Monday, August 4, 2008

Debian Testing... only with LXDE...

I feel like wiping my Suse partition... KDE4.1 is better than  thought, but still too "heavy" for my taste... 

I feel like (once again) performing a minimal Debian Testing install (32bit of course, despite my amd64 cpu...).

-So what would be the difference ??? I already wrote about that some time ago...

This time, I read here: http://debian-news.net/modules/news/article.php?storyid=4526 that "LXDE has entered Lenny, thanks to Andrew Lee. It is a quite
lightweight desktop environment that is made with the Eee PC in mind."

I don't own an EeePC, but I feel like experimenting with LXDE, and comparing it to a barebone Fluxbox install...

Also from the same blog, a goode news for me and my TVcard: "The
debian-kernel team was prompt to add atl1e to 2.6.26 which has just
been uploaded to sid. We hope 2.6.26 makes it into Lenny."

I have a dumb "Real Angel 330", which according to this site: http://www.linuxtv.org/v4lwiki/index.php/Powercolor_Real_Angel_330 "This board is supported on kernel 2.6.26 and above."

Let me tell you that I am resisting the urge to go for a Sid or Sidux experience, as I just don't feel at ease with Sid... I am more of a "Stable" kind of guy, willing to go to "Testing", when it is well-tested...

Gotta leave you guys, power just went off... will write about it later !

Saturday, August 2, 2008

The Kill A Watt device... usefull... see it here:


Saving power while using your PC...

As some of you might know, I live in a touristic beach-town on the north-east shore of the Dominican Republic. The power company (Luz&Fuerza) is privately run, this is why we are escaping those long black-outs that are common elsewhere in the island.

Of course, there's a price tag to that. Our Kwh is very expensive. It costs us roughly 0.50 US$ per Kwh !

-So what do I do ?

I am one of those guys that enjoys having my PC on all the time. I turn it off at night, because it would be too expensive. I use the "power saving" feature that shuts down the monitor (lcd) after 6 mn of not touching the mouse.

Right now, I have a quad-boot: Ubuntu 8.04(with Gnome/OpenBox mainly), OpenSUSE 11.0 (with KDE4.1), Arch (with Fluxbox only) and Windows XP.

I bought a Kill-a-watt device (see it here: http://www.p3international.com/products/special/P4400/P4400-CE.html ). It is one of those things that you plug in the wall, then plugin your item, and it measures the amount of Watts, amps, and other things that your item sucks...

I was wondering if using a distro vs another one would make any difference... Or possibly a DE vs another one, or even a WM vs another one...

My experience is by no means anything close to scientific, but I now gets precise numbers thanks to the Kill A Watt device.

As you FLOSS people would have guessed, Windows XP is the one that sucks more juice, around 166 Watts.

Disclaimer: here's what's plugged in: a homemade desktop PC (MSI K9N Neo V3 motherboard, AMD Sempron cpu, 1 gig of ddr2 ram, a big 320 gigs sata hdd, a Dell LCD monitor 18" UltraSharp, an Omega 600 UPS, a Speedtouch adsl modem, a Buffalo wireless router, an 8 ports switch, and Labtec speakers). When I measure using the Kill A Watt, I have Pidgin, Skype, Firefox, and a web-radio playing.

So here is it: not only Windows XP (sp3 with Classic looks) is way slower than any Linux distros on that hardware, but it sucks 166 Watts.

Here comes a big fat juice sucking distro as well : OpenSUSE 11.0 with KDE4.1 ! It uses about the same amount that WinXP does, sometimes a bit more... I have seen the meter going to 168 Watts... Maybe the reason is that I had some desktop effects enabled, but the very minimum ones, not the cube or any woobling windows...

Our "middle-ground" contender would be Ubuntu with the Gnome/OpenBox interface... The meter was giving me around 162 or 164 Watts... I would not see it as a big difference...

Then here's our champion of the day: Arch Linux with Fluxbox only. I am using it right now, and the meter shows only 152 Watts. I would say thios is a significant difference.

-Why is that ?

It makes me wonder... Would it be because I performed one of those famous minimal install, with only relatively lightweight apps, such as PCManFM, Fluxbox, Aterm, ePDFview, Exaile, Firefox & Opera, Pidgin & Skype... ???

-Would a "netinst" of another distro performs just the same with the same choice of apps ?

Too bad I wiped my Debian Testing partition !!!

But for me, a difference of 15 Watts is important enough to take into consideration... 

That's it for mow, but this subject is a fundamental one for me, so feel free to post your insight on the "save a watt" distro...