Basically I want my RSS reader to fetch things (images for example) needed to display every entry during updates, so I can read them offline. Images in most feed entries are referenced remotely (http://), which are usually not downloaded until the entry is actually viewed. Some feeds use enclosures but that works more like an attachment rather than for content.

I’ve tried quite a few RSS readers and Straw seems to be the only one that does full automatic image fetch during updates. However Straw’s development has been stalling, and the latest version seems to be quite unstable.

Liferea has been my RSS reader for quite a while, and so I’ve decided to do it myself with (hopefully) the simplest way possible: a Liferea conversion filter which parses a feed and fetches things for offline reading.

At the moment it works by looking for <img> tags and fetches the image using wget, and then replaces the original image src to point to the local one.

It’s a pretty simple perl script. I have written it in a way so it can be extended it to parse and fetch other things in the future, maybe embedded videos for example. It currently downloads all images, one by one. It also checks if a file is already downloaded or not. You can change $SAVE_PATH in the script as needed.

You can git (yes, git) the script at git://pigeond.net/offline_filter.git. Or alternatively get the latest version here, or browse the repo at http://pigeond.net/git/?p=offline_filter.git.

To use it, set the script as the conversion filter for the feed you want to have things downloaded and it should just work.

Now I can read all the really important stuff on the train, like xkcd and failblog ;).

The first is checking for vim python support before doing anything. I did this because in Debian vim (vim.basic) isn’t compiled with python support, and I prefer vim.basic over vim.gtk or vim.gnome as my default vim (and as for the reason, see my previous vimpress post.

The second is adding a :BlogSave command. Basically sending and saving an entry without publishing it, i.e. saving it as a draft, which I do a lot myself.

I’ve kept it in my git repo now at git://pigeond.net/vimpress.git

Gitweb at http://pigeond.net/git/?p=vimpress.git

Well, not really.

I was looking for an alternative to posting blog entries using the web interface, and was randomly trying a few apps in Debian. There were drivel and gnome-blog but they don’t seem to support tags. BloGTK seems to, but there’s something weird about how it does, and it doesn’t support multiple categories. BloGTK is also being removed from Debian.

Then I came across vimpress - a vim plugin for wordpress. It’s simple, but it does what I needed. You can get a list of your blog entires, post a new blog, or even edit an existing blog. So I’m testing it with this blog entry and see how it goes.

The only small rant, not to vimpress, but to Debian’s vim, is that vimpress needs vim’s python support, but it’s not built into the vim.basic binary, which I prefer over vim.gtk or vim.gnome. One of the reasons is when I ctrl-z to background vim, and then foreground it (fg), it sometimes (not all the time) gets stuck unless I hit ctrl-c. This does not happen with vim.basic, but only vim.gtk and vim.gnome.

I did try to trace it and it is when the function mch_suspend calling pause(), which causes vim to wait for a signal.

Perhaps I should report to vim later…

On the other hand, I’ve been having some issues with tailor. Not until much much much later (about a week ago) that I discovered the cause and possibly the remedy. At the same time one of the FlightGear developers Tim Moore has figured out why git-cvsimport wasn’t happy, and more importantly how to make it happy.

And so I’ve re-imported FlightGear and SimGear source from scratch and it’s been going well so far. But for those who have been using my repo, you will have to re-clone it again.

I’ve also imported the FlightGear data repo, but beware, it’s rather big even for git. The bare repo is almost 1G. So if you’re happy with your current CVS checkout, I suggest you not to bother with my git one. Mind you, that’s the entire history. Considering a checkout is about 1.7G, that’s not too bad. I’ve also set a 512kB/s limit for my git upstream, just in case there are like 20 people cloning the data repo at the same time :)

• cannot copy 268MB of files from and to the same harddisk (i.e. the infamous “out of memory” error)?
• cannot show me the total size of a folder because the “General” tab stopped appearing in the properties, until I rebooted?
• I have to reboot 6 times for various reasons (network not working, change of settings, etc) within 3 hours of usage?

Answer: useless piece of junk.

Please. Save yourself. Save your friends. Save the world. Tell everyone what’s wrong with Vista. I’m a Linux guy, but I’m not asking you to run Linux. I’m just asking you not to use this piece of crap. Use something that you deserve.

Unfortunately for anyone who has subscribed to my rss feed, your reader will probably see all old posts appearing as new ones. I apologies if this causes any troubles.

Switching between the two networks is not a big deal. And with the help of ifplugd, the switching is automatic. But after all, they are two different interfaces. It would be good if I could switch between them without losing any existing connections. After a bit of googling around, I’ve found one solution: Ethernet channel bonding.

From the Linux source code Documentation/networking/bonding.txt: “The Linux bonding driver provides a method for aggregating multiple network interfaces into a single logical “bonded” interface”. In my case, I’m bonding the eth0 (wired) and the eth2 (wireless) together.

Under Debian, the bonding driver is built as a module bonding. I’ve also installed the ifenslave. So I put the bonding driver into my /etc/modules so it’s loaded at system startup:

bonding mode=1 miimon=500

where mode=1 sets the bonding mode to active-backup. The bonding driver also supports a few different load-balancing modes, which you might want to check out as well. miimon=500 sets the link check interval in milliseconds.

There’s also a primary parameter which you could specify which slave interface will always be used if it is available (e.g. primary=eth0).

The bonding driver gives you a new network interface bond0. Now to setup bond0, eth0 and eth2 in /etc/network/interfaces:

# the wired interface
iface eth0 inet manual
    hostname xxxxxxx

# the wireless interface
iface eth2 inet manual
    wireless-key s:xxxxxxxxxxxxx restricted
    wireless-essid xxxxxxxxxxxx
    wireless-mode managed

# the bonding interface, eth0 + eth2
iface bond0 inet dhcp
    pre-up ifconfig bond0 up
    pre-up ifenslave bond0 eth0 eth2
    pre-down ifenslave -d bond0 eth0 eth2

bond0 is set to use DHCP, while eth0 and eth2 are not, as they will act as slave interfaces. bond0 is the master. The 2nd pre-up command for bond0 attaches both eth0 and eth2 as slaves to the bonding device. Note that it is possible to configure a bonding device via sysfs, which is also documented in Documentation/network/bonding.txt.

While ifplugd takes care of eth0 going up and down, I also want to turn off the wireless completely when the wired network is plugged in. So I added my own script /etc/ifplugd/action.d/bonding:

#!/bin/sh
set -e

if [ ! -f /proc/net/bonding/bond0 ]; then
    exit
fi

case "$1" in
eth0)
    case "$2" in
    up)
        /sbin/ifdown eth2
        /sbin/ifconfig eth2 down
        /sbin/iwconfig eth2 txpower off
        ;;

    down)
        /sbin/iwconfig eth2 txpower auto
        /sbin/ifup eth2
        /sbin/ifconfig eth2 up
        ;;
    esac

    # I need to do this sometimes, so I've put it in
    # or could someone comment on this?
    ifdown bond0
    ifup bond0

    ;;
esac

This script will turn off the wireless radio for me when my eth0 wired network is up.

I’m relying on ifplugd to bring up bond0, which will then bring up eth0 and eth2. Hence in /etc/network/interfaces I did not set any of the interfaces to be auto. I also added -l to the arguments of ifplugd in /etc/default/ifplugd:

ARGS="-q -f -u0 -d10 -w -I -l"

so that it will run the “down” script on ifplugd startup if no cable is detected. That way my script will bring up my wireless eth2 if eth0 is not plugged, say at boot time.

With this setup, the bonding driver automatically switches to use eth0 or eth2, whichever is brought up. In my setup, eth0 and eth2 won’t be up at the same time. If you use the primary parameter, the primary slave interface will always be used, even if the other slaves become available.

/proc/net/bonding/bond0 shows the status of the bonding interface. For example I’m now on the wireless network:

Ethernet Channel Bonding Driver: v3.0.3 (March 23, 2006)

Bonding Mode: fault-tolerance (active-backup)
Primary Slave: None
Currently Active Slave: eth2
MII Status: up
MII Polling Interval (ms): 500
Up Delay (ms): 0
Down Delay (ms): 0

Slave Interface: eth0
MII Status: down
Link Failure Count: 0
Permanent HW addr: 00:11:2f:9c:57:25

Slave Interface: eth2
MII Status: up
Link Failure Count: 0
Permanent HW addr: 00:11:2f:9c:57:25

It’s pretty obvious that the currently active slave is my eth2 wireless. I can run something like watch -n 1 cat /proc/net/bonding/bond0, and as I plug the ethernet cable back, the active slave will become eth0. You can also find out the currently active slave via /sys/class/net/bond0/bonding/active_slave.

For completeness, here is the output of ifconfig:

bond0     Link encap:Ethernet  HWaddr 00:11:2F:9C:57:25
          inet addr:192.168.11.3  Bcast:192.168.11.255  Mask:255.255.255.0
          inet6 addr: fe80::211:2fff:fe9c:5725/64 Scope:Link
          UP BROADCAST RUNNING MASTER MULTICAST  MTU:1500  Metric:1
          RX packets:329306 errors:0 dropped:0 overruns:0 frame:0
          TX packets:329010 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:732552585 (698.6 MiB)  TX bytes:204431117 (194.9 MiB)

eth0      Link encap:Ethernet  HWaddr 00:11:2F:9C:57:25
          UP BROADCAST SLAVE MULTICAST  MTU:1500  Metric:1
          RX packets:303936 errors:0 dropped:0 overruns:0 frame:0
          TX packets:304693 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:245381818 (234.0 MiB)  TX bytes:27905638 (26.6 MiB)
          Interrupt:5 Base address:0xd800

eth2      Link encap:Ethernet  HWaddr 00:11:2F:9C:57:25
          inet6 addr: fe80::211:2fff:fe9c:5725/64 Scope:Link
          UP BROADCAST RUNNING SLAVE MULTICAST  MTU:1500  Metric:1
          RX packets:25370 errors:0 dropped:0 overruns:0 frame:0
          TX packets:24317 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:487170767 (464.6 MiB)  TX bytes:176525479 (168.3 MiB)
          Interrupt:5 Base address:0xe000 Memory:ff9fe000-ff9fefff

which shows bond0 being the master, and eth0 and eth2 being slaves.

It almost makes plugging and unplugging the ethernet cable fun. I can just sit there and keep replugging the cable alllllll day… :)

I’ve been using this setup for at least a week now and has been working very well. Please let me know if there’s anything I’ve missed, or any even better way of setting this up.

And of course it works under that proprietary . Then we realized it has rather serious issues with the composite-in being a bit slow (both slow frame rate and sometimes laggy).

As a Linuxer, naturally I then tried the device under Linux. Only just halfly expected, there is no driver for Linux yet. That’s no better way to spend your holidays overseas? I decided to spend a little on looking into the possibility to reverse engineer the windows driver.

Doing things on windows is always painful. Though I got one of the windows USB sniffer working, it wasn’t as convenient as doing things on Linux. So I went for using QEMU booting into windows and capture all the USB traffic. Unfortunately, windows (xp) crashes (classic BSOD style) while installing the device driver. I thought for a second, maybe it’s QEMU not emulating something that is needed?

Someone on #bochs pointed me to VirtualBox, so I gave it a go. Still out of luck though, windows xp crashes at the exact same spot. Damn.

On the bright side, I have discovered VirtualBox - yet another virtualizer for x86 hardware, and it’s pretty awesome:

• (On this Pentium M 1.6GHz laptop) VirtualBox is much much faster than QEMU (with kqemu) for, well, running windows xp.
• It has a GUI frontend for setting up VMs. As much as I prefer text/command-line based apps, a GUI frontend does help new users to try things out a lot.

Having said that, VirtualBox (currently anyway) is not entirely open source. It has two editions. And the version I tried was the one with those closed-source features. On the other hand they did say “some of these features will eventually be made available with the open-source version as well”.

So, no, I didn’t go very far with the reverse engineering. It’s usually a painful and time-consuming process. I did have some of the USB traffic sniffed and logged under windows, but I haven’t done much with them yet.

I have some more photos and info of the device here, including photos of it pulled apart.

I’ve finally switched my gateway box from my 12 years old Pentium (I) 166MHz gateway box (still running Debian potato, kernel 2.2. This is Linux for you, when it works, it can work forever…), to my 5-6 years old Netwinder.

I was also very glad that the Netwinder is now (or still) officially supported by Debian. I had Debian potato installed on it for years. Upgrading it was painless.

My old gateway box was running such an old kernel, and the firewall was one of those copied-from-others script using ipchains. This time I decided to use a better way to manage my firewall rules. After a bit of research and trying things out, I’ve chosen ipkungfu. Again, it was painless to setup.

The only not-so-good bit was, I did the machine switching-over only 1 day before I went away. So as it turned out, my net connection went down in about 2 days time. I always have some scripts checking if the net is up or not, and it would reset everything (bringing down/up the interface, renewing IP, relogging on, etc). But looks like that wasn’t helping either. So I thought it might be one of those cable-modem-needed-to-be-reset situation.

I asked my uncle to power cycle the cable modem for me. And just to be safe I asked him to do the same for the gateway box too. As expected my net connection went up again. And I was happy, for a while.

Who would have thought, it went down again. Suspiciously, it also only stayed up for about 2 days. My instinct told me it wasn’t a coincident. A few days later my uncle again power cycled the modem.

I got on and check the logs again and found that my firewall was blocking (and logging) a lot of IGMP packets from some mysterious 10.x.x.x (Telstra internal network instruments) IP. It’s strange though, I haven’t seen these logs when I was testing the net. So apparently just the bpalogin client heartbeat is not enough. Seems that without the IGMP packet response (or how-ever IGMP works), after a certain timeout period (2 days?) the other end terminates the net connection and stops leasing IP, even if the cable modem is on or even there are existing network connections. So of course, I’ve then updated my firewall rules.

Lesson learnt, testing the net for 1 day is not enough, cos it will go down in 2 days ;)

Once again, many thanks to my uncle for the trouble.

The quickest hack I did was the GE version FGMap, which shows a moving map of all the online pilots on the FG MP servers. It is basically a very simple change with the FGMap CGI that translate MP server info into XML, only this time it outputs KML. Very simple and straight forward, nothing fancy at all. Check it out by opening this file in GE (if no one is on the server, you’ll see nothing! Check FGMap too).

Quickly I found this Collada exportor plugin for Blender. As a test I loaded the C172P aircraft AC3D model from FG into Blender, followed by a few trail-and-error and I figured out what I need to export for GE: 1) Convert all faces to triangles, 2) Export Triangles, Disable Physics, Use UV Image and optionally Use Relative Paths.

According to the KML tutorial I have to pack up the .dae file together with the textures, along with a textures.txt for textures path mapping. However I discovered you could actually alter the path referenced in the .dae file to a local jpeg file, or even an HTTP URL.

So now I’ve got GE to show a 3D model of an aircraft from FG. Next is to get FG to talk to GE. There are many ways (network, serial port, etc) in FG you can export real time flight data. To make things easier for now, I’ve set it up to export simply latitude, longitude, altitude, pitch, roll, heading to a file (or actually a fifo), then I wrote a little program to read and output a KML file for GE to read. Then on the GE side, I have to create a KML with a NetworkLink to refresh the actual FG KML.

Now, problems. It seems even if I set the refresh time to be 0 seconds, and I can see GE refreshing the link continuously, on screen it still only updates the position/orientation about once per second. It could well be a GE internal limitation (or feature). The bottom line is, we need a sane way to feed data into GE.

Another thing is I can never get GE to “follow” the aircraft properly. It’s either not panning faster enough, or it will go kinda crazy and jitter a lot. Guess I’m not a GE/KML yet. I need to test it more and tidy up my codes/scripts/things a bit and I’ll put them up.

Nevertheless, it’s kind of fun to “fly FG inside GE”.

Screenshots of FlightGear running side by side with Google Earth, whereby FG sending data to GE and showing the 3D aircraft. Possibly more screenshots here

KSFO

Hong Kong with Computamaps Hong Kong 3D models

Well, the only thing I want to say is, I wonder why people don’t just stop using m$ windows when they first get the BSOD :)

A little bit of surprise, it ran with Wine latest development version, at least it’s gone thru all the splash screen, characters selection, until it gets to the actual game, crashed.

After turning on some debugging and tracing, followed by consulting on #winehq, I’ve learnt that this game hits one of the famous Wine and DirectX issue, multi-threaded Direct3D.

Still I’m a bit adventurous as usual, and did a bit of code reading and hacking in Wine, trying a few different quick-and-dirty approaches to get threads to use the same GLXContext, not that I’m an OpenGL guru.

After a few tries, I gave up. However I still kind of wanna be able to play this game under Linux. So I rethink the problem from a different approach, and did a even more dirty hack, so that only the main thread can do any GL calls. Any other threads will simply returns before trying. Yes, VERY UGLY indeed. So ugly that let’s not even talk about it anymore ;)

And guess what, that does the trick. It gets into the actual game. So apparently it’s creating a new thread only for the pre-game loading screen. So now instead of the loading animation screen, I simply get a black blank screen, and the game continues after that.

And just for fun, I’ve even created a character with the top Taiwanese model’s name :P

More screenshots here.

To find out more about FGLive, see here, and make sure you read the README as well.

See screenshots and videos of FlightGear here.

At the same time I did some “sightseeing” around, and checking out one of my favorite buildings, currently the tallest building in the world, the Taipei 101. Then I noticed the difference (pretty obvious) between the satellite images from GMap v1 and v2. The one from v1 is in fact newer than v2. The one in v2 is still being built!

GMap v1

GMap v2

And so I started off doing it. It’s heaps of fun, not to mention finding out all those bugs with IE. M$ is very kind, they always do things that make people laugh. This always gives me a chance to revise a lot with Javascript.

I also ended up writing my own routines for map overlays displaying plane markers and info, instead of using the GMAPI’s GMarker. A bit more flexible, and it does exactly what I want.

There is time when Javascript is useful, and there is time where it’s just overkill. IMHO, Javascript is useful for writing “web app”. Same as using Flash, or Java applet. But btw this does NOT mean every website should be bloody web app!!

And there you go, the FlightGear server online map. A couple of FlightGearers have been using it as a flying map and as a simple ATC radar. If it’s the right time you’ll see us flying!

I’ve got plans to add more features, like airports/navids lookup. And of course, improving the usability and UI is always on the plate. Comments are also welcomed…

Also thanks to everyone who has helped me with testing and giving suggestions to the page. (ampere, AJ, thorben, vivian, gorilla, Surge, johnh51, MasseR, etc)

Project deadlines at work as usual. It was still pretty good time though. And I got 18 cans of Lipton tea for the X’mas kringle thing. And they’re all gone already!

I’ve moved again. Well just the next road down. The move was sort of quick and easy without any significant drama. (And of course, we hired removal people to do it this time. I have to say they’ve done a pretty good job.)

So, just what can you do with an ipaq running Linux? Well, while I was packing all my stuff including computers, I managed to get bpalogin (a Telstra Bigpond cable login client) running on my ipaq. It was easy. I compiled it for strongarm and that’s it. Another thumbs-up for Linux and open source software. So with all my main computers packed and unplgged, I was still able to get online. I took a picture too :P

Click it for larger version

Coding… hadn’t really got time to do much.

I got at least 5 different patches from people all around the world for the gtk2/gkrellm2 port for bfm. I have to apologize for not doing anything with bfm at the moment. Hopefully in a few months time I will be putting together another release.

I was trying to get some sort of chinese inputing on my ipaq and it wasn’t that easy. I hacked xkbd a bit and got to the point so it is showing the changjei keyboard layout. However at that time Xft and Freetype weren’t cooperative enough to work perfectly on the ipaq. I’m pushing this off a bit and will deal with it later :P

I was playing with Evas again and getting my head around with its smart object. I started off with this Evas Staff widget. Hopefully I can write some useful programs using this in the future.

Also, I’ve joined the GPE gang and started working on gpe-irc, an IRC client (obviously), together with a few other developers. I’m mainly dealing with the backend stuff.

Hmm I’ve written a quick perl script to do english-to-engrish translation. If you don’t know what engrish is, visit engrish.com.

Piano… My piano tutor has found a better job, but it’s in Kaohsiung. And so she went back to Taiwan which means I’m now teacher-less :( Still, I’m keeping up with my piano every now and then. I’ll probably look for another tutor a bit later.

A bit of credits here for my piano tutor. Claudia, she’s very good at piano and music. And I guess she understands me a lot too in terms of my knowledge of piano and music. She knows what I’m good at and weak in. She knows what, where and how I’ve been thru piano and music. And so she’s done a very good job in guiding and teaching me. Not to mention we both share the same level of enthusiasm and passion towards piano and music. We can basically just sit there and talk all about piano non-stop.

Yet I have to see if I can find another tutor like that :)

With ip_masq_h323, I got OhPhone to talk to a NetMeeting client, via my firewall/gateway box.

ip_masq_netmeeting works well with the actual NetMeeting client under my firewall, but it doesn’t like OhPhone.

So I’m sticking to OhPhone with ip_masq_h323 now, with thanks to all these opensource effort. :)

So I was testing all these with Catty in Taiwan last couple of nights, thanks too :)

Spent some time on vim.sf.net trying a lot of useful scripts. Here are some which are really helpful and I’m actually using them.

• gdbvim
• a.vim
• bufexplorer
• taglist
• showmarks