Active Topics
-
Extra softwares in Sailfish using CLI, repositories, etc (117)
to SailfishOS by nieldk - 5 hrs, 54 mins ago -
Firefox with Leste (6)
to Maemo 7 / Leste by teroyk - 1 day, 2 hrs ago - more...
 |
bzbnd |
2010-09-16
, 11:03
|
|
Posts: 209 |
Thanked: 76 times |
Joined on Mar 2010
|
#211
|
i cant wait anymore . any news?
 |
|
2010-09-16
, 20:47
|
|
Posts: 346 |
Thanked: 139 times |
Joined on Jan 2008
@ Houston Texas
|
#212
|
Yes, don't mean to be impatient. But if possible, a small update on the development would be great 
.
.
 |
|
2010-09-22
, 14:22
|
|
Posts: 2,473 |
Thanked: 12,265 times |
Joined on Oct 2009
@ Jerusalem, PS/IL
|
#214
|
School + boring lessons + classmate had a MyBook + me missed a step = hostmode reproduced with same patches but proper instructions (i.e I can reproduce it 100% of the time, unless I make a mistake)
Updates on this later on, don't thank this post :P
Last edited by MohammadAG; 2010-09-22 at 14:28.
Updates on this later on, don't thank this post :P
Last edited by MohammadAG; 2010-09-22 at 14:28.
|
|
2010-09-22
, 14:59
|
|
Posts: 1,341 |
Thanked: 708 times |
Joined on Feb 2010
|
#215
|
Originally Posted by MohammadAG
You didn't say "please".
don't thank this post :P
| The Following User Says Thank You to zimon For This Useful Post: | ||
 |
|
2010-09-23
, 10:15
|
|
Posts: 2,222 |
Thanked: 12,651 times |
Joined on Mar 2010
@ SOL 3
|
#216
|
Hostmode has gained a bit of traction recently.
PaulFertser, Klaus mrmoku Kurzmann and SHR crew, Michael mickeyl Lauer got their development devices recently (thanks to Ryan Abel, Quim Gil, and Rst38h for making this possible), and especially Paul is starting to hack away and brings a lot of new enthusiasm to the h-e-n project.
As Paul is relatively new to h-e-n, we are co-working to create a introduction and state_of_research summary paper, while he is catching up with what's been happening and got revealed during the last 6 months. Such summary should help new developers to get an idea about h-e-n and the musb-core more easily.
Regarding new developers: It seems Mohammad finally managed to track down which were the important details of his procedure when he made hostmode work with Sarah Newman's debugging kernel (plus some more patches we stacked on top of it). So today he's going to ship a proof-of-concept package (http://talk.maemo.org/showthread.php...298#post824298) or whatever, plus instructions how to reproduce his success (he claims it works on >5 different peripherals for him, all of them storage devices). Though this package with the 'shiny GUI' looks like targeted to production, it's strictly a DEBUG kernel with a lot of nasty botches. The GUI forces the kernel through all the hw controller and software state machine steps by exploiting the debugging options implemented by Sarah, and it will basically not do 'the right thing' - i.e. it doesn't detect USB connects or disconnects, it is fixed to USB 2.0 mode, it doesn't correctly negotiate power consumption with peripheral, etc pp.
[edit] It turns out all the logs of successful connections I've seen so far (not many) seem to indicate this hostmode is actually achieved by tricking the device into thinking it's on OTG and has to do a HNP (a method to swap host and peripheral roles of two devices connected in a OTG session). As this HNP needs manipulations of VBUS 5V which are rather timecritical and sensitive to electrical characteristics of all involved devices, there's probably the explanation why it fails for several people rsp their N900. h-e-n kernel devel team will look into another method now, that should yield more reliable and reproducible results[/edit]
The purpose of this package is NOT to bring hostmode to the masses, but it's meant to facilitate other devels to give a try to current state of what we basically achieved so far - and we are NOT INTERESTED in support requests, nor in "works for me" or "boohoohoooo It fails!!1!1!". What we are interested in are people having a look into the project, wrap their head around the way it works, and create decent logs to point out what in kernel needs improvement, or preferably even join h-e-n and collaborate in development and contribute patches. And that's what this package is meant for.
According to Mohammad it should work with quite the most of memory sticks, by simply connecting the peripheral to N900 directly, usually using the CA-101 cable and a USB-A-F2F (e.g.:http://www.conrad.de/ce/de/product/9...BUCHSE-ADAPTER ) adapter. Reports about N8 OTG adapter welcome.
This does not need any hardware hacks or special adapters grounding the ID pin (aka OTG cables).
Mohammad supplied a syslog on h-e-n.garage.maemo.org, and there will be further documentation and development showing up there in the near future.
Please support hostmode development, join h-e-n.garage.maemo.org if you are a kernel hacker knowing about USB, or provide nice logs if you are a technically savvy user.
And please don't spam this thread with noisy posts. If you want to thank us, there's a 'Thanks!' button on every post, if you want updates you simply wait here for them to show up, if you want details you go to h-e-n.
cheers
jOERG
Last edited by joerg_rw; 2010-09-26 at 05:26.
PaulFertser, Klaus mrmoku Kurzmann and SHR crew, Michael mickeyl Lauer got their development devices recently (thanks to Ryan Abel, Quim Gil, and Rst38h for making this possible), and especially Paul is starting to hack away and brings a lot of new enthusiasm to the h-e-n project.
As Paul is relatively new to h-e-n, we are co-working to create a introduction and state_of_research summary paper, while he is catching up with what's been happening and got revealed during the last 6 months. Such summary should help new developers to get an idea about h-e-n and the musb-core more easily.
Regarding new developers: It seems Mohammad finally managed to track down which were the important details of his procedure when he made hostmode work with Sarah Newman's debugging kernel (plus some more patches we stacked on top of it). So today he's going to ship a proof-of-concept package (http://talk.maemo.org/showthread.php...298#post824298) or whatever, plus instructions how to reproduce his success (he claims it works on >5 different peripherals for him, all of them storage devices). Though this package with the 'shiny GUI' looks like targeted to production, it's strictly a DEBUG kernel with a lot of nasty botches. The GUI forces the kernel through all the hw controller and software state machine steps by exploiting the debugging options implemented by Sarah, and it will basically not do 'the right thing' - i.e. it doesn't detect USB connects or disconnects, it is fixed to USB 2.0 mode, it doesn't correctly negotiate power consumption with peripheral, etc pp.
[edit] It turns out all the logs of successful connections I've seen so far (not many) seem to indicate this hostmode is actually achieved by tricking the device into thinking it's on OTG and has to do a HNP (a method to swap host and peripheral roles of two devices connected in a OTG session). As this HNP needs manipulations of VBUS 5V which are rather timecritical and sensitive to electrical characteristics of all involved devices, there's probably the explanation why it fails for several people rsp their N900. h-e-n kernel devel team will look into another method now, that should yield more reliable and reproducible results[/edit]
The purpose of this package is NOT to bring hostmode to the masses, but it's meant to facilitate other devels to give a try to current state of what we basically achieved so far - and we are NOT INTERESTED in support requests, nor in "works for me" or "boohoohoooo It fails!!1!1!". What we are interested in are people having a look into the project, wrap their head around the way it works, and create decent logs to point out what in kernel needs improvement, or preferably even join h-e-n and collaborate in development and contribute patches. And that's what this package is meant for.
According to Mohammad it should work with quite the most of memory sticks, by simply connecting the peripheral to N900 directly, usually using the CA-101 cable and a USB-A-F2F (e.g.:http://www.conrad.de/ce/de/product/9...BUCHSE-ADAPTER ) adapter. Reports about N8 OTG adapter welcome.
This does not need any hardware hacks or special adapters grounding the ID pin (aka OTG cables).
Mohammad supplied a syslog on h-e-n.garage.maemo.org, and there will be further documentation and development showing up there in the near future.
Please support hostmode development, join h-e-n.garage.maemo.org if you are a kernel hacker knowing about USB, or provide nice logs if you are a technically savvy user.
And please don't spam this thread with noisy posts. If you want to thank us, there's a 'Thanks!' button on every post, if you want updates you simply wait here for them to show up, if you want details you go to h-e-n.
cheers
jOERG
__________________
Maemo Community Council member [2012-10, 2013-05, 2013-11, 2014-06 terms]
Hildon Foundation Council inaugural member.
MCe.V. foundation member
EX Hildon Foundation approved Maemo Administration Coordinator (stepped down due to bullying 2014-04-05)
aka "techstaff" - the guys who keep your infra running - Devotion to Duty http://xkcd.com/705/
IRC(freenode): DocScrutinizer*
First USB hostmode fanatic, father of H-E-N
Maemo Community Council member [2012-10, 2013-05, 2013-11, 2014-06 terms]
Hildon Foundation Council inaugural member.
MCe.V. foundation member
EX Hildon Foundation approved Maemo Administration Coordinator (stepped down due to bullying 2014-04-05)
aka "techstaff" - the guys who keep your infra running - Devotion to Duty http://xkcd.com/705/
IRC(freenode): DocScrutinizer*
First USB hostmode fanatic, father of H-E-N
Last edited by joerg_rw; 2010-09-26 at 05:26.
| The Following 86 Users Say Thank You to joerg_rw For This Useful Post: | ||
aanckar, adhrie, ahmadamaj, ahmoda, allnameswereout, AlMehdi, asys3, attila77, avidscavenger, bblue, beli, Benson, bigears5000, BLIZZARD, borghal, bzbnd, cfh11, ChoMar, christoph, clasificado, Cue, debernardis, donaggio, Dousan, dvergin, e-yes, edgar2, egoshin, esiravegna, EugeneS, fw190, Fötus, GeneralAntilles, geohsia, gerdich, gryedouge, hardkorek, ivgalvez, ivyking, jgbreezer, jkq, johnel, joshv06, kolos, Larswad, LippiVan, LondonBenji, Magik, Marklar, Marshall Banana, Mentalist Traceur, mikec, mivoligo, MohammadAG, mr id, mveplus, nashith, nerfiaux, Netweaver, noobmonkey, OVK, pali, pelago, pillar, qole, RAZOR, rdwrig0ceo, Rob1n, S0urcerr0r, sjgadsby, slender, sophocha, Strawman, SubCore, ThomasAH, Transcend, twaelti, vabgeo, vkv.raju, Wanker, Wikiwide, woody14619, wovenstringz, yurka, zod | ||
|
|
2010-10-17
, 06:13
|
|
Posts: 2,222 |
Thanked: 12,651 times |
Joined on Mar 2010
@ SOL 3
|
#217
|
Seems I found a new 'magic sequence' that enables hostmode without relying on fake HNP.
Means more devices should work, no reboot should be needed, no special timings to observe.
Still needs confirmation by further tests done by the whole hostmode taskforce team, then we can fix hostmode.py in a few minutes, clean the kernel, and come up with a first semi-decent release \o/
See h-e-n ML for details, or just wait here for the rest of the story coming *really soon now* :-)
cheers
jOERG
if you want to express your excitement, please don't spam this thread! there's a thanks button right down there ->
[edit] 3 times confirmed for my devices (N900, allegedly Transcend 2GB stick, red, transparent, green LED)
For developers: https://garage.maemo.org/pipermail/h...10/000025.html (Attention! >800kB)[/edit]
[edit2]reports about devices not detected might be related to termination (cmd X/Y/Z) and speed (cmd I/i).[/edit2]
[edit3]Paul remarked there's a bug in termination setting of the patched kernel version I use. Mohammad is about to build a fixed one[/edit3]
[edit4]tried to help a little bit and checked and possibly fixed above mentioned bugs (plus some) - sent mail to h-e-n[/edit4]
Last edited by joerg_rw; 2010-10-18 at 21:47. Reason: update
Means more devices should work, no reboot should be needed, no special timings to observe.
Still needs confirmation by further tests done by the whole hostmode taskforce team, then we can fix hostmode.py in a few minutes, clean the kernel, and come up with a first semi-decent release \o/
See h-e-n ML for details, or just wait here for the rest of the story coming *really soon now* :-)
cheers
jOERG
if you want to express your excitement, please don't spam this thread! there's a thanks button right down there ->
[edit] 3 times confirmed for my devices (N900, allegedly Transcend 2GB stick, red, transparent, green LED)
For developers: https://garage.maemo.org/pipermail/h...10/000025.html (Attention! >800kB)[/edit]
[edit2]reports about devices not detected might be related to termination (cmd X/Y/Z) and speed (cmd I/i).[/edit2]
[edit3]Paul remarked there's a bug in termination setting of the patched kernel version I use. Mohammad is about to build a fixed one[/edit3]
[edit4]tried to help a little bit and checked and possibly fixed above mentioned bugs (plus some) - sent mail to h-e-n[/edit4]
__________________
Maemo Community Council member [2012-10, 2013-05, 2013-11, 2014-06 terms]
Hildon Foundation Council inaugural member.
MCe.V. foundation member
EX Hildon Foundation approved Maemo Administration Coordinator (stepped down due to bullying 2014-04-05)
aka "techstaff" - the guys who keep your infra running - Devotion to Duty http://xkcd.com/705/
IRC(freenode): DocScrutinizer*
First USB hostmode fanatic, father of H-E-N
Maemo Community Council member [2012-10, 2013-05, 2013-11, 2014-06 terms]
Hildon Foundation Council inaugural member.
MCe.V. foundation member
EX Hildon Foundation approved Maemo Administration Coordinator (stepped down due to bullying 2014-04-05)
aka "techstaff" - the guys who keep your infra running - Devotion to Duty http://xkcd.com/705/
IRC(freenode): DocScrutinizer*
First USB hostmode fanatic, father of H-E-N
Last edited by joerg_rw; 2010-10-18 at 21:47. Reason: update
| The Following 58 Users Say Thank You to joerg_rw For This Useful Post: | ||
anapospastos, arora.rohan, avidscavenger, beli, brad112358, bzbnd, cfh11, Chrome, clasificado, Cue, DaSilva, Dave999, debernardis, divan, donaggio, Dousan, edgar2, Exact, Flandry, fw190, gerdich, gravisrs, gryedouge, ironm8, ivyking, jaeezzy, jkq, johnel, jurop88, lemmyslender, LippiVan, ljo, Magik, maluka, Marklar, Mentalist Traceur, mike_shenoda, MohammadAG, mr id, mveplus, noobmonkey, osama, OVK, pierro78, RAZOR, rotoflex, S0urcerr0r, sjgadsby, sleytr, snakekontrol, Snoshrk, spanner, Strawman, SubCore, ThomasAH, treofant, xenkof, YoDude | ||
|
|
2010-10-18
, 13:11
|
|
Posts: 133 |
Thanked: 16 times |
Joined on Sep 2010
@ Balzan Malta
|
#218
|
Originally Posted by joerg_rw
I can't wait for this!! Thanks everyone for the logs and especially the developers.
Seems I found a new 'magic sequence' that enables hostmode without relying on fake HNP.
Means more devices should work, no reboot should be needed, no special timings to observe.
Still needs confirmation by further tests done by the whole hostmode taskforce team, then we can fix hostmode.py in a few minutes, clean the kernel, and come up with a first semi-decent release \o/
See h-e-n ML for details, or just wait here for the rest of the story coming *really soon now* :-)
cheers
jOERG
if you want to express your excitement, please don't spam this thread! there's a thanks button right down there ->
[edit] 3 times confirmed for my devices (N900, allegedly Transcend 2GB stick, red, transparent, green LED)
For developers: https://garage.maemo.org/pipermail/h...10/000025.html (Attention! >800kB)[/edit]
[edit2]reports about devices not detected might be related to termination (cmd X/Y/Z) and speed (cmd I/i).[/edit2]
[edit3]Paul remarked there's a bug in termination setting of the patched kernel version I use. Mohammad is about to build a fixed one[/edit3]
| The Following User Says Thank You to luketanti For This Useful Post: | ||
|
|
2010-10-22
, 21:08
|
|
Posts: 133 |
Thanked: 16 times |
Joined on Sep 2010
@ Balzan Malta
|
#219
|
Any Updates now?? How long is this going to take. At least a small update.
|
|
2010-10-22
, 21:26
|
|
Posts: 619 |
Thanked: 691 times |
Joined on Feb 2010
|
#220
|
they must be ex-blizzard devs (this is a compliment btw)
it is done when it is done
it is done when it is done