D’après le cabinet Gartner, le marché des objets connectés devraient représenter d’ici 2020 en valeur 1 900 milliards de dollars et près de 30 milliards d’unités (contre 2,5 milliards d’unités en 2009). Une explosion du marché qui s’explique notamment par le développement des technologies embarquées au sein de tout ce qui nous entoure…

Tout cela semble bien merveilleux, sauf que les journalistes ont la mémoire un peu courte. L’un des premiers objets connectés vendu fut bien le Nabaztag, mais les journalistes oublient que ça vie fut relativement chaotique, et sa mort lente après seulement 5 années. En effet comme une majorité des objets connectés développés actuellement, le lapin de Violet fonctionnait exclusivement via un accès distant aux serveurs de l’entreprise. Le lapin n’était en fait qu’une coque en plastique, 2 servos moteurs, un haut-parleur et qques maigres circuits électroniques. Le totalité des traitements (vocalisation, API de contrôle …) était opéré à distance. Le jour ou l’entreprise Violet a rencontré des problèmes commerciaux et quand la direction a décidé d’abandonner le produit, le mignon petit lapin avec des oreilles motorisées c’est transformé en vulgaire bout de plastique sans aucune utilité.

Les temps ont changé, la technologie a évolué, mais même si maintenant un service hébergé s’appelle “le claoud”, les enjeux sont les mêmes. Les objets connectés sont toujours des simples micro-contrôleurs bardés de capteurs, et dont l’intelligence est ailleurs. Il y a pleins de raisons à cela :

• Coût de l’objet lui-même : Pas besoin d’un système très évolué. Même un geek un peu débutant peut facilement développer un objet connecté, à faible cout.
• Facilité de développement et d’évolution : Pas de mise à jour à faire chez le client, SAV plus simple.. Il suffit d’apporter de nouveaux services sur les serveurs de l’entreprise. Et surtout un développeur d’API web coute beaucoup moins cher qu’un spécialiste des systèmes embarqués.
• La valeur de la donnée : Bien entendu avoir accès aux habitudes des clients (température de l’habitat, condition météo extérieure, fréquence des joggings..) sont autant d’informations monnayables. Si c’est gratuit, vous êtes le produit, comme dit l’autre.

Je dois avouer que si demain, on me demandait de designer un objet connecté, je ferais exactement la même chose que Violet en son temps, que Google avec le Nest, ou encore que Netatmo. Sauf que cela ne peut fonctionner qu’un temps, il est vrai que pour le Nabaztag cela n’a pas vraiment eu d’impact : Quelques milliers de clients déçus, mais le lapin n’était qu’un gadget pour geek attardé ou pour enfant. (Ok, j’avoue j’en ai un qui prend la poussière sur une étagère). Mais je suis pas sûr que les clients de Nest voient la chose d’un même oeil, le jour ou Google décidera d’arrêter le support, et que le chauffage de l’habitat cessera de fonctionner. Les plus croyants de mes lecteurs diront que cela n’arrivera pas, sauf qu’ils oublient une chose importante. La durée de vie d’un thermostat d’ambiance doit être de l’ordre d’une quinzaine d’année (on le change en même temps que la chaudière ou que le radiateur électrique..). Un cycle dans le domaine des nouvelles technologie c’est quoi 3 ans / 5 ans .. Dans 5 ans, que va-t-il se passer, est-ce que l’ensemble de ces objets connectés auront finis à la poubelle  ? Certains survivront sans aucun doute mais quid des milliers de clients déçus ?

Alors même si certains des produits vendus (développés) actuellement sont vraiment bien conçus, bien produits, il est important de regarder en détails comment le service est réellement opéré. Avez-vous accès aux sources, avez-vous une API locale ? Le système fonctionne-t-il sans connexion internet (en bluetooth par ex). Ou alors, êtes vous prêt à l’acheter en sachant que vous payer non pas pour un objet connecté, mais pour une dépendance assumée.

So the configuration is the same:

• Nividia 173.14.12 kernel 2.6.24 and a Q6600

First, I need to say that in the default setting the new driver doesn’t work really nicely. It’s look even slower than previous in the default configuration. So for the first time on this bench serie, I tweaked the InitialPixmapPlacement and set it to 2. In my previous bench batch, doing this tweak products bad result so I disabled this option, but this time the drivers is so slow that without this tweak the benchmark would be useless.

Ok, let’s go for the results:

First, we can see clearly the new version is really better on some points : PictOpClear is the best result. We can see the nvidia team has really work on this, and the result even outperform the ATI driver. On the other side the PicOpt[Con|Dis]jointClear is still very hight.

For the rest of the test :

To things, on quite all the result the new driver is slower than the previous on (perhaps this is a InitialPixmap side effect), but the difference isn’t really big 0.5 sec on a test which is far from 0.5 at the end.. And the ATI still outperforms clearly the Nvidia here. In fact Nvidia driver’s team claim this primitives are never used (or should be). From what I know right now, some software use this primitive. It’s look like KDE (via QT) do. Apparently Nvidia team asked the KDE dev to change their code to achieve better result on Nvidia cards … Anyways this is perhaps not the best way, but we need to wait for KDE dev answer before going foward.

The second important thing is that PictOpConjointXor has now a 0 result.

As you can see on this benchmark, the new Nvidia driver seems to perform better than the previous one. On the user perspective, it’s look like the fixes applied for PictOpClear (and perhaps PictOpConjointXor) produce some great results. Right now Firefox perform nicely, and the whole desktop is fine. I’m quite sure their is still room for improvements (look at the open source Intel driver results for PictOpOver PictOpIn…, you will see binary drivers are far from the OSS results), but this release is the first for the 8xxxx serie which perform at a decent speed, and this is a good new.

Thanks again to my friends who send me their own results to compare, and to people on various forum that helped me on this stuff.

After a couple of month, I decided it was enough, I was sick of this dirty stripes on screen. I tested every ATI driver one after one … (ATI opensource drivers have too bad performances to be used on a every days desktop, could you live without google-earth ? ) .. so I decided to go to other side, and bought a Nvidia 8600GT from ASUS. This card perform quite as the ATI in 3D, and have a affordable price. So I switched from ATI to Nvidia.

ATI offers better opensource support, but Nvidia binary driver is really nice to use and have better support today from stuffs like Compiz and Co.. and NO MORE STRIPES !! :)

A couple of weeks ago, I upgraded my Ubuntu Gusty to Hardy. Everything was Ok, since I played with firefox. Some heavy loaded pages (like Amazon, or Gmail) was damn sloowwwww ! Playing with scroll was a source …… grrrrrr ….Firefox on Hardy is 3.0b5. This version has a major “feature” the use a Xrender for the page display. And this looks like Xrender is dawn slow on Nvidia cards .. In fact, Nvidia has already work on this kind of issue before. Without looking forward I decided to run a little benchmark, with the help of friends with Xrenderbenchmark. So here the results.

Benchmarks

Benchmarks was done by me and 2 firends, on Q6600 or E6600 Intel CPU running at 2.4Gh, with kernel 2.6.24.1. The graphs only show the Plain results (not Plains + Alpha, or Transformation) but the results are quite the same anyways.

Legend:

• 8600GT/nv : Nividia 8600GT / Xorg 1.4.1git 32 bits / Nvidia GPL driver
• 8600GT/nvidia : Nividia 8600GT / Xorg 1.4.1git 32 bits / Nvidia binary driver ver: 169.12
• 8600GT/nividia-64 : Nvidia 8600GT / Xorg 1.4.1git 64 bits / Nvidia binary driver ver: 169.12
• Intel GMA X300 : Intel GMA 3000 / Xorg 1.4.0.9 64 bits / Intel GPL driver
• ATI HD2600PRO : ATI HD 2600 Pro / Xorg 1.4.1git 64 bits / ATI GPL driver

I split the results into two graphs for convenience.

As you can see on this first part, numbers are really small, the Nvidia GPL driver is the worst : 5 times slower than any other one. Not a good news, and the binary one have some bad results on 2 tests. ATI HD and Nvidia drivers offer quite the same results, but remenber this is the GPL ATI driver ! … The Intel doesn’t have a lot of linearity on this part.

But the next graph give us absolutely different picture !

For every graph, Nvidia drivers (GPL, or binary, 32 or 64 bits) are at least 6 time slower. Intel perform very well, no surprise, this card are damn cool, perfect driver for linux.. but to slow in 3D to really rock. And ATI GPL driver is the clear winner of this benchmarks tests.

As my issue is the Nvidia one, I can comment the results, the GPL driver performs better than the binary one. This is not a big surprise cause, I can see it in Firefox, even it’s slow. There is difference between 64 bits and 32, but I guess this is more kernel related than the driver itself.

I’m not a video guru and only do that figure out what’s going on my computer. I publish in the hope it might help somebody else, and to find help.

Update : The numbers can be found here

Thanks to Ludo and Christian for their help !

Important update : Check the new driver results !!

My new box description:

• CPU: Intel Q6600 Quad Core at 2.4Ghz (G0 => it’s running at 1.6Ghz on idle)
• Motherboard: Gigabyte GA-P35C-DS3R (Intel P35 Express)
• A massive big CPU cooler Arctic Cooling Freezer 7 Pro
• Case: Cooler Master Centurion 534
• Hard-drive: Seagate Barracuda 7200.10 SATA – 320 Go 7200 RPM 16 Mo SATA II
• 2Go of generic DDR2-SDRAM DDR800 PC6400
• Power Supply: Seasonic S12II-430
• A passive Radeon HD2600
• A nice (and big) Samsung Syncmaster 223BW
• A Asus DVD player: DRW-1814BLT

Here a little photo:

So let’s take a look closer.. I bought the Q6600 because it has a good price right now.

For the motherboard I wanted to buy a serious brand (I had some bad issues with ASRock ..). And Asus build some good ones, but doesn’t have a PS2 keyboard plug. I choose the Gigabytes because the bridge cooler is on the right, so I can use a big fan in the case back (no cpu-fan) without the bridge fan issue.

For the harddrive, I already have this type of Seagate on another box (in ATA, not SATA), and it’s damn cool ! Low noise and great performances .. a substain 58Mo/sec (a little bit more than my previous LVDS-SCSI config at 10 000 rpm !)

The perfect box should have two 120mn fans. So this CoolMaster is fine.
I only use the front fan right now, but this may change in future. The case is quite big (too much for my needs), easy to mount, and serious quality. The hard-drive combo stuff, isn’t really easy to use but it does it’s job.

For the power-supply, after reading a lot of website, I bought the Seasonic. This stuff cost a little money, but it’s a really silent one, and seems to be robust.

As I bought a silent power supply, I needed a silent cpu cooler two. Everybody knows that default Intel fan are awfully noisy, so I decided to test this Freezer. And that was a good choice. On idle, the fan runs at 450 rpm, (hey that’s low!!) and CPU is at 31°. At full load, the fan runs at 1000 rpm (still really low!!) and the CPU is at 46°. My initial plan was to remove the fan on the cooler and to place a big 120 mn on the case back, but haven’t done this yet.

For the video card, I was looking for something with good bench on Linux, so ATI was a good challenger (beside it only works with the latest ATI binary shit). And finding a passive HD2600 is a bit hard.

For the DVD, I didn’t have any ideas. Most DVD players I used are noisy, have unstable speed. My previous Samsung was a real bullshit. I went to me local computer store, and bought the cheaper they have on stock. And this is a really good surprise. This DVD burner is really cool. So, if someone ask, give it a try :)

So, as you can imagine, this little stuff is quite silent and powerfull :)

I will upload some pictures soon …

Update: I took some photos

Enjoy cores

The phone itself is pretty nice, use Linux (yes !! kernel 2.6.10 inside ) and can be hacked: http://opentwin.org/

You can use it to surf quite well. Opera does a really good job, and you can even watch TV for free.. This phone is a good piece of stuff for the price.

Hum, I ‘m guessing you will see some new post about Asterisk or other fun stuff here soon ;)

/Enjoy VoIP

The only issue: I need to take it back to my computer when I want to upload a new film, that’s it .. When I first received this gift, I went back to the shop to exchange it against a networked one. But 1) they cost a lot more money, 2) not available at the shop. (even if it’s a really big one).

Last week, somebody sent me a mail. He found some cheap wireless access point with a USB port: Netgear WGT634U. (50Euro) Hum, ok let’s give it a try. First test, plug a Media Player to this USB port… ok It’s working fine. But you know, I don’t have a network cable near my bed, so I decided to transform the Access Point in a Wireless client. By this way, I can simply plug the Media Player on the WGT634U and remotely put some Divx via FTP, without any wire.

# hit Ctrl-C on the bootloader
CFE> ifconfig eth0 -auto
CFE> flash -noheader tftp_host:openwrt-wgt634u.bin flash0.os
CFE> reboot

config wifi-iface
option device   wifi0
option network  lan
option mode     sta
option ssid     MyDummySSID
option hidden   0
option encryption none

#### LAN configuration
config interface lan
option type     bridge
option ifname   "eth0.0"
option proto    'static'
option ipaddr   '192.168.3.2'
option netmask  '255.255.255.0'
option gateway  '192.168.3.1'
option dns      '192.168.1.254'

#!/bin/sh

# clear all firewall rules
for T in filter nat mangle ; do
iptables -t $T -F
iptables -t $T -X
done

# drop incomming packet
iptables -P FORWARD ACCEPT
iptables -P OUTPUT  ACCEPT
iptables -P INPUT   DROP

# accept traffic on localhost
iptables -A INPUT -p tcp -i lo -j ACCEPT
iptables -A INPUT -p udp -i lo -j ACCEPT

# accept ftp only from my home network
iptables -A INPUT -s 192.168.1.0/24 -p tcp -i br-lan --dport 21 -j ACCEPT

# accept incoming http / ssh
iptables -A INPUT -p tcp --dport 80 -j ACCEPT
iptables -A INPUT -p tcp --dport 22 -j ACCEPT

# finaly accept already open Cnx
iptables -A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT

#!/bin/sh

while true
do
{
STATE=$(iwconfig ath0 |grep  00:13:13:53:DA:D1 | wc -l)
[ $STATE -eq 1 ] && echo 0 > /proc/diag/led/power
[ $STATE -eq 0 ] && echo 1 > /proc/diag/led/power
sleep 20
}
done

Step 6 / Finally

Here the result:

The finished product (grrr) works really fine and managed to achieve something like 850Ko/sec, I mean full speed ;) (11Mbs network) I need to remove the serial port right now, and close the box..

Important update

I read on Engadget that I use this to stream video to my TV, (like the Apple TV do), but this is absolutely wrong. This media player has a 320Go hard-drive. Why would stream film on this ? I just put it on that’s it :)

This is really amazing how people doesn’t read the article to see what I’ve done. It’s not the first time this happen in fact.

Here in France, most advanded users have some network TV from a long time, and this little boxes can play network stream (like the Apple TV) for a long time now. Beside you don’t pay for this service (rent for the boxes are included in DSL bill), the only issue is that you must have another computer to stream.

I guess nobody here (in France) would by a Apple TV, since we already have this kind of products for free for a long time. Look at a Freebox for example.

It’s time to watch a film :)

I have been looking at eBay for a while, I bought a pair of matched CPUs. Today, I received them. After a bios upgrade, I put the CPU “as is” on the motherboard. In fact, my motherboard haven’t the x 6 jumper settings.. but despite the jumpers settings are totally wrong, the CPUs work at 600Mhz.

Yes, I get 1200 BogoMips for 42 Euros :)

Thanks eBay ! and this faq !

I decided to change it by myself. But finding the right fan is really hard, because it’s 80mn x 80mn x 1.5mn fan. After a long search i decided to go to a tunning shop. (They have a bunch of various CPU fan).. The guy said to me: Ok, No problem. We have a video card fan that can feet .. Oh really a video fan ?

He give me a Zalman ZM-OP1

Despite this fan doesn’t have the exact same caracteristics as the default one, it perform quite the same. And it does even less noise. So i’m really happy :)

Many thanks to the shop: Tuning Informatique (http://www.tuninginformatique.com/)

update: fix the Zalman URL

Here on my shuttle .. not working

aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: IXP [ATI IXP], device 0: ATI IXP AC97 [ATI IXP AC97]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 0: IXP [ATI IXP], device 1: ATI IXP IEC958 [ATI IXP IEC958 (AC97)]
Subdevices: 1/1
Subdevice #0: subdevice #0

Here on another box (VIA C3 based)

aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: V8235 [VIA 8235], device 0: VIA 8235 [VIA 8235]
Subdevices: 4/4
Subdevice #0: subdevice #0
Subdevice #1: subdevice #1
Subdevice #2: subdevice #2
Subdevice #3: subdevice #3
card 0: V8235 [VIA 8235], device 1: VIA 8235 [VIA 8235]
Subdevices: 1/1
Subdevice #0: subdevice #0

So yes the C3 can play 4 stream at the same time, while the AC97 doesn’t.

Another way to do this is to use software mixing

Update:

To use the software mix in Alsa: simply change the pcm.default to pcm.dmix

#pcm.default cards.pcm.default
pcm.default pcm.dmix

I have tested several way to fix this issue. The only that’s working fine right now, is to use swsup. So to make it work:

• install the latest 2.6.9 kernel (the 2.6.8 has a issue w/ agpart and drm)
• in kernel select ‘ACPI -> sleep state’
• and ‘ACPI -> swsup’
• build the kernel
• create a swap partition if you don’t have one (i resized a ntfs w/ qparted)
• mkswap /dev/hda4 (for example)
• add ‘resume=/dev/hda4′ in your boot prompt (in grub for example)

Now your system is ready. Reboot :) ..To put the the system in standby and dump the current state in the swap: echo 4 > /proc/acpi/sleep.

Next step simply bind the power button on this action. This is done w/ the acpid.

#!/bin/sh
# /etc/acpi/powerbtn.sh

# put the box in swsup mode
echo 4 > /proc/acpi/sleep

Beside the S2/S3 ACPI doesn’t work right now, swsup permit to have suspend to disk without too much trouble.