Re: tangoGPS and good user experience

2010-04-12 Thread Gilles Filippini
Marcus Bauer a écrit , Le 11/04/2010 22:07:
 On Sun, 11 Apr 2010 21:01:23 +0200
 Gilles Filippini p...@debian.org wrote:
 BTW I've found gpsd upstream very responsive each time I've submitted
 a bug to them.
 
 Currently there are 5 different GPS API's
 
  * gpsd old (pre 2.92)
  * gpsd new
  * Maemo
  * moblin
  * fso
 
 As gpsd just dropped its own dbus interface and maemo plus moblin will
 be meego, it would be great if gpsd would implement the meego dbus
 interface, thus unifying the various APIs.
 
 Doing this, the addition of more fine grained GPS power management
 would be great: many gps chipsets support a low power mode. For example
 SIRFIII chips can switch to a 5mW mode with a fix every 10secs (good
 enough for walking) and the normal mode at 50mW with a continous fix.
 
 With a 1000mW battery this means either 20 hours or 200 hours...
 
 Last not least do all chips nowadays support AGPS - would be great to
 have that in gpsd too.

You'd better discuss this directly with gpsd upstream. They have a
public developer mailing list[1] and an IRC channel[2] where - as
already said - I've found them very responsive.

Thanks,

_g.

[1] http://lists.berlios.de/mailman/listinfo/gpsd-dev
[2] #gpsd at irc.freenode.net



signature.asc
Description: OpenPGP digital signature
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Stefan Fröbe
Hi,
as I recently struggled with the same issues I'd like to comment on some of
the suggestions
On Mon, Apr 12, 2010 at 1:19 AM, Joshua Judson Rosen
roz...@geekspace.comwrote:

 ...

Most importantly: How should I keep up with work that you're doing
 upstream between releases? Maintaining the patches that I use is vital
 to me, and it's harder to do that if I have no idea where you're going
 upstream--if I have no idea what's going to change out from under me
 with the next release.


I can only second that - I ported my geocaching enhancements from version to
version, but am becoming more and more reluctant to do so with every new
release. Any public repository would be helpful in distributing and
maintaining these features, even if they wouldn't be included immediately in
any official release.

As I said, I *do* really *like* tangoGPS--I owe you a big thanks for
 giving me a good base application on which to build!


I could not agree more - and yet I'd really appreciate it if the full
potential of community driven enhancements would also be available. In the
current state, my features collect virtual dust on my HD and will probably
never ever see a release, with chances decreasing as I loose interest in
developing them further with all the burdens of porting and adapting.

I can fully understand Marcus' focus on quality, and TangoGPS is a great
result of that approach - but still I hope that someday a reasonable public
access is possible with his help instead of forking.

Stefan
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: qualitiy is important for tangoGPS

2010-04-12 Thread Sander van Grieken
On Sunday 11 April 2010 19:42:23 Marcus Bauer wrote:
 On Sat, 10 Apr 2010 16:29:15 +0200
 Sander van Grieken san...@3v8.net wrote:
  No. Just email your patches to Marcus Bauer. Expect no reply nor use
  of patches.
  
  But, you don't have the right to complain. You have the right to
  fork, though.
 
 Well, looks like you are a bit upset that your patch was not accepted

Not at all, I don't care, I build my own tangogps. I just warning people what 
to expect. 
It's quite different from other GPLed projects. 

 but quality is important for the longterm success of tangoGPS.

Yes that's why GPL projects usually attempt to also build a _developer_ 
community.

 Criteria for software quality are (amongst others):
 
  * correctness (i.e. bug free)
  * maintainability
  * robustness
 
 Your patch about speed-up is very invasive in core parts of
 tangoGPS, was not well documented, not minimalistic and introduced
 several bugs. 

Yes all true, there were some issues still that needed attention, and I didn't 
expect you 
to merge it in. I expected some feedback on the direction though. Never got it 
(until now 
:)

 In general it falls in the category of premature
 optimisation which will cause enhancements like other map datums as
 WGS84 or other projections as Merkatoor significantly more difficult
 and error prone.

Nonsense. I mean, reloading and parsing all PNG tiles on every map drag? come 
on, you can 
do better than that. And other projections? BS, you're just stacking tiles.

That's not premature optimisation, that's called caching.

 There is plenty of documentation about how to contribute to open source
 projects and my advice is to check that first.

Very funny. Where is the mailing list, where's the bugtracker, where's the 
public tree, 
where's the *feedback*?

If you're really interested in the long-term success of TangoGPS I suggest you 
start 
building on the developer community aspects. I know it's hard to let other 
developers make 
changes to your project, but you're still the owner of your own tree, and 
decide what has 
high enough quality standards for you. For not-quite-ready patches (like mine), 
there's a 
thing called branches, which you can use to give other devs a place for their 
work. 
Another option is to use a bugtracker, where patches can be attached to bug 
descriptions. 
At least then developers don't get the impression that their hours of work fall 
into a 
deep black void.

If you don't set up this critical infrastructure, or even have the courtesy to 
give 
feedback on patches, eventually all interested developers lose interest.

grtz,
Sander

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customers

2010-04-12 Thread The Rasterman
On 09 Apr 2010 12:37:00 +0200 openm...@pulster.de (Christoph Pulster) said:

  Perhaps I should clarify that I don't mean to make fun of your
  situation. We're all in the same boat here.
 
 No problem Werner, all your optinions and comments are very welcome.
 As you already said, companiess who built a solution based on the  
 Openmoko need a reliable product AND a reliable company behind.
 The product, Freerunner, had a lot severe bugs (GPS not working, GSM  
 buzzing, #1024 suspend problems). Also after fixing this with new  
 revisions and third partys (Dr.Nikolaus), the product is not ready to  
 compete in the market. Very poor battery life to mention just one  
 serious no-go.
 
 The company Openmoko Inc. missed to give customers a reliable support  
 and long-term concept. The open idea stopped behind the doors, no info  
 about stock availability, spare parts supply etc.
 CEO Sean is an visionary, not a sales guy. Steve Mosher's part of the  
 game was not evident. (BTW, what is the status of Steve according Qi ?)
 
 
 I had a project asking for 2500 units as a first order only.
 Openmoko Inc. failed to provide this customer a infrastructure.

this is the problem with phones. the big boys are beginng to get it - but for
them 2500 units is what they do for a verification run - and then throw out.
they don't get up out of bed in the morning for a request for 2500 units. :)
they will want at least 2 or 3 zeros added to that to even give you the time of
day. openmoko never made it to be big enough to continue - and ye once you get
big enough, the kind of thing you talk about no longer make business sense (as
you are busy shopping around to telcos who will order millions of devices).
catch 22 :-S

 I can confirm sold units worldwide is 20.000 maximum.
 which is a very poor result and can not give a living for anyone.
 The game openmoko was only possible with the financial injection of  
 FIC. the funds are gone, the Wikireader is an anachronistic product and  
 will not give any cashback.
 
 Not to misunderstand, the game opensource is just starting.
 if you need to, remember Openmoko as a early hero.
 
 
 Christoph
 
 ___
 Openmoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community
 


-- 
- Codito, ergo sum - I code, therefore I am --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Timo Jyrinki
2010/4/12 Joshua Judson Rosen roz...@geekspace.com:
 So, how *should* I collaborate? What mailing-list should I use?
 When I fix an upstream bug, in what BTS should I publish the patch?
 Where can we talk on IRC about current developments in tangoGPS,
 if anywhere?

I agree. After my second translation patch (fixing a few typos and
adding a few more translations) was not included, found out only by
waiting a few months for the next TangoGPS code drop, I dropped
contributing until there is a proper version control system where one
can see if changes are committed or not, and a mailing list. It's now
been a year or two since that.

There is unofficial, non-endorsed IRC channel nowadays. I wouldn't
mind if there was an unofficial, non-endorsed version control system
and mailing list as well.

-Timo

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


how to contribute to open source projects

2010-04-12 Thread Marcus Bauer
On Sun, 11 Apr 2010 19:19:26 -0400
Joshua Judson Rosen roz...@geekspace.com wrote:

[...]

Let me summarize:

 * you are well aware that forks are potentially damaging
 * you wrote one patch that was not accepted
 * now you spread fud and threaten to damage the tangoGPS project

Well okay, you certainly have the freedom to damage the tangoGPS
project which is a major contribution to the free software world... 

But:

tangoGPS is an excellent piece of software, actively maintained and
developed and very focused on the user. Any good quality contribution
is highly welcome and this means: User experience first. Developer ego
last.

And nope, I don't have the time to hang out on IRC - that's because I
have plenty of life in the real world. And I guess that's why I develop
GPS software - you can make best use of it if you leave your desk and go
out.

All major software projects have pretty high hurdles of participation.
One of the most excellent software projects is certainly Debian - and
they are well known for a veeery lengthy process to become a Debian
Developer. If you go there, submit an invasive patch to apt and demand
to become a Debian Developer or otherwise you fork Debian - well,
people will only laugh at you. That's because Debian is so big that
your fork wont do any damage.


Probably one of the key aspects of tangoGPS is its simplicity, that's
why so many people like it. Keeping it that way takes a tremendous
amount of discipline and thought. The easier it looks, the more work
and the more thought has been spent on it.

It is always easy to add more buttons, more menus, more patches more
everything. Everybody has different ideas about what is needed and for
any feature you will find someone who wants to have it - finally you
end up with plenty of buttons everywhere. The totally overloaded
toolbars of Openoffice or Word are a good example for this.


Under the hood it all this leads to code obesity. It is like eating a
cookie here, a cake there, some fish and chips, and one day you wake up
and you have to carry 140kg body weight with you around at every step.

I found a nice picture about patches:
http://www.globalnerdy.com/2009/02/26/always-keep-salt-in-the-server-room/

And patches are like patch-cables: they potentially do something, but
the question is how much of an improvement they are and how they are
done.

There is a good quote of Antoine de Saint-Exupery: Perfection is
achieved, not when there is nothing more to add, but when there is
nothing left to take away. The success of the iPods is a great example
for this.


Good contributions and long term commitement to tangoGPS are always
welcome!! I will stay commited to the user community and the continued
success of tangoGPS.

My special thanks to all the people who have given me encouragement
with their friendly feedback and support, and to all people who have
actively contributed, especially packagers and blog-writers.

Have fun,
Marcus





___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Marcus Bauer
On Mon, 12 Apr 2010 13:47:55 +0300
Timo Jyrinki timo.jyri...@gmail.com wrote:

 I wouldn't mind if there was an unofficial, non-endorsed version
 control system and mailing list as well.

Essentially it is the constant threads about forks that have been
started and fuelled over and over again by Risto that keep me having
always a few aces in my sleeve and restricting access to people who I
have confidence that they keep the project over their ego.

As long as nobody speaks up and tells them to shut up, I'll
certainly keep having aces in my sleeve :-)

Nevertheless, this totally doesn't prevent anybody to participate in
the development. In any case up to today nobody here came to me and
said: hey, how can I help out with the project? and was refused.

It is important to show that these are a few -albeit very vocal-
people who first want to serve their ego and then the users of the
project.

The fork shouters are known as poisonous people and they can befall any
project. There is a Google talk about the subject from the SVN
developers.

Marcus


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Marcus Bauer

Hi Stefan

before it gets lost in all the other discussions: I'd still be happy to
integrate this.

Remarks:
 * it must be usable for people that have not signed for the 30$
   premium account
 * everything must work from the user interface, command line is not an
   option
 * no python dependency - you can use either libsoup or libcurl

Send me your last working tarball and I'll have a look and make
suggestions or give you a hand.

Marcus



On Mon, 12 Apr 2010 09:46:46 +0200
Stefan Fröbe frob...@googlemail.com wrote:

 Hi,
 as I recently struggled with the same issues I'd like to comment on
 some of the suggestions
 On Mon, Apr 12, 2010 at 1:19 AM, Joshua Judson Rosen
 roz...@geekspace.comwrote:
 
  ...
 
 Most importantly: How should I keep up with work that you're doing
  upstream between releases? Maintaining the patches that I use is
  vital to me, and it's harder to do that if I have no idea where
  you're going upstream--if I have no idea what's going to change out
  from under me with the next release.
 
 
 I can only second that - I ported my geocaching enhancements from
 version to version, but am becoming more and more reluctant to do so
 with every new release. Any public repository would be helpful in
 distributing and maintaining these features, even if they wouldn't be
 included immediately in any official release.
 
 As I said, I *do* really *like* tangoGPS--I owe you a big thanks for
  giving me a good base application on which to build!
 
 
 I could not agree more - and yet I'd really appreciate it if the full
 potential of community driven enhancements would also be available.
 In the current state, my features collect virtual dust on my HD and
 will probably never ever see a release, with chances decreasing as I
 loose interest in developing them further with all the burdens of
 porting and adapting.
 
 I can fully understand Marcus' focus on quality, and TangoGPS is a
 great result of that approach - but still I hope that someday a
 reasonable public access is possible with his help instead of forking.
 
 Stefan

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Stefan Fröbe
Hi Marcus,

I guess with those prerequisites I won't make any geocaching submission soon
- while all the features are usable as long as you have obtained a GPX file
with all the data, the conversion to an sqlite3 db is done on the CLI with
(one call to) bash/python scripts.

Looking at gpxview [1], for example, I do not want to re-implement all that
again - and as that program now also compiles cleanly without any Maemo
dependancies on my desktop and also has maps support, there might no longer
be a point to patching TangoGPS. GPXView just is more specific and may
therefore well be the better tool for geocaching.

But thanks for the offer, I'll check my history for any other patches I got
queued and come back to you,

Stefan

[1] https://vcs.maemo.org/viewvc/trunk/?root=gpxview


 before it gets lost in all the other discussions: I'd still be happy to
 integrate this.

 Remarks:
  * it must be usable for people that have not signed for the 30$
   premium account
  * everything must work from the user interface, command line is not an
   option
  * no python dependency - you can use either libsoup or libcurl

 Send me your last working tarball and I'll have a look and make
 suggestions or give you a hand.

 Marcus



 On Mon, 12 Apr 2010 09:46:46 +0200
 Stefan Fröbe frob...@googlemail.com wrote:

  Hi,
  as I recently struggled with the same issues I'd like to comment on
  some of the suggestions
  On Mon, Apr 12, 2010 at 1:19 AM, Joshua Judson Rosen
  roz...@geekspace.comwrote:
 
   ...
 
  Most importantly: How should I keep up with work that you're doing
   upstream between releases? Maintaining the patches that I use is
   vital to me, and it's harder to do that if I have no idea where
   you're going upstream--if I have no idea what's going to change out
   from under me with the next release.
 
 
  I can only second that - I ported my geocaching enhancements from
  version to version, but am becoming more and more reluctant to do so
  with every new release. Any public repository would be helpful in
  distributing and maintaining these features, even if they wouldn't be
  included immediately in any official release.
 
  As I said, I *do* really *like* tangoGPS--I owe you a big thanks for
   giving me a good base application on which to build!
 
 
  I could not agree more - and yet I'd really appreciate it if the full
  potential of community driven enhancements would also be available.
  In the current state, my features collect virtual dust on my HD and
  will probably never ever see a release, with chances decreasing as I
  loose interest in developing them further with all the burdens of
  porting and adapting.
 
  I can fully understand Marcus' focus on quality, and TangoGPS is a
  great result of that approach - but still I hope that someday a
  reasonable public access is possible with his help instead of forking.
 
  Stefan

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Timo Jyrinki
2010/4/12 Marcus Bauer marcus.ba...@gmail.com:
 Essentially it is the constant threads about forks that have been
 started and fuelled over and over again by Risto that keep me having
 always a few aces in my sleeve and restricting access to people who I
 have confidence that they keep the project over their ego.

I don't know what egos you are discussing (I haven't followed previous
TangoGPS discussions that closely). You can restrict any access even
if you allow read-only following of the development version source
control. It's essential for any open source development project with
more than one person interested in developing it. The fear of forks is
probably simply because currently there is no way to really have a
development project where others than one person can eventually affect
the future of the project. E-mail patches without return channel or
follow-up possibility does not really work.

Anyway, I personally wouldn't want a fork, if it can be avoided. The
developer community, if there is more than one person, just completely
needs developer tools like version control and mailing lists, which
are currently not available for this project, and have been in absence
even despite months of silence on the front (so no shutting up of
anyone needed). I was thinking about a system where latest source code
is imported and patches can be tracked, which are applied, which are
pending review, which are rejected etc.

-Timo

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customers

2010-04-12 Thread Werner Almesberger
Dr. H. Nikolaus Schaller wrote:
 In business strategy planning it is quite common to separate e.g.  
 consumer, project, service, solution business types and  
 markets because they have quite different requirements and attitudes  

Yes, I was familiar with the concept but I didn't know the term
project customer. I used terms like integrator and similar,
which cover only part of the activities. Project customer is
much more precise. I'm happy I finally found the word for them :)

 At least one visitor during our presence at the SYSTEMS 2008 fair  
 expressed he is interested in 6 units [...]

Hmm, if we assume a contribution towards RD and QA of about USD
50 per device, that would be 3 millions. A while ago, Maddog and I
did a rough estimate of how much it would cost to make a new phone,
similar in style to gta02-core, but with updated components, etc.

We came up with a cost of about 2.5 millions before production (but
including transfer, certification, etc.). This visitor may be one of
those who would have enough resources to have their own design made
(*), but may not realize it.

(*) With the proviso that modules are used for most or all the RF
parts. A design with RF components at the chip level would add
difficulties and increase development cost.

 So for successful project business, one must either provide a  
 product that is available for 5-7 years or at least a consistent  
 roadmap/upgrade plan.

I think either is very difficult to provide for an entire device
using mobile phone technology. But I think that much of the push
towards technological change can be buffered by having an Open
design - one may not be able to avoid changing the bare metal, but
interfaces and the software layers above the hardware can live as
long as anyone cares to keep them alive.

- Werner

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customers

2010-04-12 Thread Werner Almesberger
Carsten Haitzler wrote:
 day. openmoko never made it to be big enough to continue - and ye once you get
 big enough, the kind of thing you talk about no longer make business sense (as
 you are busy shopping around to telcos who will order millions of devices).
 catch 22 :-S

This is where an Open hardware design can help :-) No matter which
role you play, you always have the purchase power of the whole group
behind you.

Openmoko Inc. found many doors open that would normally be closed
for such a flyspeck of a company, because it promised manufacturers
access to the Linux market.

The Open hardware design also increases the scalability - the small
garage company that makes 100 customized phones for the local
shopping mall has access to the same design resources and can have
access to much of the production resources as the largest member of
the group.

- Werner

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Marcus Bauer
On Mon, 12 Apr 2010 16:32:59 +0200
Stefan Fröbe frob...@googlemail.com wrote:
 I guess with those prerequisites I won't make any geocaching
 submission soon
 - while all the features are usable as long as you have obtained a
 GPX file with all the data, the conversion to an sqlite3 db is done
 on the CLI with (one call to) bash/python scripts.


Well, .LOC should be possible and for reading GPX files there is
already code in tangoGPS. Running a loop over the GPX and inserting it
in a db is probably in a few dozen lines of code possible. I could have
a look into it. Up to you.

Marcus

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customers

2010-04-12 Thread The Rasterman
On Mon, 12 Apr 2010 12:19:43 -0300 Werner Almesberger wer...@openmoko.org
said:

 Carsten Haitzler wrote:
  day. openmoko never made it to be big enough to continue - and ye once you
  get big enough, the kind of thing you talk about no longer make business
  sense (as you are busy shopping around to telcos who will order millions of
  devices). catch 22 :-S
 
 This is where an Open hardware design can help :-) No matter which
 role you play, you always have the purchase power of the whole group
 behind you.
 
 Openmoko Inc. found many doors open that would normally be closed
 for such a flyspeck of a company, because it promised manufacturers
 access to the Linux market.

too late for that. the others are in on the game. and now being open enough
is all that's needed. window of opportunity for om and the likes has closed -
or at the best is very close to closed.

 The Open hardware design also increases the scalability - the small
 garage company that makes 100 customized phones for the local
 shopping mall has access to the same design resources and can have
 access to much of the production resources as the largest member of
 the group.

depends on who is buying the units to make it scale - if it's a telco, chances
are they want it far from being open - that includes the hw design. chaning
that doesn't come from a small company like om screaming. it comes from someone
big enough to change the rules and power balance so its you have to come to us
and beg for a phone - then we set the rules. apple are in such a position
right now for example. :)

-- 
- Codito, ergo sum - I code, therefore I am --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Stephen Pape
Aces in your sleeve? You mean you're keeping the development process
intentionally closed to discourage forkers? What?

There's no reason you can't have a public SVN with read only access to
everyone except your chosen few. Google code or sourceforge would work
fine.

Why does all this discussion go on in the openmoko mailing list
anyway? Why don't you have your own project forums/bug
tracking/mailing list/whatever?

On Mon, Apr 12, 2010 at 9:41 AM, Marcus Bauer marcus.ba...@gmail.com wrote:
 On Mon, 12 Apr 2010 13:47:55 +0300
 Timo Jyrinki timo.jyri...@gmail.com wrote:

 I wouldn't mind if there was an unofficial, non-endorsed version
 control system and mailing list as well.

 Essentially it is the constant threads about forks that have been
 started and fuelled over and over again by Risto that keep me having
 always a few aces in my sleeve and restricting access to people who I
 have confidence that they keep the project over their ego.

 As long as nobody speaks up and tells them to shut up, I'll
 certainly keep having aces in my sleeve :-)

 Nevertheless, this totally doesn't prevent anybody to participate in
 the development. In any case up to today nobody here came to me and
 said: hey, how can I help out with the project? and was refused.

 It is important to show that these are a few -albeit very vocal-
 people who first want to serve their ego and then the users of the
 project.

 The fork shouters are known as poisonous people and they can befall any
 project. There is a Google talk about the subject from the SVN
 developers.

 Marcus


 ___
 Openmoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Marcus Bauer
On Mon, 12 Apr 2010 12:02:05 -0400
Stephen Pape srp...@gmail.com wrote:

 Aces in your sleeve? You mean you're keeping the development process
 intentionally closed to discourage forkers? What?

I happily explain it again to you: forks are potentially damaging to any
project. Here and actually only here on openmoko there have been some
people actually massively aggressing me and threatening me on any
occasion with a fork. So yes, you got that right :)

I do not always have time to work on tangoGPS and still can do a
constant flow of new features. You are a funny man - big scandale that
not all my brain activity is monitorable :) I know that RMS would like
to enclose all software developers in a gulag with constant thought
monitoring - yeah! Any software thought must be freed at once :)

Just to mention it: the linux kernel had been developed for the first
years purely based on tarballs and patches without any public
distributed VCS and collaboration has worked without problems. Simply
because in 1992 pretty much nobody had a 24/7 permanent online internet
access.

It is not as if 80% of the code changes in a week. And it is not as if
there would be legions of willing developers. Actually any open source
project that I know is short of good developers and much more software
doesn't even get developed because people lack skill, focus, energy and
long term commitment.

Last not least I am replying here because the thread was opened here
and because many people have bought Freerunners to run tangoGPS on them.

A last word: the vast majority of the members of the Openmoko community
are here for the opportunities that open hardware and open software
offer. Hwoever a small fraction has a misunderstanding about free
software. I have seriously received bizarre emails of people telling me
what I have to do because I am a free software developer and as they
don't know how to write software I have to do it for them.

Best regards,
Marcus - creator of successful free software :-)






___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS is a very successful user orienteted map and gps viewer

2010-04-12 Thread Stefan Monnier
Reminds me of the following feature request:
when downloading levels N...N+4 (or 5, 6, younameit), please also
download all the levels above (it's a negligible fraction of the
corresponding disk space and network bandwidth anyway).
I.e. instead of downloads N to N+4, it should be
download levels 1 to N+4.


Stefan


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


[shr-u] key bindings

2010-04-12 Thread jeremy jozwik
can someone running shr-u greater then 20100306 please print there key
binding action params from illum2?
my power key has been altered and i would like to get the default
action parms back.

thank you to anyone who can do this for me.

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Stephen Pape
 I do not always have time to work on tangoGPS and still can do a
 constant flow of new features. You are a funny man - big scandale that
 not all my brain activity is monitorable :) I know that RMS would like
 to enclose all software developers in a gulag with constant thought
 monitoring - yeah! Any software thought must be freed at once :)

No one is asking you to dump every thought you have for public
viewing. I don't think a publicly viewable SVN server, so people have
some sense of what's going on, takes it to your extreme scenario. They
would, however, like to know if their patch was accepted, if there was
a problem with it, or if someone's working on a similar feature at the
same time.

 Just to mention it: the linux kernel had been developed for the first
 years purely based on tarballs and patches without any public
 distributed VCS and collaboration has worked without problems. Simply
 because in 1992 pretty much nobody had a 24/7 permanent online internet
 access.

Sure, that was fine way back then. Times have changed, and people do
have 24/7 permanent online internet access. You'll notice the kernel
developers shifted away from that development approach.

 A last word: the vast majority of the members of the Openmoko community
 are here for the opportunities that open hardware and open software
 offer. Hwoever a small fraction has a misunderstanding about free
 software. I have seriously received bizarre emails of people telling me
 what I have to do because I am a free software developer and as they
 don't know how to write software I have to do it for them.

This has nothing to do with people telling you what to do because they
can't write software.

I'm not saying you have to give up control and let people hijack the
direction you want to go in. I'm saying when someone takes the time
and energy required to write a patch, they should know if it's going
to be accepted or rejected without having to wait for the next
release. They should get a chance to clean it up if there's a problem.
There should be discussion so someone can ask if they should even
bother doing the work and get feedback.

Your current approach discourages anyone from contributing back to
you, and yet you discourage anyone from creating a fork. Have you seen
the people trying to help? Someone mentioned feeling like they're
trying to hit a moving target, while blindfolded, just to help out.

Forking is not a bad thing. In fact, sometimes forks end up being more
successful than the original project. Xorg from XFree86 ? How about
gcc? Apache, maybe?
I'd suggest you read this if you have a fear of forking
http://linuxmafia.com/faq/Licensing_and_Law/forking.html (WHY LINUX
WON'T FORK: And why being able to fork is still A Good Thing)

I tried to find something explaining how forking hurts open source
software, but I couldn't find any useful results with Google. If
someone feels they can do a better job, I'd encourage them to try
forking.

-Stephen

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customers

2010-04-12 Thread Werner Almesberger
Carsten Haitzler wrote:
 too late for that. the others are in on the game. and now being open enough
 is all that's needed. window of opportunity for om and the likes has closed -
 or at the best is very close to closed.

I think the advantage is still there, it's just harder to communicate.
Also in this regard, Open Design Hardware helps: you still don't get
anything like this from the now open players.

And from what I've heard and keep on hearing, there are lots of project
customers who want to modify the hardware. They often also have the
engineering resources to perform their changes. But also doing the rest
of the phone would be too much for them.

The Open phones would be sort of a reference designs created by the
Open hardware development process, but not the one and only results.

 depends on who is buying the units to make it scale - if it's a telco, chances
 are they want it far from being open - that includes the hw design. chaning
 that doesn't come from a small company like om screaming.

A small telco may be happy enough to finally be able to brand their
products, too. I wouldn't try to deal with large telcos for now.
Don't sleep with a girl who eats more than your own weight for
breakfast :-)

- Werner

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Marcus Bauer

Hi Stephen,

thanks for your feedback - especially as you are a programmer as you
once stated on the openmoko mailing list [1].


On Mon, 12 Apr 2010 14:36:38 -0400
Stephen Pape srp...@gmail.com wrote:

 but I couldn't find any useful results with Google. 

Dang! I have exactly the same problem when trying to find 'useful
results with Google' when searching for your contributions to Openmoko:

stephen pape openmoko - 18 results (2 from this discussion)
http://www.google.com/search?hl=enq=%22stephen+pape%22+openmokostart=10


Well, have a nice evening and dream well :-)
Marcus



[1] http://lists.openmoko.org/pipermail/community/2008-April/014875.html

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customer

2010-04-12 Thread Jon 'maddog' Hall
Hello Christoph,

Good luck. Maddog made a lot words about the Brasilian universitary  
which should continue the Openmoko project. Nothing happend.

As far as I know the University still stands ready and willing to help
with the GTA02-core project as soon as that is ready to move forward.
Professor Zuffo has not de-committed from that project as far as I can
see.

As to going further than that, the University is still interested in
working on an Open Phone.  But as I have seen here over the past
several days, there does not seem to be much agreement as to how to move
that forward.

From my viewpoint a bit of this comes from a tacit disagreement in the
project as to what is open, and even less of a plan as to how to
finance a project that requires real money.  Openmoko's financing always
seemed to be on a shoestring, and never (for example) included the money
to fix problems.   Other companies might do a recall and fix the issue
at the factory. I think it was by good luck that the different issues
that happened with the phone were able to be fixed with a capacitor
here and a resistor there..and people stepped up to the bug fix
parties...but there are still a lot of people out there with unfixed
phones.

Also he cooperate with silly companys like Koolu, who bargain Openmoko
down to blood and damaged all the project.

Koolu had its faults, and I will not say it didn't, but after several
days of you writing and lambasting everyone about everything (other than
yourself, of course) I think blaming Koolu for damaging all the
project is a bit harsh.

I had a company in Brazil that was all set to license the designs from
Openmoko and manufacture the phone in large quantities.  They had a good
SMT line, channels to distribute the phone in Brazil, and from Brazil
throughout Latin America, and we had a good business plan to market to
the VARs that were mentioned in another email.  Even though the phone's
components were a little dated, we felt we had a good market in people
who had to change the OS to create the applications they wished to have
for small and medium business.  An example of that can be seen here:

http://www.youtube.com/watch?v=9dVch2nSuBA

The licensing of the design would have generated money to create the
next design.

Then we tried to find out where to buy the parts, and how many parts
were still being manufactured, and for how long.  That was when we began
to realize that the marketplace for Openmoko parts was very limited.  As
several people on this list have mentioned, to create a market for cell
phones that is profitable takes hundreds of thousands, or even millions,
not tens of thousands.

By the time that the company in Brazil ramped up to produce the phone,
did the manufacturing and certifications and testing that were
necessary, and did the certifications, built the channel, did the
advertising, they would probably run out of parts.  It would have been
unprofitable for them.

In the end I recommended that the company not try to produce the
Openmoko V7, even though I had spent a lot of time and money helping
them evaluate the possibilities.

So from my viewpoint, if there was one thing that killed the Openmoko
project, it was lack of a thorough, over-all, realistic business plan
that showed how the project was going to be sustainable into the future.

And the lack of agreement among all of the people involved as to what
the marketplace was for the phone.

md



___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: how to contribute to open source projects

2010-04-12 Thread Rui Miguel Silva Seabra
Hi Marcus,

Reading your comments, I'd say you're quickly running several steps of
How to destroy your community - http://lwn.net/Articles/370157/

I won't go into the details of listing which is which, and please note
that my felling on that is *PURELY* based upon your emails and not on
what others claim or feel or actually passed through.

Rui

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Stephen Pape
Okay. I'm not sure what you're getting at, or why you're completely
changing the subject to make a personal attack on me over an argument
involving software development practices.
I haven't been very active with openmoko, so I guess that makes
anything I say invalid? And you're the one complaining about
developers with egos?

You still haven't been able to produce any useful argument about why
forking is bad for open source, and how keeping the development
process under your own strict control helps open source.

 Best regards,
 Marcus - creator of successful free software :-)

Maybe you should look inwards regarding your complains about developer
ego. I suspect this is why you don't want anyone else forking your
code.

Good luck.

-Stephen

On Mon, Apr 12, 2010 at 4:00 PM, Marcus Bauer marcus.ba...@gmail.com wrote:

 Hi Stephen,

 thanks for your feedback - especially as you are a programmer as you
 once stated on the openmoko mailing list [1].


 On Mon, 12 Apr 2010 14:36:38 -0400
 Stephen Pape srp...@gmail.com wrote:

 but I couldn't find any useful results with Google.

 Dang! I have exactly the same problem when trying to find 'useful
 results with Google' when searching for your contributions to Openmoko:

 stephen pape openmoko - 18 results (2 from this discussion)
 http://www.google.com/search?hl=enq=%22stephen+pape%22+openmokostart=10


 Well, have a nice evening and dream well :-)
 Marcus



 [1] http://lists.openmoko.org/pipermail/community/2008-April/014875.html

 ___
 Openmoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customer

2010-04-12 Thread Werner Almesberger
Christoph Pulster wrote:
 Good luck. Maddog made a lot words about the Brasilian universitary  
 which should continue the Openmoko project. Nothing happend.

I think it's Sao Paulo you're talking about. USP never promised to
continue the project (even though the press may have mis-interpreted
this) but to give gta02-core free use of their SMT line, which is
more than generous by any standard.

It's not USP's fault that nothing happened so far. We still need to
obtain the components, make the layout, produce the PCB, and only
then we can use the SMT facility at the university.

- Werner

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: glamo backlight

2010-04-12 Thread mobi phil
 The correct full sysfs location for JBT should be:
 /sys/bus/spi/devices/spi2.0/lcd/jbt6k74-lcd/ .  Getting the not
 initialised message indicates that the correct code is compiled into
 the kernel at least (the message is generated by the JBT driver, and
 the Glamo driver won't compile without the JBT driver after my dirty
 hack). If the device really doesn't exist, it'd be good to test with the
 main 2.6.32 (non-KMS/DRM) branch (without trying to enable DRM in the
 config, of course).  The 2.6.32-gdrm branch will not work with a non-KMS
 configuration (i.e. normal glamo-fb compiled in) unless by sheer luck.
 Main Openmoko 2.6.32 without DRM also still doesn't work, as far as I
 know, but I could be slightly out of date there.

 I've attached the config I currently use when compiling 2.6.32-gdrm.

 Are you sure that the backlight is not coming on, rather than it being
 on but the display being all black?  Neither the Glamo nor JBT driver
 are actually responsible for the backlight coming on - the backlight
 itself is just an LED-powering output from the PCF chip.  Is there
 anything under /sys/class/backlight?

Thanks for your answer... Unfortunately I do not receive the emails.
Found the answer on the mailinlist homepage...
please cc my address directly, if any.

your config file helped !!!

CONFIG_BACKLIGHT_PWM=y

was missing!

furthermore you have:
CONFIG_LEDS_TRIGGER_BACKLIGHT=y

I had:
CONFIG_LEDS_TRIGGER_BACKLIGHT=y



so I think it would not be bad to add as many config files as many
relevant changes/patchces happen... and maybe they should be named
arch/arm/configs/gta02.patch.blahblah


so drm works (your example gdrm-waitq etc. etc)


I can have fun again with the drm .. :)

rgrds,
mobi phil

being mobile, but including technology
http://mobiphil.com

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: how to contribute to open source projects

2010-04-12 Thread Marcus Bauer

Hello Rui,

you certainly know that I value your opinion a lot. 

Just the other day there was a storm going on about a remark of Scott
James Remnant from Ubuntu who is an outstanding developer, very much
committed to advancing Ubuntu. People who have never done any
contribution flooded the bugzilla and flamed him.

Foul behaviour needs to be sanctioned, otherwise it ruins the whole
basket of apples. User experience is key. 

Actually I have received several supportive emails today from people
that have similar amounts (or more) of Google hits in connection with
Openmoko than I do have. As wrong as it is to go for weeks to argue
with these people it is to let it go through without response.

There were at least three other projects in the last four years with a
similar scope than tangoGPS. Where were the people complaining about
not being able to participate in the development of tangoGPS when those
projects were open for participation, where are their contributions? 

Foul play needs to be clearly identified and pointed out. I rather
doubt that those people have got plenty of invitations today to join
other projects but I'd be happy if that would be the case.

I don't know what your current main project is, but feel free to invite
them. Let me know in 4 years if they are still actively contributing
and I'll cover myself in ashes :-)

Have fun with them and please take them away from me and thousands
of happy tangoGPS users and integrate them in your projets :p

Marcus



On Mon, 12 Apr 2010 21:17:30 +0100
Rui Miguel Silva Seabra r...@1407.org wrote:

 Hi Marcus,
 
 Reading your comments, I'd say you're quickly running several steps of
 How to destroy your community - http://lwn.net/Articles/370157/
 
 I won't go into the details of listing which is which, and please note
 that my felling on that is *PURELY* based upon your emails and not on
 what others claim or feel or actually passed through.
 
 Rui
 
 ___
 Openmoko community mailing list
 community@lists.openmoko.org
 http://lists.openmoko.org/mailman/listinfo/community

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Marcus Bauer
On Mon, 12 Apr 2010 16:19:25 -0400
Stephen Pape srp...@gmail.com wrote:

 Okay. I'm not sure what you're getting at, or why you're completely
 changing the subject to make a personal attack on me over an argument
 involving software development practices.

So when you use Google it is okay and when I use Google it is a
personal attack? Hey, you are such a cutie :-) You should try to sue
Google for diffamation - you could get rich quick! :D And I will join
you because Linus Torvalds and Openmoko has a lot more Google hits
than me and that guy has never done any work on Openmoko at all.

I'll right tomorrow write him an email about what's wrong with his
completely fair scheduler. I think there are situations where it is not
fair enough. And I'll be really upset if he disregards my email just
because I have never contributed to the kernel (almost at least).

 Maybe you should look inwards regarding your complains about developer
 ego. 

I have scheduled 4 hours for this task next weekend. You think that's
enough?

 Good luck.

Thank you. And thank you for your example that the best way for
avoiding questions about forking is not to make any open source
software at all. 

...gosh, what a cutie Stephen is... :D

HAVE FUN. Last mail, day over. Thanks guys.




___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: tangoGPS community development, patches (was: tangoGPS magnify patch)

2010-04-12 Thread Stephen Pape
Now it's just getting silly.

 So when you use Google it is okay and when I use Google it is a
 personal attack? Hey, you are such a cutie :-)

You're just twisting words and sidestepping arguments. You used Google
to specifically try and prove that I don't know what I'm talking
about. You're turning a discussion on development methodology into
name calling. (Cutie? Really?)

I used Google looking for an argument against forking, trying to
figure out what you're talking about. You keep saying it hurts open
software. As far as I can tell, the open software community seems to
think the ability to fork at any time is a major benefit. You can get
around projects that have limitations or refuse to move forward. If
your fork fails, it's no big deal, but there's always a chance that
it'll become the next great thing.

Of course you don't want to discuss it, it's easier to be petty and
make it personal.

 You should try to sue Google for diffamation - you could get rich quick! :D
 And I will join you because Linus Torvalds and Openmoko has a lot more
 Google hits than me and that guy has never done any work on Openmoko at all.

Who said anything about suing anyone? When did I ever indicate that I
care who has more Google hits? I couldn't care less.
This isn't a competition for fame, stop making it out to be one.

 I'll right tomorrow write him an email about what's wrong with his
 completely fair scheduler. I think there are situations where it is not
 fair enough. And I'll be really upset if he disregards my email just
 because I have never contributed to the kernel (almost at least).

I'm not challenging your code itself, or telling you that you don't
know what you're doing. It's not comparable to your example.

 Thank you. And thank you for your example that the best way for
 avoiding questions about forking is not to make any open source
 software at all.

When did I make an example that you should make closed source software
to avoid questions about forking?
Obviously you're going to continue being unreasonable, and now you're
down to name calling and blatantly making things up, so I won't bother
with this line of discussion anymore.

I feel bad for all of the developers that want to help. I wouldn't
want to have to work with you.

-Stephen

___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customer

2010-04-12 Thread The Rasterman
On Mon, 12 Apr 2010 17:39:28 -0300 Werner Almesberger wer...@openmoko.org
said:

 Christoph Pulster wrote:
  Good luck. Maddog made a lot words about the Brasilian universitary  
  which should continue the Openmoko project. Nothing happend.
 
 I think it's Sao Paulo you're talking about. USP never promised to
 continue the project (even though the press may have mis-interpreted
 this) but to give gta02-core free use of their SMT line, which is
 more than generous by any standard.
 
 It's not USP's fault that nothing happened so far. We still need to
 obtain the components, make the layout, produce the PCB, and only
 then we can use the SMT facility at the university.

ah. components - where so much fun happens :)

-- 
- Codito, ergo sum - I code, therefore I am --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customers

2010-04-12 Thread The Rasterman
On Mon, 12 Apr 2010 16:52:03 -0300 Werner Almesberger wer...@openmoko.org
said:

 Carsten Haitzler wrote:
  too late for that. the others are in on the game. and now being open
  enough is all that's needed. window of opportunity for om and the likes
  has closed - or at the best is very close to closed.
 
 I think the advantage is still there, it's just harder to communicate.
 Also in this regard, Open Design Hardware helps: you still don't get
 anything like this from the now open players.

if it's hard to communicate - you don't have a sales point. if someone is to
spend money on something they need to be able to be told a simple thing and
get it and go aha! that's just what we want!. until the market is actively
seeking the kind of freedom you want to provide (schematics, cad design, 100%
open source os in all ways), you are the guy in the street with a sign i have
anchovie flavored chocolate. you really want some. the problem here is the
market is happy with good enough. at least the market that buys millions of
units. :) (that's why i mean by market btw - ie the mass market. of course
niches will exist!) :)

 And from what I've heard and keep on hearing, there are lots of project
 customers who want to modify the hardware. They often also have the
 engineering resources to perform their changes. But also doing the rest
 of the phone would be too much for them.
 
 The Open phones would be sort of a reference designs created by the
 Open hardware development process, but not the one and only results.

sure - but it seems those project customers want to feed off a stable supply
line - and for that you need a mass market to consume it to have that
production and thus supply line run to keep costs down, ensure basic quality of
build, design, components etc. (thus why i focus on mass market).

  depends on who is buying the units to make it scale - if it's a telco,
  chances are they want it far from being open - that includes the hw design.
  chaning that doesn't come from a small company like om screaming.
 
 A small telco may be happy enough to finally be able to brand their
 products, too. I wouldn't try to deal with large telcos for now.
 Don't sleep with a girl who eats more than your own weight for
 breakfast :-)

hahahahahahahahahhaha! :) maybe - the the small telcos are competing with
bigger ones. the big ones get to attract customers with oooh we have an
iphone! or check out this droid. branding is a nice to have... *IF* you can
match the competition. you need to get there first before small telco might
consider it. remember telco is trying to sell these phones to average joes -
and those average joes see shiny sexy iphone, then see a freerunner... guess
which one (and which telco) they choose? :) i know that to you, or to many
freedom advocates all this fancy eyecandy, sexy design, high end components
etc. seems all irrelevant to the goal of freedom - and if anything makes it
harder, and you have a point - but that point imho vanishes with the market
realities - to produce enough units to keep cost down, keep production flowing
etc. you need mass market appeal. and that means matching, or beating, the
competition in h i like that for the average joe. that means sexy
swishy animations, beautiful graphics, good screen, responsive touch surface
(capacitive), nice case/design, powerful cpu/gpu to power all the sexiness, 3g,
and then apps and lots of them and so on... you need to at least provide what
people now EXPECT from a phone. yes... even make and receive calls from
reliably from day 1 the phone ships. :)

it's a tough spot. what i see as more viable is making those that already
produce phones, more open, and gradually prying things open. getting schematics
is likely to simply never happen - you are talking different cultures even
within such companies. the hardware sides just don't even want to hear the
arguments. the software sides either get it already (and fight internally
politically, or have tough tradeoffs to make - like making it more open will
make your big customers go away as they can't close it down as easily), or are
beginning to get it. life would be easy if they all already got it and did it.
but... that's not the case. the closest to an open production-level phone today
is the n900 - and it has been a pretty rocky start there.

-- 
- Codito, ergo sum - I code, therefore I am --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customer

2010-04-12 Thread The Rasterman
On Mon, 12 Apr 2010 16:08:08 -0400 Jon 'maddog' Hall mad...@li.org said:

yo maddog!!

btw - before i launch in... i'm biased towards maddog - he's always been a well
intentioned chap at a minimum, if not and experienced and seasoned veteran who
really cares about open - in my experience that has been open source. i'd trust
maddog with a LOT (ok - not my code... sorry maddog - i'm not letting you near
my pixel processing stuff! :)), but... when it comes to contacts, strategy,
industry, ideas, cheering openness on - and more, i'd have maddog in the mix.

so...

 Hello Christoph,
 
 Good luck. Maddog made a lot words about the Brasilian universitary  
 which should continue the Openmoko project. Nothing happend.
 
 As far as I know the University still stands ready and willing to help
 with the GTA02-core project as soon as that is ready to move forward.
 Professor Zuffo has not de-committed from that project as far as I can
 see.
 
 As to going further than that, the University is still interested in
 working on an Open Phone.  But as I have seen here over the past
 several days, there does not seem to be much agreement as to how to move
 that forward.
 
 From my viewpoint a bit of this comes from a tacit disagreement in the
 project as to what is open, and even less of a plan as to how to

indeed. thats why i bring up open enough. to me - schematics are irrelevant.
datasheets of course are very useful. to others they dont consider a phone open
without schematics and cad designs and so on. everyone is different. i'm even
willing to concede that u likely will never have open 3d unless you design your
own 3d unit. not at the embedded level. all the players are closed with no
signs of going open - unless you (maddog) can convince them? :) your only other
choices are to create your own software 3d via dsp's and other auxiliary
processors. even those are in short supply of being open.

 finance a project that requires real money.  Openmoko's financing always

that's the biggest issue. financing.

 seemed to be on a shoestring, and never (for example) included the money
 to fix problems.   Other companies might do a recall and fix the issue
 at the factory. I think it was by good luck that the different issues
 that happened with the phone were able to be fixed with a capacitor
 here and a resistor there..and people stepped up to the bug fix
 parties...but there are still a lot of people out there with unfixed
 phones.

indeed.

 Also he cooperate with silly companys like Koolu, who bargain Openmoko
 down to blood and damaged all the project.
 
 Koolu had its faults, and I will not say it didn't, but after several
 days of you writing and lambasting everyone about everything (other than
 yourself, of course) I think blaming Koolu for damaging all the
 project is a bit harsh.

i'd call koolu misguided. to me they were just uninteresting. why a freerunner
when i can get a android g1 dev phone that was signficantly better hardware
(though by todays standards its totally shot and useless).

 I had a company in Brazil that was all set to license the designs from
 Openmoko and manufacture the phone in large quantities.  They had a good
 SMT line, channels to distribute the phone in Brazil, and from Brazil
 throughout Latin America, and we had a good business plan to market to
 the VARs that were mentioned in another email.  Even though the phone's
 components were a little dated, we felt we had a good market in people
 who had to change the OS to create the applications they wished to have
 for small and medium business.  An example of that can be seen here:
 
 http://www.youtube.com/watch?v=9dVch2nSuBA
 
 The licensing of the design would have generated money to create the
 next design.
 
 Then we tried to find out where to buy the parts, and how many parts
 were still being manufactured, and for how long.  That was when we began
 to realize that the marketplace for Openmoko parts was very limited.  As
 several people on this list have mentioned, to create a market for cell
 phones that is profitable takes hundreds of thousands, or even millions,
 not tens of thousands.

this is the problem. components are not a problem - if you go and buy 10
million of each, the suppliers will be happy to talk to you and provide you
with those. for this you need to design very high-end to make those components
have longevity. and then you need the money to buy them all well in advance (ie
commit to at least large initial stock and regular shipments with payments on
time or in advance - if they don't know who you are - they won't trust you and
want money up front). so this means serious dough. and serious volume to make
the suppliers sit up and take notice .. and this all comes back to having mass
market appeal. :)

 By the time that the company in Brazil ramped up to produce the phone,
 did the manufacturing and certifications and testing that were
 necessary, and did the certifications, built the channel, did the
 advertising, they 

drm/glamo Re: glamo backlight

2010-04-12 Thread mobi phil
fyi:


after successful booting and running  the gdrm-waitq example I got on the
log:
glamo-drm: Fence seq#157 was not signalled
(with increasing #ref numbers)
and usb network seems to be frozen

after second reboot and running intensive directfb test applications problem
cannot be reproduced...


directfb and other frambuffer applications run on /dev/fb0, however the old
bug with display shift to the right is again present. (everything is shifted
~150 pixels to the left)

the same with qtmooko... if you could provide the patch straightforward,
would be nice, if not, will try to see your old patch...


rgrds,
mobi phil

being mobile, but including technology
http://mobiphil.com
___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community


Re: project customer

2010-04-12 Thread Jon 'maddog' Hall
Rasterman!

We do go back a long way, don't we?  And don't worry...even though I
know a bit about the X Window System, I bow to you for the real
bit-jamming.

all the players are closed with no signs of going open - unless you
(maddog) can convince them?

It won't be me who convinces themit will have to be their
customers who buy in large quantitiesand customers who say I won't
buy your stinking 3D unless you tell me how to program it so I can
maintain it into the future.

that's the biggest issue. financing.

And here is where I saw a lot of disagreement, and could never see the
path forward to a sustainable design business plan.

Everyone wanted everything in Openmoko to be completely gratis (other
than being willing to buy the phone itself).

From my viewpoint the circuit diagrams should be open and free so
people can comment, improve, etc.  I also liked the fact that the case's
cad design was open so you could change the case.

However, printed circuit board layouts, gerbers, etc. are grunt work and
could be licensed out with a decent license that would allow
universities to make phones for free, hobbyists could make a phone or
two (or even ten) for free, but companies that wanted to manufacture or
sell it would have to pay 1-2 dollars a unit license fee.  Then by the
time a mega-unit of phones were made (and it could be small factories
each making 100K phones) you would have the money to design and test the
next phone.

i'd call koolu misguided. to me they were just uninteresting. why a
freerunner when i can get a android g1 dev phone that was signficantly
better hardware (though by todays standards its totally shot and
useless).

A hobbyist/developer might have been interested in getting an Android
G1 Dev phone.  Unlocked and Unsigned.  But each person could only buy
one.

Imagine developing a kick-ass SMB application that could not just be
delivered as an app on top of Android.  You have to change the OS.

Do you tell each of your SMB customers that they have to sign up to be
an Android developer just so they can get one of those phones?

Now the Nexus One...different story.  It sells unlocked and unsigned.

The factory in Brazil was all set to completely buy out Openmoko's
inventory (if Openmoko could have told them how much inventory they
had), but then they started looking beyond that andno one could tell
them how many GSM modules were out there, and how much they would cost
as the quantities available dropped close to zero.

the suppliers will be happy to talk to you and provide you
with those.

Sometimes.  Other times they have simply gone End Of Life with that
part and they don't want to tie up their engineers and lines with old,
obsolete partsbecause they are selling too much of the new stuff and
they are short on capacity to make both.

But if you are making your phone out of beginning of life components
that other people are also using and that have a bit of life to them,
you can sometimes get some components without having to buy 10 million
of themparticularly if you are a university...and particularly if
you have a business plan to license out the design to lots of small
companies for manufacture.

For example, it can cost one-half million dollars just to get TI to talk
to you as part of their partner program.  That half-million buys you
some TI's engineering consulting time, etc. but what it really does is
get rid of the kids and lets the big boys play.

From what I saw, the FreeRunner was EOL, with EOL components.and its
software was still a bit undercooked in places more than a year after
its design was done.

Everything else you said I agree with, and we both agree that it takes
lots of up-front moneyor a smaller amount of money and a track
record of success.

By the way, I think that both the hardware team and the software teams
did a great job given the circumstances, and I have the greatest respect
for most of the community members.

So Rastermannext time we meet we can have a beer.

md


___
Openmoko community mailing list
community@lists.openmoko.org
http://lists.openmoko.org/mailman/listinfo/community