23/Sep: Kernel Source

[Guest rjm2k]

I have passed your comments on but i am. Also confused re the notification led and wakeup. If i miss a call or. Txt my back key flashes when the screen is off and in this state pressing the back key turns on the screen. I have flashnotify installed but not sure it changes things.

[Guest targetbsp]

Not on charge or usb, no music players running, my lock never been installed. All non OS apps killed - except the notification app otherwise it definitely won't do its thing. Using the rom that came with it, de-crapified using titanium backup.

Using the Vodafone text back test, I miss a text and then the back button flashes and can be used to bring the screen on. Only the back button (and the top button obviously!) can wake it - and only when the led is actually flashing at that precise time. If I don't have a missed text, or the phone is between bouts of flashing (I have it set to flash every 10 seconds), then the back button does nothing. So it must be asleep if my other buttons don't wake it right? Another reason I suspect it's in sleep mode, is that unlike when the screen is on, the flashing is erratic. It flashes different numbers of times and sometimes half flashes and only roughly on the schedule I requested. Like it's not sparing much power for accurate timing.

I think you're missing an app you need rather than there being anything wrong with the phone. Missed Calls is another such app, and much more popular, but it's not been updated in ages so I passed on that one. According to marketplace there are between 350,000 and 750,000 users of these apps combined so if it is a fault of the phone that it can't do this itself then it's a fault affecting a heck of a lot of handsets! And that's not counting the fact that the text clients can do it too with the repeat reminder setting - though all of them only seem to go as low as once per minute which isn't often enough for my tastes. And I don't know if their button behaviour is the same as above.

Edited November 15, 2010 by targetbsp

[Guest oh!dougal]

IF I'm complaining about standard Android behaviour, then "my bad".

But I don't think I am.

Sebastian404 and k0nrad, who seem fairly knowledgable android types, are both of the impression that what the standard phone does (or rather does not do) is NOT right.

Looking up "Missed Call" I found this

According to the standard notification functions that come with the Android device, the user is notified about missed calls and text messages by the continous blinking of the LED light on the device.

http://www.androidfanatic.com/component/co...or-android.html

I simply cannot believe that STANDARD android behaviour is to give NO notification flashing on battery power when asleep.

What these 'notification' programs seem to be doing is BRIEFLY waking the phone (but not the screen) to give the notification, then putting it back to sleep until the next flash (or vibrate or whatever).

That world explain --

1/ why pressing the button to wake only works during the flash (that's the only time the phone is awake, otherwise its ignored, as usual)

2/ why the flashing is slightly erratic (its overlaid on top of the standard flashing pattern that hasn't been 'getting out' to the LED, because the phone is asleep until Missed Call briefly wakes it).

I note that that review of Missed Call specifically says that it impacts battery runtime. How much it impacts is going to depend on what all gets turned on when the phone wakes for each flash, GPS? Bluetooth? Wifi? And battery impact will vary depending on how long, and how frequent, your choice of notification might be. Your mileage may will vary ...

Given that these programs permit enormous flexibility in tailoring these notifications, I will not accept that their popularity is evidence that STANDARD ANDROID notifications do not work AT ALL on battery power.

They meet a need for tailored notification.

This phone needs notification working even, and particularly, when asleep on battery power.

I grant that such programs might well provide a viable workaround, BUT -

I am trying to get feedback through to ZTE about the anomalous (and wrong-headed, and generally disliked) behaviour of their STANDARD product.

Frankly, anyone saying "there's an app for that" when the standard phone is not doing what a STANDARD ANDROID phone should do, seems to be rather missing the point.

And it bothers me more than it should, when people don't realise what the App they are recommending is actually doing - or what standard behaviour it is concealing.

Mine flashes in sleep mode if i set it to do so. Wasn't frequent enough for my tastes though so I installed Missed Reminder to have it flash more frequently.

Also confused re the notification led and wakeup. If i miss a call or. Txt my back key flashes when the screen is off and in this state pressing the back key turns on the screen. I have flashnotify installed but not sure it changes things.

/rant :rolleyes:

Edited November 15, 2010 by oh!dougal

[Guest targetbsp]

What these 'notification' programs seem to be doing is BRIEFLY waking the phone (but not the screen) to give the notification, then putting it back to sleep until the next flash (or vibrate or whatever).

That world explain --

1/ why pressing the button to wake only works during the flash (that's the only time the phone is awake, otherwise its ignored, as usual)

2/ why the flashing is slightly erratic (its overlaid on top of the standard flashing pattern that hasn't been 'getting out' to the LED, because the phone is asleep until Missed Call briefly wakes it).

That makes sense. And to back that up, I should also point out that the home button does indeed wake the screen as long as the back button is flashing. All times I'd tried it before were I guess prior to me discovering it will only awake while flashing. :rolleyes:

[Guest Sebastian404]

IF I'm complaining about standard Android behaviour, then "my bad".

But I don't think I am.

Sebastian404 and k0nrad, who seem fairly knowledgable android types, are both of the impression that what the standard phone does (or rather does not do) is NOT right.

Looking up "Missed Call" I found this http://www.androidfanatic.com/component/co...or-android.html

I simply cannot believe that STANDARD android behaviour is to give NO notification flashing on battery power when asleep.

Being the anti-social type that I am (bah-humbug) I've never really noticed the notification thing on any of my other Android devices, tho I tend to use my iPhone as my main phone, so the very few people who call me wont be doing it on an Android device...

BUT, yes my understanding of that is they it should flash while alseep... I will see if I can find someone with an Android device and have an experiment..

[Guest oh!dougal]

... tho I tend to use my iPhone as my main phone ...

Oh! The irony! (But we promise not to tell!) :rolleyes:

Edited November 15, 2010 by oh!dougal

[Guest dadashi]

silly me, i thought there was some new news on the kernel....

[Guest rjm2k]

ZTE Response

New version of open source code

I asked about the source for the newer releases, france, hungary etc

The oped source code which you got from our website is the newest copy. Because when we fixed and released our open source code, the kernel is stable and we didnot modify unless there are serious errors.

ZTE will release V880 in the nearly 1 month in China, and i will post the open source code on our website as soon as possible.I have no permission to spread the code untill product list.

Wifi chip sleeping/Notification LED (back key)

I asked about the wifi sleep policy of never being ignored and the confusion over the led notification working or not working

I think this two point is very useful to enhance the user experience. I will convey this information to our User Experience Promote Project Manager and we will evaluate them for work times/ difficulty/ Feasibility etc.

Compatibility Issues

I asked about compatibility with routers in particular the netgear which was reported as having issues

Chinese people like domestic so we choose Chinese brand home routers such as ZTE/TP-link etc. So we have no idea on this issue...

Edited November 16, 2010 by rjm2k

[Guest Ally Weir]

this is good news, yes?

[Guest buneech]

They seem to be cooperative, hope something comes out of this. I guess they learned that they need to listen to customer feedback.

[Guest dadashi]

With the interest in their top seller San Francisco aka. ZTE Blade / Libero / Tactile etc... they are going to be if not already big players in smartphone markets...

HTC were unknown 6 years ago, selling OEM products under different names for O2, Telekom, T-mobile, etc... ZTE have been doing just the same but on a very short time scale, from OEM supplier to own brand in one year!

[Guest kendon]

The oped source code which you got from our website is the newest copy. Because when we fixed and released our open source code, the kernel is stable and we didnot modify unless there are serious errors.

haven't we established that this is a lie? has anyone compiled a working kernel from their sources without modifications?

I think this two point is very useful to enhance the user experience. I will convey this information to our User Experience Promote Project Manager and we will evaluate them for work times/ difficulty/ Feasibility etc.

right, bugs go to the "User Experience Promote Project Manager", i guess that explains quite a lot.

Chinese people like domestic so we choose Chinese brand home routers such as ZTE/TP-link etc. So we have no idea on this issue...

read: f you western people, buy our other crappy products to work with the crappy products you already bought.

idea is you won't complain about the wifi not working on your phone if your router isn't working in the first place.

not sure how any of this is good news. empty promises, mixed with marketing BS.

[Guest goatee]

To be fair, when I had a Pulse and Hero, I don't know of getting this kind of level of engagement with Huawei or HTC. It may not be perfect, but it's great to have a line in :rolleyes:.

haven't we established that this is a lie? has anyone compiled a working kernel from their sources without modifications?

right, bugs go to the "User Experience Promote Project Manager", i guess that explains quite a lot.

read: f you western people, buy our other crappy products to work with the crappy products you already bought.

idea is you won't complain about the wifi not working on your phone if your router isn't working in the first place.

not sure how any of this is good news. empty promises, mixed with marketing BS.

[Guest mELIANTE]

this is good news, yes?

I just read a bunch of FUs :rolleyes:

As for the Netgear router problems, I have one (CBVG834G) and I don't seem to have the problems experienced by most... I don't understand.

[Guest kendon]

To be fair, when I had a Pulse and Hero, I don't know of getting this kind of level of engagement with Huawei or HTC. It may not be perfect, but it's great to have a line in :rolleyes:.

i am happy to withdraw my accusations once we see improvement, but for now the facts are clear:

- we don't have the shipping kernel's sources

- wifi just sucks

i agree at least we got an answer, but we also got answers from htc all the time...

[Guest rjm2k]

The 2.1 stuff is an aside really, what we really want is the 2.2 stuff when it comes out, we can only hope that they will release the 2.2 source pretty much at the same time as the phone so that the source and binary are consistent with each other this time. I ask questions re 2.1 but tbh doubt we will get very far with it, 2.2 is the way to go.

[Guest goatee]

i am happy to withdraw my accusations once we see improvement, but for now the facts are clear:

- we don't have the shipping kernel's sources

- wifi just sucks

i agree at least we got an answer, but we also got answers from htc all the time...

Fair enough - I wasn't aware we got answers from HTC. BTW, thanks for your ROM - I was about to install it, when I saw flibblesan's 2.2 ROM which looked to be stable enough to use as my day to day ROM, and (for the last couple of hours since I installed it) seems to have proved itself.

[Guest rjm2k]

haven't we established that this is a lie? has anyone compiled a working kernel from their sources without modifications?

I suspect the comment re the source being the latest is not a lie, which is where the problem comes from, they appear to have only ONE copy of the kernel source for ALL phones, so they work on the SF kernel, then work on another phone altering the same source, then release the source by which time it's been broken for the SF. Poor source control is the issue.

[Guest rjm2k]

As for the wifi issue, is wifi not part of the MSM7227 chipset? If so, can't we just extract wifi drivers from other phones based on the same chipset to see if they are improvements or does that depend on being able to compile the kernel too?

http://pdadb.net/index.php?m=pdamaster lists many devices with android 2.1 based on this chipset.

[Guest kallt_kaffe]

I suspect the comment re the source being the latest is not a lie, which is where the problem comes from, they appear to have only ONE copy of the kernel source for ALL phones, so they work on the SF kernel, then work on another phone altering the same source, then release the source by which time it's been broken for the SF. Poor source control is the issue.

Found another problem caused by this... The touchscreen driver in the released source is not the same as the one used in the shipped kernel. With a stock kernel pinch zoom does work in the stock web browser but with my compiled kernel it does not. Multitouch test apps does detect two finger though. Propably some change in the way it handles multitouch.

The other issues (battery & power button) has been solved but it would really help if ZTE could give us the synaptic_i2c_rmi.c that they used with the stock kernels for the Orange and the French phones.

[Guest rjm2k]

Found another problem caused by this... The touchscreen driver in the released source is not the same as the one used in the shipped kernel. With a stock kernel pinch zoom does work in the stock web browser but with my compiled kernel it does not. Multitouch test apps does detect two finger though. Propably some change in the way it handles multitouch.

The other issues (battery & power button) has been solved but it would really help if ZTE could give us the synaptic_i2c_rmi.c that they used with the stock kernels for the Orange and the French phones.

Ok, have passed this on, with your compiled kernel what does 'cat /proc/touchscreen/ts_information' return? on the stock it says its a 2000 series with firmware TM1541

[Guest kallt_kaffe]

Ok, have passed this on, with your compiled kernel what does 'cat /proc/touchscreen/ts_information' return? on the stock it says its a 2000 series with firmware TM1541

Doesn't have that in my /proc

# find . -name touchscreen*

find . -name touchscreen*

# find . -name synapt*

find . -name synapt*

./sys/bus/i2c/drivers/synaptics-rmi-ts

./proc/irq/93/synaptics_touch

#[/code]


There's also an cypress_touch driver and a cypress_firmware.bin in /system/etc/


But I think that the synpatics-rmi-ts is the one that our blades use because when I replaced the source for it with a different version my touchscreen stopped working completely.

[Guest rjm2k]

Doesn't have that in my /proc

# find . -name touchscreen*

find . -name touchscreen*

# find . -name synapt*

find . -name synapt*

./sys/bus/i2c/drivers/synaptics-rmi-ts

./proc/irq/93/synaptics_touch

#

There's also an cypress_touch driver and a cypress_firmware.bin in /system/etc/

But I think that the synpatics-rmi-ts is the one that our blades use because when I replaced the source for it with a different version my touchscreen stopped working completely.

yeah, I think they built in drivers for 2 types of hardware, /proc/touchscreen/ts_information being missing could be a symptom of the issue you are having with the kernel source.

The full output from mine is

touchscreen module

name : synaptics

i2c address : 0x22

ic type : 2000 series

firmware version: TM1541

module: synaptics + TPK

[Guest kendon]

I suspect the comment re the source being the latest is not a lie, which is where the problem comes from, they appear to have only ONE copy of the kernel source for ALL phones, so they work on the SF kernel, then work on another phone altering the same source, then release the source by which time it's been broken for the SF. Poor source control is the issue.

ok, didnt read and comprehend that in full. so if the term "newest sources" is true, then we are back to a gpl violation. they need to provide the sources for the binary they ship, exactly those and not somehow modified sources.

and you actually believe the crap about using one kernel for all phones? you can't be serious. that would mean they wouldn't be able to compile a new blade kernel right now. imagine an oem shortage, they need to replace a random piece of hardware. how would they implement that in the kernel, go back to start and do it all again? or just get an tarball from the kernel archives storage, patch the sources for the changed hardware, and compile the kernel... and exactly this tarball is what we want, and what they are obliged to give to us. not the latest source, but the sources they compiled the kernel shipped on the phones from. down to the last character.

[Guest Frankish]

Then get back on at them about it.