1. GUADEC - last day

    Wednesday was last day of GUADEC for me. Getting from bed was hard — party and night talks made it really hard. We came to conference and started to put OH stand working and waiting for other guys to get here.

    There was GMAE meeting which I attended to get some idea what is going on. Many people, many different projects etc… One of interesting parts was during break when I got Apple iPhone in my hands — it’s interface is really fast and has some interesting ideas in it. For example Safari browser renders only lthat part of page which is displayed on screen — you can see it when fast scroll page. On screen keyboard is very easy to use with bare fingers — no need to look at it if you remember layout of standard QWERTY one. Many reviews noted ability to automatic screen rotation. It is not automatic — application needs to support it otherwise nothing will change when you rotate. Another extra stuff is multitouch screen — zooming works really great in apps which I tried (Maps, Photos, Safari). And the great part of touchscreen is that it is integrated into top of case — it is totally flat.

    Today I attended Lighting Talks — some of them are interesting (like GPE^2) some are not.

    Evening will be other — food at OH apartments, some beers and talking about anything. Tomorrow early morning wakeup and long trip back home — taxi, train, plane, taxi (about 10h in total including stops).

    Written by Marcin Juszkiewicz on
  2. GUADEC continued

    Tuesday I spend mostly on OpenedHand stand. We had ST Nomadik board there and it was my part of work to get it booting — with serial cable it was fast stuff. Then I updated it’s software to latest build. We also had few OpenMoko phones with some experimental software on it — new Today application is great — old one can not be compared.

    I decided to skip most of presentations as I did not found any interesting. There was one which I decided to saw — something about multi-user session handling. It was quite interesting but I think that it will require lot of work to get it done.

    During evening there was a GUADEC party at “Workabout” — few beers, lot of talks etc… There was a guy from iRex Technologies which we talk with — he will work on integrating his work with upstream OpenEmbedded version. Some parts will be not integrated because they can not (or not want) to open everything.

    After party we had discussion in OH apartment. I gave some ideas to Chris and Neil about new stuff for OpenMoko Today/Dialer stuff. Finally I was able to get to sleep.

    Written by Marcin Juszkiewicz on
  3. GUADEC - day 0

    I decided to call this day ‘Day 0’ because most of it was taken by going to GUADEC instead of being on conference.

    I had to wake up at 5:20 and get first train at 6:25 (Częstochowa -> Katowice). When I arrived at station I found out that there is one more train at 6:20 so I took it as this gave me few extra minutes in Katowice to catch bus to airport.

    After 50 minutes in bus I got to airport. It was full of people waiting for my flight — Wizzair to London Stansted. Queue was moving quite fast and I spent time talking with one guy about embedded devices (do not remember for which company he works — something with WiFi and long range comes to mind). After boarding we got first surprise — Wizzair do not number plane seats so first on plane == better choice (and we were last ones :( — so no talk during flight). Then plane captain gave us another bad information — flight was delayed due to local ‘rush hours’ in air ;( Finally we started — with over 40 minutes delay so I was not be able to catch train from Standsted Airport to Birmingham New Street at 12:25.

    Flight was ok — standard problem with lack of space for legs and some turbulations during landing. And I finally ended reading a book — Robert Ludlum ‘Covert One’ which I read for over 6 months.

    There was a funny part with my passport — on the photo I look like terrorist (which I am not) so I decided to show passport in Poland and my ID at Britain to avoid any extra questions.

    After coming to airport I took my laggage and bought train tickets (as I will go back very early there was no discounted tickets for it so buying tickets in advance was the same as buying Standard Return Ticket) and (as I had over hour to spend) took a look at airport’s terminal. I found Rynair desks so will not have to do it on Thursday, got something to eat/drink and returned to -1 level to catch the train.

    British trains… nearly no space for legs like in Airbus 320 and one interesting difference between Polish trains — space for luggage. Instead of solid rack where you can put everything you have to put it on special rack right after you enter wagon (because overhead racks are designed only for small items). Before that trip I read and heard that British trains are slow and unreliable but that one is going quite fast.

    Now it is time to shutdown laptop as battery in my D400 gives me less then 40 minutes of work (with backlight set to lowest readable). After GUADEC I will know do I need to buy a new one or can I live without it.

    Finally I got to Birmingham and take a walk to hotel. It took me about one hour to go that one mile because I got lost few times. Too many streets without any names…

    During evening there was OpenEmbedded meeting when we discussed many things which should improve our project.

    It was busy day…

    Written by Marcin Juszkiewicz on
  4. GUADEC: preparations to travel

    Time to really prepare for GUADEC trip. Tomorrow we are going to another wedding in family (3rd in this year) and during Monday I will travel to Birmingham to meet OpenedHand team (without Dodji ;( — visa laws are strange), OpenEmbedded guys, maybe someone from OpenMoko (Mickeyl does not count) and many other people working on misc projects.

    Travel times noted (trains, flights, buses), maps from train station to hotel printed, some GBP bought — so now it only left to pack. I even bought backpack for laptop so will be able to test how does my D400 behave (especially it’s battery as it does not give too much time). Good that I have that huge plug which is used in England already — one thing less to care about.

    And I hope that I will have more luck then people which were on Akademy and my luggage will follow me ;)

    Written by Marcin Juszkiewicz on
  5. Neo1973 now available!

    If you are one of those developers (or even normal users) who want to buy FIC Neo1973 (GTA01 version) you can do it online via OpenMoko webshop. Two versions available (Base and Advanced), payment only via credit card, shipment via UPS.

    So go and get it before they will send all available ones! Then unpack, hack, write software let OpenMoko platform grows!

    Written by Marcin Juszkiewicz on
  6. How to buy Neo1973 phone

    During last months most popular search which made people visit my page was like “neo1973 buy” (which is nothing strange as my website is on first page for that query). Also from time to time someone asks me on IRC about it.

    So how to buy Neo1973 phone? This is good question but it is hard to give proper answer. Basically few hundreds of them are already built and occupy space somewhere in China or Taiwan. Currently OpenMoko company works on making a way to sell them to developers (at least I think that this blocks selling). When phones will be ready to order proper information will be sent to announce mailing list. Until then there is no way to buy Neo1973.

    I hope that this will give answer to some people — but keep in mind that I am not a part of OpenMoko team so nothing in that post is official.

    UPDATE: Sean wrote on announce ML that phones will be on sale starting from 9th July 2007. Two configurations on sale:

    Neo Base — everything the mobile application developer needs to enjoy the benefits of the first freed phone, the Neo 1973:

    • Neo 1973 (GTA01B_v4)
    • Battery
    • Stylus
    • Headset
    • AC Charger
    • Phone Pouch
    • Lanyard
    • SanDisk 512MB MicroSD Card
    • Mini USB Connectivity Cable

    Neo Advanced — everything the mobile device hacker wants to get down and dirty with the first freed phone, the Neo 1973:

    • Neo 1973 (GTA01B_v4)
    • Battery (2x)
    • Stylus
    • Headset
    • AC Charger
    • Phone Pouch
    • Lanyard
    • SanDisk 512MB MicroSD Card (2x)
    • Mini USB Connectivity Cable (2x)
    • USB Host Mode Cable
    • Debug Flex Cable
    • Debug Board v2 (JTAG and serial console)
    • Ruggedized Toolbox with shoulder strap
    • Guitar Pick (for opening case)
    • Torx T6 screwdriver
    Written by Marcin Juszkiewicz on
  7. Wedding organizing continues

    14th August 2007 will be important day for me — I will marry Ania at 16:30 in Szczecin, Poland.

    It took many days (usually weekends) to organize whole stuff — find restaurant/hotel for wedding party, choose music band, photographer, camera man… Invitations was another thing to select from many different templates but we found nice one and it was even not so expensive (included personalization).

    Still need to buy suit and wedding rings and probably lot of other stuff but I feel that it is nearly done as most of things to do can be done in Poznań.

    And I still have to finish webpage about our wedding (Polish only)…

    Written by Marcin Juszkiewicz on
  8. Extracting diffs from vendor kernels

    From time to time there is situation when you need to use kernel sent to you by vendor which you support (or use device which they made). The problem is that often all you can get is one big tarball with “our current code” which means Linux upstream + extra patch(es). How to get something usable from it?

    First step is checking version of vendor kernel (if unknown). If you are lucky then main Makefile contain informations such as “2.6.19-rc3” which is a sign what to get as clean base.

    Then clean sources — simple make mrproper will remove all results of compilation so diff will have less work to do.

    Third phase is taking first diff:

    diff -Naurpw linux-original/ linux-vendor/ >vendor.diff
    

    Result will be probably very big but often can be made slower by looking inside of it. Check for backup files, extra log files etc crap — here I usually use Midnight Commander due to it’s “patchfs” which allow to go inside of patch and work on “per-file” changes.

    Next step can contain searching for popular kernel patchsets which can be in vendor kernel. This is hardest part as there are many of them and often only parts of them are applied.

    Same rules apply to other code from vendors — toolchains, libraries and other forks.

    Written by Marcin Juszkiewicz on
Page 80 / 105