http://wiki.openmoko.org/index.php?title=Talk:SHR_Party_Page&feed=atom&action=historyTalk:SHR Party Page - Revision history2024-03-29T09:48:29ZRevision history for this page on the wikiMediaWiki 1.19.24http://wiki.openmoko.org/index.php?title=Talk:SHR_Party_Page&diff=80275&oldid=prevRakshat: rough answers to SHR questions.2010-03-09T19:56:43Z<p>rough answers to SHR questions.</p>
<p><b>New page</b></p><div>1. What is SHR? <br />
<br />
2. Can you tell a bit about the history of SHR? Why was the development of SHR started? What were the aims and objectives?<br />
<br />
3. What are the target platforms for SHR? Does SHR target only Openmoko phones or is it designed as a generic mobile communication/ PDA solution?<br />
<br />
4. How does the development process of SHR work ( As in who/ who all decide what is to be doe at what priority)?<br />
<br />
5. Who are the main developers working on SHR? What are their areas of interest wrt SHR and what are they working on these days?<br />
<br />
6. How far has the development of SHR progressed? Is the current SHR image usable out of the box?<br />
<br />
7. Can you explain the structure of SHR (as in kernel, middleawre windows manager, data storage structure, application layer etc)?<br />
<br />
8. If a S/W developer wishes to contribute to SHR what should she/ he do? Does she/ he need to have any specific skill sets to contribute to SHR? <br />
<br />
9. How does a developer make sure her/ his application runs on SHR. Is there any centralised verification process or a repository where verified applications are stored?<br />
<br />
10. Can you list the main areas where you would appreciate help in the SHR development process and are looking for volunteers.<br />
<br />
11. What is the relationship of SHR with FSO ( A bit about FSO too). Is SHR an user interface on top of FSO or does it offer something more?<br />
<br />
12. Can someone highlight the Short term, medium term and long term benefits of using SHR to Hardware hackers and hardware manufacturers? <br />
<br />
13. What milestones does SHR hope to hit in 2010?<br />
<br />
13. Where do I find all the code and documentation for SHR?<br />
<br />
14. When will SHR/ Stable be available :)<br />
<br />
ANSWERS<br />
<br />
Martin.Jansa@gmail.com<br />
<br />
4. How does the development process of SHR work ( As in who/ who all decide<br />
> what is to be doe at what priority)?<br />
<br />
<br />
Every developer does what he likes/needs the most, when there is some<br />
controversial change then we usually discuss it before on<br />
#openmoko-cdevel.<br />
<br />
<br />
<br />
> 6. How far has the development of SHR progressed? Is the current SHR image<br />
> usable out of the box?<br />
<br />
<br />
That's questions more for our users, but AFAIK all devs are able to use<br />
images on daily basis and I think it's usually usable out of the box.<br />
But of course there is some breakage from time to time, because it's<br />
really living development and almost no testing before building stuff<br />
in shr-u (that's why shr-t exists)<br />
<br />
<br />
<br />
> 8. If a S/W developer wishes to contribute to SHR what should she/ he do?<br />
> Does she/ he need to have any specific skill sets to contribute to SHR?<br />
<br />
<br />
Personally I think that best start is to build own images<br />
http://wiki.shr-project.org/trac/wiki/Building%20SHR<br />
and then start improving any aplication you like or just improving<br />
recipes which helps a lot too.<br />
<br />
Normal linux hacker skills like C/python/shell knowledge are usefull for<br />
fixing build failures, creating patches.<br />
<br />
<br />
<br />
> 9. How does a developer make sure her/ his application runs on SHR. Is there<br />
> any centralised verification process or a repository where verified<br />
> applications are stored?<br />
<br />
<br />
I don't know about any verification and IMHO it doesn't make much sense<br />
when distro is moving fast (ie EFL libs are upgraded often, fso,<br />
phoneuid, phonefsod is developing fast), so even verified app is not<br />
quaranteed to run a month later.<br />
<br />
And keep backward compatibility in this state is non'sense (as our<br />
upstream also doesn't guaranetee compatibility - yet).<br />
<br />
> 10. Can you list the main areas where you would appreciate help in the<br />
<br />
> SHRdevelopment process and are looking for volunteers.<br />
<br />
><br />
> 11. What is the relationship of SHR with FSO ( A bit about FSO too).<br />
<br />
> Is SHRan user interface on top of<br />
<br />
> FSO or does it offer something more?<br />
<br />
<br />
FSO is whole middleware, SHR consists of UI on top of FSO and also<br />
bundles it with other usefull applications to easy flashable images.<br />
<br />
<br />
> 12. Can someone highlight the Short term, medium term and long term benefits<br />
> of using SHR to Hardware hackers and hardware manufacturers?<br />
><br />
> 13. What milestones does SHR hope to hit in 2010?<br />
<br />
<br />
Hi hope to see 2.6.32 kernel used by default as it will make next<br />
upgrades to 2.6.33 and newer version much more easily and thanks to<br />
great work of Lars-Peter Clausen, Thomas White and other kernel hackers<br />
is as usable as 2.6.29-rc3 was and much better shape.<br />
<br />
<br />
> 13. Where do I find all the code and documentation for SHR?<br />
<br />
<br />
Build metadata<br />
http://git.openembedded.org/<br />
Apps sources<br />
http://git.shr-project.org/git/<br />
FSO sources<br />
http://git.freesmartphone.org/<br />
And other stuff (urls defined in build metadata - recipe files).<br />
<br />
<br />
> 14. When will SHR/ Stable be available :)<br />
<br />
<br />
No idea, I don't think we should hurry with that. Even with shr-testing<br />
I see the case when users are expecting new stuff which happens only in<br />
shr-unstable to be available in shr-testing soon.<br />
<br />
If there is shr-stable I guess it will be just released and then no more<br />
upgrades to it, so from user POV I don't see any advantage over<br />
shr-testing.<br />
<br />
<br />
<br />
<br />
<br />
Sebastian Krzyszkowiak<br />
<br />
> 1. What is SHR?<br />
<br />
<br />
SHR is a embedded distribution aimed to give users power of free<br />
software in their smartphones. It's based on OpenEmbedded build system<br />
and on FSO phone stack, which is the best approach to smartphone<br />
middleware for us (well, at least in my opinion).<br />
<br />
<br />
> 2. Can you tell a bit about the history of SHR? Why was the development of<br />
> SHR started? What were the aims and objectives?<br />
<br />
<br />
I joined SHR development a bit later, but AFAIK SHR was created after<br />
announcement of ASU (later called Om2008) and its first goal was to<br />
port 2007.2 applications to FSO stack and to use them in ASU<br />
environment (that's why it's called Stable *Hybrid* Release).<br />
<br />
<br />
> 3. What are the target platforms for SHR? Does SHR target only Openmoko<br />
> phones or is it designed as a generic mobile communication/ PDA solution?<br />
<br />
<br />
ATM we target only Openmoko phones, but only because they are the most<br />
open and most of us have them. We're keeping our eyes on effort to<br />
port FSO to HTC phones, Palm Pre and other devices and we're also<br />
interested in running SHR on them. So our aim is to be generic<br />
smartphone operating system and to run wherever FSO runs.<br />
<br />
<br />
> 4. How does the development process of SHR work ( As in who/ who all decide<br />
> what is to be doe at what priority)?<br />
<br />
<br />
We try to always look on what community wants and to not decide<br />
anything important without community acceptance. For organizational<br />
reasons, there is SHR core team, which consists of the most active SHR<br />
and FSO developers. Members of coreteam should be listed somewhere on<br />
wiki, I won't write it now because I don't want to skip anyone :)<br />
<br />
<br />
> 5. Who are the main developers working on SHR? What are their areas of<br />
> interest wrt SHR and what are they working on these days?<br />
<br />
<br />
I think TAsn and mrmoku (working mostly on libphone* libs, but not<br />
only) and JaMa (OpenEmbedded) are the most active developers ATM.<br />
Also, spaetz is maitaining shr-testing OE branch. I used to work on<br />
shr-settings, shr-installer, opimd and few others - but recently I<br />
didn't have time to do active development, so I don't know if I'm the<br />
most active anymore :) Also, there are some people who are sending to<br />
us patches to shr-devel maillist.<br />
<br />
<br />
> 6. How far has the development of SHR progressed? Is the current SHR image<br />
> usable out of the box?<br />
<br />
<br />
It depends, but mostly - yes. Most of us is using their SHR based<br />
Freerunners as primary, daily phones (me too).<br />
<br />
<br />
> 7. Can you explain the structure of SHR (as in kernel, middleawre windows<br />
> manager, data storage structure, application layer etc)?<br />
<br />
<br />
Kernel - Linux. Actually 2.6.29, but we're looking forward newer<br />
kernels (SHR supports them, but not out-of-box)<br />
As window manager we're using Enlightenment with Illume module (but<br />
there started first efforts to drop Illume and replace it with<br />
Illume2). Middleware is obvious - freesmartphone.org :) As data<br />
storage we're using one of FSO daemons - opimd. About applications -<br />
they are different. Some of them comes from GPE project, some of them<br />
were written by Openmoko community, and some of them were written by<br />
us for us ;)<br />
<br />
About architecture of SHR apps and libs - I think TAsn and mrmoku are<br />
the ones who should talk about it :)<br />
<br />
<br />
> 8. If a S/W developer wishes to contribute to SHR what should she/ he do?<br />
> Does she/ he need to have any specific skill sets to contribute to SHR?<br />
<br />
<br />
Of course no! There are many different ways users can contribute to<br />
SHR - code, API organization, distro mainternance, artwork, themes...<br />
We have running patchwork system on our shr-devel maillist, so patches<br />
are welcome and they won't be forgotten. Also, developers can just<br />
hang on IRC - #openmoko-cdevel channel on freenode - in order to talk<br />
about what they can do. Obtaining commit access to our git repos is<br />
easy, just send few patches, be active - and then, when you'll ask for<br />
access, you'll get it :)<br />
<br />
<br />
> 9. How does a developer make sure her/ his application runs on SHR. Is there<br />
> any centralised verification process or a repository where verified<br />
> applications are stored?<br />
<br />
<br />
Just send us one file - it's called "bitbake recipe" and has *.bb<br />
extension. More about writing them you can read on OpenEmbedded<br />
websites. After sending us this file, your app will be integrated in<br />
our main repositiories and we will have to keep eye on package - if it<br />
works, builds etc. - not app developer (and personally: that's why I<br />
don't like opkg.org :P)<br />
<br />
<br />
> 10. Can you list the main areas where you would appreciate help in the SHR<br />
> development process and are looking for volunteers.<br />
<br />
<br />
SHR testing and SHR stable maintainers with skills in OpenEmbedded.<br />
Also, some new themes would be welcomed :)<br />
<br />
<br />
> 11. What is the relationship of SHR with FSO ( A bit about FSO too). Is SHR<br />
> an user interface on top of FSO or does it offer something more?<br />
<br />
<br />
SHR is both user interface on top of FSO, and OE based distribution<br />
containing FSO, SHR apps and some other utilities.<br />
<br />
<br />
> 12. Can someone highlight the Short term, medium term and long term benefits<br />
> of using SHR to Hardware hackers and hardware manufacturers?<br />
<br />
<br />
For hardware hackers: it gives you a full access to hardware. It's<br />
unlocked, full Linux, like on PC.<br />
<br />
<br />
> 13. What milestones does SHR hope to hit in 2010?<br />
<br />
<br />
I think SHR stable release. Also, we look forward to improvements in<br />
opimd done by TAsn and polishing user interface of phone apps.<br />
<br />
<br />
> 13. Where do I find all the code and documentation for SHR?<br />
<br />
<br />
http://git.shr-project.org/<br />
<br />
<br />
> 14. When will SHR/ Stable be available :)<br />
<br />
<br />
When it'll be ready :)<br />
<br />
(I know I didn't answer exhaustive to all questions, but I hope<br />
answers from others will be complementary to mine ;))<br />
<br />
Regards!<br />
<br />
<br />
> > 1. What is SHR?<br />
SHR is the abbreviation for 'Stable Hybrid Release', which comes from its<br />
origins. In the beginning of OpenMoko there was a GTK based distro, which then<br />
got replaced by an EFL based one. SHR was trying to combine both worlds.<br />
That's where the 'Hybrid' comes from - at least that's what I remember :)<br />
<br />
> ><br />
> > 2. Can you tell a bit about the history of SHR? Why was the development<br />
> > of SHR started? What were the aims and objectives?<br />
See 1. It was started because existing distros for the OpenMoko phones did<br />
suck (and at that time there were not that plenty as today). The other<br />
objective was to have a community based distro, which fullfills the needs of<br />
the community and not of some company.<br />
<br />
> ><br />
> > 3. What are the target platforms for SHR? Does SHR target only<br />
> > Openmokophones or is it designed as a generic mobile communication/<br />
> > PDA solution?<br />
SHR started as a Distro for the OpenMoko phones. We would really like to see<br />
it (and I'm convinced we *will* see it) running on other devices too. Next<br />
most probable candidates are the Palm Pre and the HTC Dream.<br />
<br />
> ><br />
> > 4. How does the development process of SHR work ( As in who/ who all<br />
> > decide what is to be doe at what priority)?<br />
><br />
> Every developer does what he likes/needs the most, when there is some<br />
> controversial change then we usually discuss it before on<br />
> #openmoko-cdevel.<br />
><br />
> > 5. Who are the main developers working on SHR? What are their areas of<br />
> > interest wrt SHR and what are they working on these days?<br />
><br />
> Most patches in last few months flow from:<br />
> mrmoku - Klaus Kurzmann<br />
> - whole SHR telephony stack<br />
> TAsn - Tom Hacohen<br />
> - whole SHR telephony stack<br />
> spaetz - Sebastian Spaeth<br />
> - whole SHR telephony stack, shr-testing maintainer<br />
> dos - Sebastian Krzyszkowiak<br />
> - whole SHR telephony stack, themes, enlightenment stuff<br />
> JaMa - Martin Jansa<br />
> - build metadata<br />
> Heinervdm - Thomas Zimmermann<br />
> - build metadata<br />
><br />
> but also lots of developers working on stuff which makes SHR better<br />
> all OE devs<br />
> - build metadata<br />
> Michael 'Mickey' Lauer<br />
> - incredible work on FSO<br />
> Lars-Peter Clausen<br />
> - guru for new kernel version<br />
> Thomas White<br />
> - guru for new DRM driver<br />
> Jesus McCloud<br />
> - great themes<br />
> Other 3rd party app developers<br />
> - great apps :)<br />
><br />
> I hope I didn't forget someone<br />
><br />
> > 6. How far has the development of SHR progressed? Is the current SHR<br />
> > image usable out of the box?<br />
><br />
> That's questions more for our users, but AFAIK all devs are able to use<br />
> images on daily basis and I think it's usually usable out of the box.<br />
> But of course there is some breakage from time to time, because it's<br />
> really living development and almost no testing before building stuff<br />
> in shr-u (that's why shr-t exists)<br />
><br />
> > 7. Can you explain the structure of SHR (as in kernel, middleawre windows<br />
> > manager, data storage structure, application layer etc)?<br />
SHR is based on a linux kernel, FSO as middleware and Enlightenement/Illume as<br />
window manager. On top of that we have our telephony stack consisting of the<br />
following two parts:<br />
- phonefsod: a system daemon interacting with the FSO middleware, eg. reacting<br />
to incoming phone calls<br />
- phoneuid: a GUI daemon implementing the user interface, eg. the dialer<br />
screen<br />
<br />
> ><br />
> > 8. If a S/W developer wishes to contribute to SHR what should she/ he do?<br />
> > Does she/ he need to have any specific skill sets to contribute to SHR?<br />
><br />
> Personally I think that best start is to build own images<br />
> http://wiki.shr-project.org/trac/wiki/Building%20SHR<br />
> and then start improving any aplication you like or just improving<br />
> recipes which helps a lot too.<br />
><br />
> Normal linux hacker skills like C/python/shell knowledge are usefull for<br />
> fixing build failures, creating patches.<br />
><br />
> > 9. How does a developer make sure her/ his application runs on SHR. Is<br />
> > there any centralised verification process or a repository where<br />
> > verified applications are stored?<br />
><br />
> I don't know about any verification and IMHO it doesn't make much sense<br />
> when distro is moving fast (ie EFL libs are upgraded often, fso,<br />
> phoneuid, phonefsod is developing fast), so even verified app is not<br />
> quaranteed to run a month later.<br />
><br />
> And keep backward compatibility in this state is non'sense (as our<br />
> upstream also doesn't guaranetee compatibility - yet).<br />
><br />
> > 10. Can you list the main areas where you would appreciate help in the<br />
> > SHRdevelopment process and are looking for volunteers.<br />
One thing we would need very much is a maintainer for SHR stable. Other than<br />
that any help is of course appreciated :-)<br />
<br />
> ><br />
> > 11. What is the relationship of SHR with FSO ( A bit about FSO too).<br />
> > Is SHRan user interface on top of<br />
> > FSO or does it offer something more?<br />
><br />
> FSO is whole middleware, SHR consists of UI on top of FSO and also<br />
> bundles it with other usefull applications to easy flashable images.<br />
><br />
> > 12. Can someone highlight the Short term, medium term and long term<br />
> > benefits of using SHR to Hardware hackers and hardware manufacturers?<br />
The immediate benefit for Hardware hackers probably is the openess - having a<br />
standard Linux system without black holes (with the obvious exception of some<br />
parts as the GSM firmware). For Hardware manufacturers I don't know. Maybe a<br />
benefit _could_ be to get the software for 'free'. I doubt they are interested<br />
in that though :/<br />
<br />
> ><br />
> > 13. What milestones does SHR hope to hit in 2010?<br />
><br />
> Hi hope to see 2.6.32 kernel used by default as it will make next<br />
> upgrades to 2.6.33 and newer version much more easily and thanks to<br />
> great work of Lars-Peter Clausen, Thomas White and other kernel hackers<br />
> is as usable as 2.6.29-rc3 was and much better shape.<br />
In addition I hope to see it running on other devices such as the Palm Pre or<br />
the HTC Dream.<br />
><br />
> > 13. Where do I find all the code and documentation for SHR?<br />
><br />
> Build metadata<br />
> http://git.openembedded.org/<br />
> Apps sources<br />
> http://git.shr-project.org/git/<br />
> FSO sources<br />
> http://git.freesmartphone.org/<br />
> And other stuff (urls defined in build metadata - recipe files).<br />
User manual<br />
http://wiki.openmoko.org/wiki/SHR_User_Manual<br />
<br />
><br />
> > 14. When will SHR/ Stable be available :)<br />
><br />
> No idea, I don't think we should hurry with that. Even with shr-testing<br />
> I see the case when users are expecting new stuff which happens only in<br />
> shr-unstable to be available in shr-testing soon.<br />
><br />
> If there is shr-stable I guess it will be just released and then no more<br />
> upgrades to it, so from user POV I don't see any advantage over<br />
> shr-testing.<br />
yes, I completely agree with Martin. There is too much stuff missing so that<br />
users always will want to have new stuff merged into it. We were talking about<br />
renaming the current SHR-Testing to SHR-Stable, when the new SHR-Testing is<br />
ready. Personally I think that would just be stable in the sense of not<br />
getting updates - and not in the sense of having a complete, stable phone ;)<br />
<br />
BUT, stuff is progressing fast these days. FSO2 (the vala-rewrite of FSO) is<br />
not missing much (many thanks to Mickey Lauer for that - he's doing that<br />
incredible work almost alone). And our phone stack is progressing too. I would<br />
not be surprised if we reach something I would call a good and stable phone<br />
distro before this year ends :-)<br />
<br />
Regards<br />
--<br />
Klaus 'mrmoku' Kurzmann</div>Rakshat