Re: [jokosher-devel] Nearing the freeze

2007-01-26 Thread Michael Sheldon
On Fri, Jan 26, 2007 at 12:10:01AM +, Jono Bacon wrote:
 Hi all,
 
 Well, an exciting time is afoot, and Jokosher is nearing its 0.9/1.0
 freeze. There is still confusion as to whether we are 0.9 or 1.0, and
 just when we were resigned to us being 0.9 because multi-channel card
 support may not go in, it seems that the GStreamer side may be fixed
 up and we may well have the support in. Elleo is testing it tonight.
 :)
 
 So, where are we? What needs doing? How are we going to make this
 *the* Jokosher release for people to use? I really get the feeling
 that people are looking at Jokosher for actual recording - more bug
 reports, podcasts saying they are considering it (such as the Linux
 Action Show) and my test podcast proves we can do it. Now the
 dependencies are hitting distros, this is making life *much* easier.
 Exciting days. :)
 
 So, here are some things to think about, comments are encouraged:
 
 -- Multi-channel support
 
   I think if we can get this in, this is essential. Right now
 everything we need is in a released gst, but multi-channel support
 will require an updated audioconvert and the deinterleave element -
 can we use caps detection to see if the correctly patch audioconvert
 is available for correct recording - i.e. if the user has an
 audioconvert that does not have the bug fixed, can we disable
 multi-channel support?

 Checking audioconvert should be fairly simple, it's just a matter of
seeing if the user's version of audioconvert can handle less channels
then their card. Even with an old version we won't need to disable it
for users who have 8 or less simultanious inputs; although we will
still have to check for deinterleave.

 
 -- docs vs. userdocs
 
   I have been thinking a lot about the docs sites, and I think doc.j.o
 is becoming irrelevent. I suggest we ditch it, make userdocs docs.j.o
 and theme it like the main site. I will then poke the site and make
 clear separate sections so that official docs appear on the front page
 and people can link to development docs.
 
 -- Bugfixing and testing
 
 When we freeze we have a month of bug-fixing, which is essential to
 make Jokosher rock-solid. When we released 0.2, we got a lot of
 recording bug reports, much of which seemed to be down to dependencies
 and certain sound cards (such as the big SBLive bug report thread). I
 think we need to encourage as much testing and bug reporting as
 possible. Rest assured, I have a plan for this, and I am going to go
 on a test Jokosher, report bugs pimping mission. It would be amazing
 if we could nail the vast majority of our bug reports. :)
 
 -- Stuff left to add
 
 Not much left to hack in, but mainly:
 
  * Mixdown Profiles - Aq is working on this and assures me it will be
 in before the freeze. I think Aq needs to hack this, and we should
 determine how stable the code is. There is no reason why itcan't go in
 and we disable it from the UI if we are not confident it is mature
 enough. We not only need the facility, but some actual profiles too,
 and I would suggest update a podcast feed is the first one to use.
 
  * Freesound - the freesound extension needs some UI love. The code
 actually works, and is awesome for demos, but it needs refining and to
 have a non-blocking GUI.
 
 -- Effects and presets
 
 I would like to ship some preset packs with Jokosher that can go on
 the website and be included. Would be useful to have a preset pack
 installation option if we have time to include it. I hope to spend
 some time sitting down and making a bunch of presets that sound
 decent. I did for 0.2 and managed to get some really nice sounds. Of
 course, these are listed at the instrument level, which is cool. :)
 
 -- Accessability
 
 Elleo has been hacking on some pretty incredible accessibility stuff,
 and of course, our effects code is so much more accessible now. Would
 be good to get a disabled user or two to test this.

 Yes, it'd be great if we could get someone with real experience to
test things out, there are still some feature I know are completely
impossible to handle without a keyboard (such as fades), but I'm 
confident I can make this accessible given time.

 
 So, I think those are the main points. Pretty exciting days, and I am
 realy confident that if we can ensure Jokosher works for most people
 and we can avoid the recording bugs that we were getting, we should
 actually start hearing things recorded in Jokosher. :)
 
 As ever, thanks for all your efforts folks. We have an awesome
 application and an awesome development and user community forming. :)
 
   Jono
 ___
 jokosher-devel-list mailing list
 jokosher-devel-list@gnome.org
 http://mail.gnome.org/mailman/listinfo/jokosher-devel-list
 

-- 
Michael Sheldon
http://www.mikeasoft.com
___
jokosher-devel-list mailing list
jokosher-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/jokosher-devel-list


Re: [jokosher-devel] Nearing the freeze

2007-01-25 Thread Laszlo Pandy
Jono Bacon wrote:
 Hi all,
 
 Well, an exciting time is afoot, and Jokosher is nearing its 0.9/1.0
 freeze. There is still confusion as to whether we are 0.9 or 1.0, and
 just when we were resigned to us being 0.9 because multi-channel card
 support may not go in, it seems that the GStreamer side may be fixed
 up and we may well have the support in. Elleo is testing it tonight.
 :)
 

I thought this was already decided. Personally I don't think 
multi-channel recording should be the only thing that makes or breaks us 
having a 1.0 release. Things like accessibility and total keyboard 
navigation of the interface is looking really good thanks to Elleo, but 
this needs to be complete before 1.0. At this point, with 7 weeks until 
release I still see too many rough edges to call it 1.0. The name 1.0 
to me says 100% polished, and this is especially true with open source 
projects (even though they take much longer than a year to get to 1.0).

 -- Multi-channel support
 
   I think if we can get this in, this is essential. Right now
 everything we need is in a released gst, but multi-channel support
 will require an updated audioconvert and the deinterleave element -
 can we use caps detection to see if the correctly patch audioconvert
 is available for correct recording - i.e. if the user has an
 audioconvert that does not have the bug fixed, can we disable
 multi-channel support?
 

This is the gstreamer CVS thing again, eh? Since its not a killer 
feature for everyone, we should probably enable it but tell people who 
want multi-input cards to use CVS. Yet another reason to not make it 1.0.

 -- docs vs. userdocs
 
   I have been thinking a lot about the docs sites, and I think doc.j.o
 is becoming irrelevent. I suggest we ditch it, make userdocs docs.j.o
 and theme it like the main site. I will then poke the site and make
 clear separate sections so that official docs appear on the front page
 and people can link to development docs.
 

Agreed, I was never really sure why we set up two sites in the first place.

 [snip]

  * Mixdown Profiles - Aq is working on this and assures me it will be
 in before the freeze. I think Aq needs to hack this, and we should
 determine how stable the code is. There is no reason why itcan't go in
 and we disable it from the UI if we are not confident it is mature
 enough. We not only need the facility, but some actual profiles too,
 and I would suggest update a podcast feed is the first one to use.
 

Hmm, maybe we can get some partially functional code committed so we can 
start testing/hacking/writing mixdown profiles? As our freeze is 6 weeks 
long this time (and the last one was only 20 days) mixdown profiles will 
be eligible for a maximum two week exemption from the freeze (if this 
idea gets approval from two other core devs).

  [snip]
 
 -- Accessability
 
 Elleo has been hacking on some pretty incredible accessibility stuff,
 and of course, our effects code is so much more accessible now. Would
 be good to get a disabled user or two to test this.
 

It would be nice to have some accessibility experts give us some 
feedback and advice, but for the rest of us we can just close our eyes 
like Elleo does when testing it. I have a wireless keyboard (but not a 
wireless mouse) and sometimes I will sit on the other side of the room 
and see how easily I can make Jokosher work using only the keyboard. 
Everyone should attempt both of these to see how far away (and close) we 
are from full accessibility.


Laszlo
___
jokosher-devel-list mailing list
jokosher-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/jokosher-devel-list