keesj

A few days ago the Maemo 5.0 Beta SDK was announced. I was more exited then I expected myself. I already played with the Alpha SDK and was able to install the Beta without pain on my amd64 based machine.

The install on amd64

The amd64 bits architecture is not officially supported as host for the SDK and in the past I have had troubles installing the SDK. This time the installer worked pretty well. But before you install the sdk and /or sbox you need to add "vdso32=0" to your kernel command line. this can be done in /boot/grub/menu.lst . I used the “tar.gz” based install on a “not supported” arch to install scratchbox.

root@ijssijs:~# ./maemo-scratchbox-install_5.0beta.sh -F -s /scratchbox

The only drawback of this method is that you have to start and stop scrachbox by hand when you want to use it. The “default” ./maemo-scratchbox-install_5.0beta.sh -F appeared to work at first but the install of the Maemo SDK failed after wards (I guess this is something related to fakeroot) . After that I follow the normal install using

keesj@ijssijs:~/downloads$ ./maemo-sdk-install_5.0beta.sh -y 

The biggest difference compared to the Maemo 4.x SDK in that the Xephyr X-server now requires the composite extension where it used to be disabled explicitly when starting the server

Xephyr :2 -host-cursor -screen 800x480x16 -dpi 96 -ac 

Trying some QT stuff

Trying some qt is really childs play. I always likes qt and apparently it only gets better. the qt4 project on Maemo created packages that by now are in the fremantle extras repository. simply add

http://repository.maemo.org/extras-devel/ fremantle free 

to your /etc/apt/source.list and apt-get install libqtgui4 libqt4-dev and qt4-demos . After that you can go to /usr/lib/qt4/examples and looks at the many demo's. The nice thing about this is that the demos are small get you started very quickly. I really fancy the clock example. it is very easy to create nice apps based on this.

OpenGL ES

Trying out opengl samples should be as easy as downloading the sdk from here
http://www.imgtec.com/powervr/insider/sdk/KhronosOpenGLES2xSGX.asp

and copying the file to the SDK directory.

SDKPackage/Builds/OGLES2$
 cp -r Include/* /scratchbox/users/$USER/targets/FREMANTLE_X86/usr/include/
 cp LinuxPC/Lib/* /scratchbox/users/$USER/targets/FREMANTLE_X86/usr/lib/
 cp -r LinuxPC/Include/* /scratchbox/users/$USER/targets/FREMANTLE_X86/usr/include/

however this doesn't work for me

Next for me will be to create a small app in qt!

(sorry for the edits' a wiki page would have been a better solution)

Velocity, the rapidity of motion, describes the speed at which we move as community. It describes the speed at which we can make changes to the Maemo world. We need to understand a few things about velocity in the community in order to function properly.

Understanding more about maemo velocity. First of all velocity can be hard to measure. We want to measure the progress Maemo is making. You might think that it is easy enough to measure but it is not.It is NOT is the sum of all the work that is done (work doesn't mean progress). It is not the sum of all the progress that is done either as not all the progress results in Maemo moving. It is the intersection of the progress made(more on that in the next paragraph). What would be a good measure to measure our velocity? Thinking in terms of maemo 5.0 would be wrong as 5.0 without 3rd apps would be useless. I think that "new users / week" or "new developer / week" might be a good unit as it makes everybody happy( Nokia and US) .

Given an amount of people we have in the community we have a certain amount of available horse-power that we can spend moving Maemo to increase it's velocity. The problem is that velocity != horse-power. And certainly Until LinuxTag the amount of horse-power we had available was not changing that much. What did we do with that power? We tried to keep up that's all we where not able to move forwards and spread our wings. We are an eco system that works at a certain speed and there is not much you can do about that can we? Would it help to enter more bugs in the bugtracker? I don't think so.

So we had LinuxTag and now have Niels and Dave. I am very happy to see them so busy. They certainly make progress in the documentation and packaging area. They also surely will increase our velocity in the end. What is this post about then? Niels and Dave are doing stuff we really wanted for a long time. Can we keep up with them as community or must we help increase the velocity by doing different things? can we simply wait for new developers to come? I certainly have a hard time keeping up with then.

Whe had a nice chat on the #maemo irc. It started by trying to define how a maemo hacker should
perform presentations at linuxtag. Because the obvious answer is called "noBounds" we started brainstorming
about the other possibilities. many possibilities involve having a "real" server and controlling
the real server using the nokia. This can be done using web 2.0 or vnc. an other alternative is using the tablet as mass-storage device but of course that is not sexy!.

We did not find the the answer , but we did came up with some nice ideas of what could be a nice goal for LinuxTag or a reason for gathering. I mean what is the point of see each other if you don't have a common goal!

Here are some ideas:

• Create a 4 times HD big wall of nokias
  
  • That is 10*9 devices
• Create a scroller using different tablets
• Allow people to use a tablet to draw on screen
• Create 3d images using images captured using tablets
  
  • Have the tablet create a sound chain reaction
  • record and playback
• Create a messhed network
• Use BT to show you are wearing a device
  
  • Broadcast your name
  • Play war, don't get beeped

I have never liked to use dos or windows. Perhaps it is because I am a Dutch an honest guy and did not want to run illegal software and neither did I want to pay for it. The possible communication I was able to have with real (windows running) people has always been compromised. When I used pine , they used html based emailing. GnomeNeeting and other hacks never worked because of NAT problems. nokia video calling beta also did not work as expected. That is why i have been asking my parent who live in France to install different programs over the years. But last week it happend

BIG SUCCESS: Skype video calling beta for linux worked with skype for windows and my kids where able to talk with their grand-parents. If that works I call it Ubiquitous! we had no problem with echo's and we talked for about one hour. for me this is really a big step in the right direction.

In the same week I have a good working gizmo video call with a other n800 user. That was a smaller success of course.

So guys. The real problem currently is that all that software has some constraints,
Skype video is not there for the n800 , gizmo video is not there for a normal linux distro and google/nokia beta calling is not usable.

Who will introduce Ubiquitous video calling?

A few brave men announced software and hardware in a fashion and openness I have never see before. I can not understand that we as community can even think about complaining about "Delays and Discount codes". We are informed in an unusual open fashion. With the it2008 beta for the N800 we even get a software preview. What more do you want?

Please have some respect for the people that offer such openness and think about what it means for a developer to be working against such an open deadline with open kitchen beta's