Computing

CyanogenMod does list Samsung Galaxy S in its list of devices, but given that it is marked experimental I was not very confident of trying it out. But yesterday, I took the plunge. Even though experimental, the ROM is pretty good. And the Cyanogen goodness just makes it awesome. This ROM (also being AOSP GingerBread) is Super Fast.

So in a day of use, my impressions:

• Battery performance has turned out to be better than what I was used to.
• Cyanogen’s enhancements are pretty impressive.
• The CyanogenGalaxyS team’s work is of very good quality, even though they’ve marked it experimental. (Thanks codeworkx, coolya and the team)

The CyanogenMod experience is also good because it makes me feel like home, just like our Debian project.

While upgrading the memory (RAM) of my notebook I did a lot of study as to what exact kind of RAM would my notebook use without problems. By the end, while I finalized on what RAM I need for my notebook, I had compiled a small document on Basic RAM Information courtesy Crucial® FAQs.

I hope this guide to be helpful to you but with no warranty though. :-)

Do I have to buy the same size upgrade as the memory module currently

installed in my computer or can I mix different sizes?

The latest kernel upload (2.6.32-16) brings goodies to SystemTap in Debian. This version has added support for kprobes, on which systemtap has a major dependency, for many of its features.

Most of the systemtap instrumentation should work now and all of this will be part of the Squeeze release. Instrumenting the kernel modules still needs some work (DBTS: #555549) but can be done.

It all started long back when I worked for a giant computer manufacturing company. Certain IT policies led fo the need for an Offline APT Package Manager

While I got it working for a long time, I didn’t have the aggression to polish and push it for general usage. Thanks to my friend appaji, [apt-offline](http ://apt-offline.alioth.debian.org) (a.k.a pypt-offilne) is now part of Debian.

I’ve been using KDE for a while now, probably 9yrs. I’ve also been an early adopter of KDE 4.x. While KDE 4 is still far way behind in proving its worth of the radical core changes it made (take for example: Nepomuk, Strigi, Phonon, Decibel - I still wonder when they are going to be ready for the user), I still find KDE apps far far ahead of GNOME.

Probably, many would disagree. Possibly, they might flame me too.I mean everyone supports GNOME as the default - Red Hat/Novell/Ubuntu. And I always wonder WHY.

Once upon a time CPU power was low. In those days, what we have today, was termed to be Super Computers.

Thanks to tough competition and great engineering, we now have CPU in the range of Gigahertz and multiple cores. But on Destkops/Laptops, do we really have applications that utilize the ability of these processors ?

So I thought about making these powerful CPUs to do some work.

There are different views about Backups. My preference has always been that my backup should be an identical image of my entire OS. That’d include the cache, the packages installed, my personal data, my mail spool et cetera. Given the requirements, my preferred choice of Backups has been LVM Snapshots.

Backups

Most of the users using computers have a very high dependency on it. Day-by- Day, our data is getting digitized. Everything is getting into electronic formats ( Movies/Pictures/Music et cetera). If you are one, you know how important it is to have a backup.

Lately, I haven’t been using Microsoft Windows on a daily basis. So I’ll comment on Linux here.

The definition of Backup can be different. People like backing-up only the Important data. The problem is that the term Important is very volatile. What is important to X is not necessarily important to Y.

[Read More]

IMO, Linux Distributors should seriously think about Role Based Operations for their distributions.

Every linux user is going to be using his copy of the distribution in a different way. But from the distributor’s point of view, it should boil down to:

Desktop Users

Laptop Users

Server Users

Currently, some of the distributions apply this thought only to packages. For example, Ubuntu rolls out a desktop flavor and a server (generic) flavor of its kernel package. I think the distributions need to think broad and realize that NOT One Size Fits All.

I’ve been working on a bug recently related to udev which under certain circumstances triggers an OOM on the Linux kernel. On another bug, I also wanted to look for the max open processes for a single file. The results I found and pretty good. The Linux Kernel’s OOM feature has imporved a lot.

This is the one liner used to trigger the OOM.

while true; do tail -f /tmp/test.txt & x=expr $x + 1; echo $x; done

[Read More]

Many of us might be aware of the block level encryption features (LUKS for example) provided in Linux.

Most of the articles that I referred to over the internet mention the usage of the traditional scsi block device when creating it as an ecrypted device. Eg. sudo cryptsetup luksFormat /dev/sdb1

This actually poses a problem. The problem is with the way Linux detects SCSI devices. While many distributions have shifted to using ID/LABEL based access to devices, users still use the old traditional way of using a device. This turns out to be a problem in cases like suspend/resume where devices are rediscovered. Technically there is no guarantee that the device you see currently as sda will be named sda again on reboot/resume. While we don’t see issues like this with attached disks, the issue can be easily reproduced in a SAN Enviornment or in cases where users use external hard drives using USB.