great to hear (no pun intended) :) so, which modules are still left to be REd? In /usr/lib/pulse-0.9.15/modules I see a lot of them, but I think some of these are open source... Aren't only the one with prefix module-nokia-* closed ones? If so, only the module-nokia-voice.so is left to RE? :) Next question - does it mean that upgrade of pulseaudio (or porting these changes to newer pulseaudio in e.g. Debian or Arch on N900) is now possible and makes sense (providing one wants to do it)?

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

yes only(some ~300k ARM binary) module-nokia-voice is left to be REed :P.

That is one of the reasons we started REing - to be able to use those modules with other distributions as well.

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

A small update on the PA closed bits - yet another one is ready - the famous xprot (speakers protection) module. See https://gitorious.org/pulseaudio-nok...0bcd2023df9399

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

I see the times are about 20% better in each test case. Do you see a difference/noticable improvement in CPU usage as well?

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

Not sure I understand your question, so will try to elaborate:

all the timings were performed on n900 with CPU locked @ 500MHz by using the test program. We still don't have all the pieces together, so running xprot "in action" is still not possible. Despite that, I expect to see lower CPU usage with replacement, as the above timings are on the algorithm that does the job. i.e. faster processing == lower CPU usage.

Re: the Fremantle Porting Task Force, or "how to run maemo on Neo900"
 

Been a long day in a very long week...again.

Yeah, that's what I was trying to get at. I know there have been issues in the past due to higher than desirable CPU usage during audio playback, such as in emulators.