Htc Universal, ... running Debian EABI Options

[Sam Liddicott]

The gtkdialog I have doesn't work properly on the universal, most of the button clicks handlers are not processed, but rather a critical assertion given.

This is a shame because glade-3 (which also runs on the phone!!) was helping do some nice phone GUI.

Sam

[Chero]

The gtkdialog I have doesn't work properly on the universal, most of the button clicks handlers are not processed, but rather a critical assertion given.

This is a shame because glade-3 (which also runs on the phone!!) was helping do some nice phone GUI.

Sam

I remember this problem with glade,
Try configuring the app with

LDFLAGS="-export-dynamic"

it helped me out with gnocky.
Chero.

[datathief]

I've added titchy-phoneui-base and titchy-phoneui-console to the Titchy repository.

There is one problem I think with whiptail now, which might be associated with key bindings. When in X, when in a call, the hangup button on the display flap, doesn't hangup. I.e. what hangs, I've found, in X, is double ESC, while the hangup button is not mapped to it. I vaguely recall it hanging up outside X, though don't know if I'm imagining, or it indeed does that. If yes, than probably some xbindkeys or xmodmap issue.

This is currently working for me, with the above packages.

And last thing missing from a default phone UI, well two actually, well three now that I think of it, are associated with loudness. I.e. already mentioned, when in a call, the volume keys on the side, should increase/decrease loudness.

Working.

When the phone rings, moving the volume key should mute.

Working.

And last thing is profiles, i.e. silent, vibe, ring.

I have a few ideas for this, but nothing working yet...

[Cagnulein]

Am i the only one who have problems with the duration of the battery? After 2 weeks of titchy i've obsverved that it runs for about 2 days with the full charge against the 4-5 days with WM6.

Is it normal? i've used the phone with wifi off.

[ShiroiKuma]

I've added titchy-phoneui-base and titchy-phoneui-console to the Titchy repository.

I get 403 Forbidden when trying to aptitude install the packages.

There's some permissions problem on your site.

From the browser:

CODE

Forbidden

You don't have permission to access /titchy/titchy-phoneui-base.deb on this server.
Apache/2.2.3 (Debian) mod_python/3.2.10 Python/2.4.4 PHP/5.2.0-8+etch7 mod_ssl/2.2.3 OpenSSL/0.9.8g Server at debian.neilandtheresa.co.uk Port 80

[datathief]

I've added titchy-phoneui-base and titchy-phoneui-console to the Titchy repository.

I get 403 Forbidden when trying to aptitude install the packages.

There's some permissions problem on your site.

From the browser:

CODE

Forbidden

You don't have permission to access /titchy/titchy-phoneui-base.deb on this server.
Apache/2.2.3 (Debian) mod_python/3.2.10 Python/2.4.4 PHP/5.2.0-8+etch7 mod_ssl/2.2.3 OpenSSL/0.9.8g Server at debian.neilandtheresa.co.uk Port 80

Oops. Fixed.

[Sam Liddicott]

Am i the only one who have problems with the duration of the battery? After 2 weeks of titchy i've obsverved that it runs for about 2 days with the full charge against the 4-5 days with WM6.

Is it normal? i've used the phone with wifi off.

Same problem here.
I loose about 2% battery per hour while on standby.

Sam

[Sam Liddicott]

The gtkdialog I have doesn't work properly on the universal, most of the button clicks handlers are not processed, but rather a critical assertion given.

This is a shame because glade-3 (which also runs on the phone!!) was helping do some nice phone GUI.

Sam

I remember this problem with glade,
Try configuring the app with

LDFLAGS="-export-dynamic"

it helped me out with gnocky.
Chero.

Good tip, thanks. Do you have any cross-building deb tips, or may I shall just re-build it on my phone (with swap over NFS over USB to save my poor SD card)

Sam

[ShiroiKuma]

Yes, WM keeps the correct time. But perhaps it's as you say, since I've set the "automatically get time from network" or whatever that is...

There's a command get the time from the modem, but it doesn't return the date.

It's possible that we could create some awful hack, like:

• Write the date to a file
• Reboot
• Read date from file
• Retrieve time from modem

This would only work if you restart Linux on the same day you reboot...

I wrote this hack. You can get the deb file from http://sumoudou.org/index.php?cmd=%E7%9B%B...tween%20reboots

[datathief]

I've added titchy-phoneui-base and titchy-phoneui-console to the Titchy repository.

There is one problem I think with whiptail now, which might be associated with key bindings. When in X, when in a call, the hangup button on the display flap, doesn't hangup. I.e. what hangs, I've found, in X, is double ESC, while the hangup button is not mapped to it. I vaguely recall it hanging up outside X, though don't know if I'm imagining, or it indeed does that. If yes, than probably some xbindkeys or xmodmap issue.

This is currently working for me, with the above packages.

Actually, this wasn't working. And with dash as the default shell, dialling out was quite broken. I've uploaded a fixed titchy-phoneui-console now.

[datathief]

Am i the only one who have problems with the duration of the battery? After 2 weeks of titchy i've obsverved that it runs for about 2 days with the full charge against the 4-5 days with WM6.

Is it normal? i've used the phone with wifi off.

Same problem here.
I loose about 2% battery per hour while on standby.

Sam

I'll do some testing, to try to find out what's not getting powered down properly.

[Sam Liddicott]

Am i the only one who have problems with the duration of the battery? After 2 weeks of titchy i've obsverved that it runs for about 2 days with the full charge against the 4-5 days with WM6.

Is it normal? i've used the phone with wifi off.

Same problem here.
I loose about 2% battery per hour while on standby.

Sam

I'll do some testing, to try to find out what's not getting powered down properly.

I'm guessing it's things waking up periodically.

When I worked for Orange on smartphones htc supplied an application which records when it wakes up to each wake-up level, so we could easily see if our customization apps were stopping the phone sleep properly.

I was thinking this morning that we may need something similar.

[Sam Liddicott]

Mono apps...

rather than tackle re-compiling gtkdialog (I'm still more at home with rpm's than debs - and cross-debs is no joke) I tried some mono phone apps.

Sadly, mono runtime(s) take at least 22% of phone RAM.
gtkdialog takes up 15% of phone RAM (according to top) (with more widgets too)

So.... I'm not too sure yet....

[datathief]

I'm guessing it's things waking up periodically.

When I worked for Orange on smartphones htc supplied an application which records when it wakes up to each wake-up level, so we could easily see if our customization apps were stopping the phone sleep properly.

I was thinking this morning that we may need something similar.

The different "levels" of suspend that Windows Mobile and presumably Symbian use don't really apply under Linux. If the processor is configured to wake on a given interrupt, it wakes fully. You can find out how often this is happening by looking at syslog:

CODE

grep htcuniversal_resume /var/log/syslog

You should find it only wakes on a GSM event (call or SMS), an alarm, or when you press the power button.

I know there are problems with the wifi driver's power management - not least that the kernel powers up wifi by default (loading the htcuniversal-acx module powers it down again). There may be more.

Edit: It looks like the bluetooth module is powered up by default. I've updated titchy-hardware-support to turn it off. I've no idea whether it will make much difference, though.

[Sam Liddicott]

Just for general interest;
here's my /etc/network/interfaces entry from my desktop:

CODE

allow-hotplug usb0
auto usb0
iface usb0 inet static
        address 192.168.2.1
        netmask 255.255.255.0
        post-up echo 1 > /proc/sys/net/ipv4/ip_forward; iptables -D FORWARD -j ACCEPT; iptables -A FORWARD -j ACCEPT; iptables -t nat -D POSTROUTING -s 192.168.2.0/24 -o eth0 -j MASQUERADE; iptables -t nat -A POSTROUTING -s 192.168.2.0/24 -o eth0 -j MASQUERADE

It doesn't go down when I unplug and so I have to ifdown/ifup when I re-plug or it doesn't get an address.

Here is my interfaces from my universal:

CODE

allow-hotplug usb0
iface usb0 inet static
        address 192.168.2.2
        netmask 255.255.255.0
        gateway 192.168.2.1
        post-up echo nameserver 4.2.2.2 > /etc/resolv.conf

And I finally got a kernel compile, not sure what I was doing wrong, but now I can start on the wlan stuff.