I'm far from being an urxvt expert so if you have better solutions please write me!

#! perl

sub on_start {
   my ($self) = @_;

   $self->{browser} = $self->x_resource ("urlLauncher") || "sensible-browser";

   ()
}

sub on_user_command {
   my ($self, $cmd) = @_;

   $cmd eq "selection2browser"
      and $self->exec_async ($self->{browser}, $self->selection);

   ()
}

I have this script saved as a file named selection2browser in my ~/.urxvt/perl/ directory and in order to activate it via Ctrl-Alt-w I put the following lines in ~/.Xresources:

URxvt.perl-lib: /home/shammash/.urxvt/perl/
URxvt.perl-ext-common: default,[other stuff],selection2browser
URxvt.keysym.C-M-w: perl:selection2browser

]]>

Roberto received his new LiveScribe Pulse. After playing a bit with his new toy he decided to attach it to the computer, but unfortunately the pen got bricked during a firmware upgrade.

So with my friend Filippo we tried to perform various kinds of hard reset, following what's described in 29101 - Mac Master Reset and firmware update utility, but with no success.

We were about to give up when, just out of curiosity, I decided to attach the pen to my Linux machine. Using "lsusb -vv" I've discovered that the pen was stuck in its Device Firmware Upgrade mode waiting for something interesting to happen.

Fantastic! We found the pen firmware inside LiveScribe Flash Utility and we uploaded it using "dfutool" from BlueZ... The pen came back to life, very good!

After this forced upgrade the pen was basically working, but it lacked some functionalities like voice recording. But now official LiveScribe utilities were able to see the pen, so we re-flashed using them and we obtained a 100% working device.

These are the steps we used to flash the LiveScribe Pulse from Linux:

$ wget http://www.livescribe.com/downloads/support/LivescribeSmartpenFlashUtility_2.1.1_Mac.zip
$ unzip LivescribeSmartpenFlashUtility_2.1.1_Mac.zip
$ cd 2.1.1\ Pen\ Flash\ Utility.app/Contents/Resources/Bundles/
$ unzip SystemSoftwareV2.1.1_C.bnd -d SystemSoftwareV2.1.1_C
$ cd SystemSoftwareV2.1.1_C/Files/
$ dfutool upgrade bali_ams3.dfu
Filename  bali_ams3.dfu
Filesize  3365572
Checksum  40559ec9 (valid)

Available devices with DFU support:

  1) Bus 004 Device 002: ID 1cfb:1020 Interface 2

]]>

Filippo on some performance analysis. At some point we
needed to send some traffic between two different interfaces on the same Linux
box, but we did not want the traffic routed through the local loopback.

The main problem here seems to be the lack of a direct and clean way to say in
Linux: "yes, this address is on this machine, but do not use the local loopback
for the traffic originating from the same host which needs to reach it".

Different flavours of this problem have been approached in many ways, there is
also a Send-To-Self patch for the kernel.

So this is my sol^W^Wan hack to achieve that using four IP addresses, NAT, and
a few simple changes to the routing table.

ip l s tap1 up
ip l s tap2 up
ip a a 10.10.1.2/32 dev tap1
ip a a 10.10.2.2/32 dev tap2
ip r a 10.10.2.1/32 dev tap1
ip r a 10.10.1.1/32 dev tap2
iptables -t nat -A POSTROUTING -o tap1 -j SNAT --to-source 10.10.1.1
iptables -t nat -A POSTROUTING -o tap2 -j SNAT --to-source 10.10.2.1
iptables -t nat -A PREROUTING -i tap1 -j DNAT --to-destination 10.10.1.2
iptables -t nat -A PREROUTING -i tap2 -j DNAT --to-destination 10.10.2.2
arp -i tap2 -Ds 10.10.1.1 tap1
arp -i tap1 -Ds 10.10.2.1 tap2

]]>

]]>

here.
I hope it will get upstream.
]]>

classe delle medie, mediamente piu` alti di tutti i libri di testo impilati.

Vabbeh, mettiamolo giusto per aggiornare questo blog :)

]]>

Last login: Wed Dec 24 01:40:40 2003 from 192.168.1.3 ]]>