Backup solutions for a Linux PC

Which is the “best” backup solution for my laptop? That’s what occupied my mind until very late last night. It seems like a very classical question, as old as the homo numericus. So you would assume that the answer as been disussed over and over and is now solved for good.
Que nenni.
Depending on the nature of the data, of the quality of service you expect, on the harware you have or you will buy, on the volume of your data, on your network, on the security policy you wish to comply with, on the interface you feel comfortable with, and so on…, well you might end up making a very personal decision.
And that’s what I did. Let me tell you how and why.  Once upon a time, an ordinary Linux user tried to use…

… The Backuppc server
Backuppc is a server side backup script with a web frontend. The script periodically searches for pcs to backup on the network and when it finds one, it connects to it (via ssh for example) and runs rsync (but other options are available) from the pc to backup with the server as a destination.
First the drawbacks: you have to set up an account on your pc for backuppc to connect to it, you have to make it secure by using encryption keys. Then rsync execution over a wifi network takes forever. Plus, I am not really confident in rsync copying ALL my files properly. You don’t really know how the files are stored on the server, so if you want to recover, you have to trust backuppc, and I don’t like it. When it come to data backups, I want to know exactly what’s what.
Advantages: If you test it properly, you can easily backup dozens of pcs. I only have one.

I trashed my backuppc setup lastly.

… The Cobian backup software (Windows only)
Cobian backup is exactly what I want on linux. It is clear: the software is on the pc, it copies and possibly compresses files from one point to another, which can be a remote server.
It is perfect to backup a single pc or various pcs assuming pc users are responsible for their backups.

Problem is I am a Linux person.

… The short Keep (kde) experience
Keep is like Cobian, but for kde and not as fancy: it can’t even connect to a remote server.
Plus I often switch between kde and gnome when I upgrade my ubuntu system.

So Keep is not good either for me.

What now? A good old bash script!!

Yes. That’s the solution. Of course I would prefer a good software like Cobian that can run everywhere (such as keepass). But it seems one has to live without it.
Anyway a script is portable, standard. So here’s what I’ll do: a bunch of scripts using tar, gzip, cron, time, sftp running on my pc. A sort of home made cobian.

Yapuka. Aka let’s do it.

Posted in backup, home server, software | Leave a comment

GPS logging newbie

I bought this electronic gadget on ebay to grow my collection of “toys”, it is a gps data logger. I still didn’t have time to tinker with it a lot, but I am really excited to play with this new toy.

gps_iblue747_s.jpg

Applications are thrilling: I’ll be able to geotag my photos, post them on the net, I’ll be able to share my trips, places I went. I’ll be able to play back my track in google earth.
I can think of countless applications:

  • analyse your driving profile regarding safety
  • analyse and compare your ecological impact given the frequency, the length of your trips. You can mix that up with your means of transport.
  • you could come up with interesting numbers like the total distance you made by feet, by car, by plane. The energy you burnt while doing sport.

This is my skiing track on March 17th this year.

Val D’Isere track

And if you zoom, you can even see the turns:

valdiseretrack2.jpg

On the technical side, being a Linux user, I am really pleased to see that 2 wonderful programs will be real helpful:

  • mtkabel: extract data from and configure the data logger.
  • gpsbabel: convert, filter, manipulate different gps file formats. Converts from gpx to kml for instance.

Yum yum yum…

Posted in gps | Leave a comment

Founders at Work: great book

foundersatworkI am currently reading an excellent book about entrepreneurs in the hardware and software industry. The book is called founders at work, it is a collection of interviews of the people who built and are still building what we now take for granted: computers, emails, internet. This book in fascinating because you get to see behind the curtain, the founders share their personal experience and give insights. They are stories of craftsmanship, of hard work, of faith, of passion. What is striking among all these stories is the faith these people have in what they are doing, and above all in what they can accomplish. They do not fear the future, they just throw their lives in a project because they believe in it.

This book gives humanity to the computer world, it reminds us that it is above all a story of men and women trying to produce something useful and valuable. You can really feel the violence of the feelings they had to go through: exhaustion of long waking hours, excitement of success, pride of going public, doubt.

If you ever tried to invent things, if you have ever thought of starting something on your own, if you are interested in innovation and business at its essence, you must read this book.

Posted in book, software | Leave a comment

This site is openID enabled

Open ID is THE authentication protocol on the rise. Software is moving to the cloud, zillions of web apps out there can do everything from photo editing to money management, everybody is on facebook, you can do spreadsheets in google. All you could do on your desktop is now on the web, even storage. The major drawback is the number of passwords you have to remember (around 30 for me). With OpenID, all you need to log in an openID enabled service is your url (mine is http://www.thibautwdowiak.com). The website wanting to check your ID no longer have to keep your credentials, they ask an OpenID provider (to which you register) if the url you entered is really yours. This way, you only have to authenticate to your OpenID provider. Isn’t this cool stuff!

There is even more to this: you can leave comments using your openID, and if your openID is also the url of your blog, you might get traffic from the places where you left comments.

This blog is openID enabled!!

Posted in openID, software | Leave a comment

Cv editing is really a pain

In this previous post, I exposed a set of scripts to produce resumes in html and pdf. I have showed the pdf output around and gathered some comments and advice. Turned out that the structure of the cv could be tuned to reflect my experience more accurately. It opened my eyes on the drawbacks of the dtd – xml approach for cvs: While it is really practical to change the contents, the structure is not flexible at all. Or if you want to make it flexible, it has to be more complicated.
Or you have to have tools to make it easy.

See, it is the exact opposite of word documents:

word xml dtd
no logical structure at all logically structured
hard to maintain and deploy easy to maintain and deploy (html, pdf, french, english all in sync)
easy to change the visual aspect hard to change visual aspect.

For example: I had this structure:

Work experience / Experience / Company, Other element.

That supposed I had to specify the company for each experience.
What if I want to describe 2 experiences in the same company. I have to make it:

Work experience / Company / Experience.

And adjust the html and pdf outputs accordingly.

Anyways, this version of cvmaker tries to be more generic to handle the case above. It could tidy it up and make it smarter but arf… maybe the next version. You can download it and check out the result here.

Posted in CVs, development | Leave a comment

Where do I go?

Just random thoughts coming across my mind as I was daydreaming about various software solutions…
One preoccupation I had was: how to make something useful, purposeful and valuable? Well, I realized it was a tough question, To escape the resulting frustration, I started to wander among the nearby thoughts.

I believe the creeping feeling of lack of purpose is a major characteristic of our society: there is not any more just a few paradigms to help us organize the world mentally. Our world is a globalized network (network of people, of financial connections, of ideas, of computers) in which it is difficult for the individual to find its way, to have a vision of where the world is going. Our actions and lives depend less on our close environment and more on remote connexions. Physical proximity, hierarchy, inclusion in a group (may it be family, company, religion) used to provide a meaningful structure for societies. It is less and less the case. We cannot understand our world with the sole help of our good old paradigms. The gap between those who try to adapt, try to understand (which requires education, willingness and efforts) and the others who want to live and think by the same old standards is bound to widen.

As a computerized, connected (whether you want it or not, whether you are aware of it or not) agent who produces value, what should I do to feel I am alive in my environment, to feel that what I do matters.
I think I have to clarify my statement at this stage. I do not mean you cannot be happy in life nowadays because of networks and complexity: there will always be simple things in life that make people feel happy and fulfilled. Here, I am interested in those forces which drive societies, companies, industry, and which have repercussions on how we understand our environment and how we perceive the part we play in the game.

I mean, people need to understand why their bank just failed to go bankrupt for 5 days and is fine 3 days after, people need to know why they are paid so little and still have their factory shut down for costs control reasons, they need to know why some agents (may it be companies, persons, products, services) are incredibly successful and why others are not. They need to know where the value comes from, where it goes and above all, how they capture (or not) some of it. This is not the answer to life, the universe and everything but it doesn’t hurt to wonder.

Posted in opinion | Leave a comment

screen

How I take advantage of a low perf proc to perform tasks in background.
This software is just awesome: simple, straightforward, practical. This is how I use it:

I have a small box (kurobox) with little power, little memory, little everything, except the hard drive, which is BIG. I often connect to it from the internet to perform various tasks and sometimes these tasks take several hours of processor usage. Screen enables me to disconnect, shut down my laptop and reconnect later. I can compile, copy batches of files etc… without having the client (my laptop for instance) on all the time through. Simply with this:

apt-get install screen #install
screen #start a screen
#do stuff
Ctrl+a d #disconnect from screen
#disconnect from the server
#reconnect to the server
screen -r #reconnect to the screen
#do stuff again

Have it a try!

Posted in home server, kurobox, software | Leave a comment

SNCF vending machines (try to) run windows

SNCF Vending machines Windows
Location: Poissy (Paris Area)
Date: 2008-01-23

That’s why these train tickets vending machines are soooo slow when they are not down !!! Now I understand…

Posted in opinion, photos | Leave a comment

Wintersports in Flaine (Alps)

Do you know Flaine? It is skiing resort in the french Alps, built in the 60′s, in the aftermath of WWII, when concrete was fashionable… This is a picture from the slopes on a foggy day.

Flaine from the slopes
Location: Flaine (French Alps)
Date: 2008-01-05

Posted in photos | Leave a comment

vlc 0.8.6x on ppc 603e as of December, 11th 207

Plot:
I am running a machine with a powerpc processor MPC8241 derived from the ppc 603e and managed to get vlc 0.8.6c running as a daemon on it. This way, I am able to stream my mp3s off my little machine to a multimedia box provided by my ISP as part of the service. This box has RCA output.

Challenges:
It took me months to discover the right set of libraries and patches to include in my installation and the options to compile them. So here we go for a little how-to.

How-to:
I quickly decided to go the hard way, that is by hand compilation rather than the debian apt way (oh, yes I am running debian) given the specifics of my processor and the minimal config I desired. First thing to do is download the libraries and vlc 0.8.6c (maybe it also works with the freshly released 0.8.6d):
Then extract, configure and build them.
I configured them with –disable-altivec, because the mpc8241 does not support the altivec instructions set.
Patch vlc:
This part is the nastiest part, not that it is particularly difficult but it took me ages to find this was THE hack I needed. Let me explain: there was a bug related to altivec support in vlc (link), patched in the 0.9.x branch (trunk) of the vlc development tree. But this patch never made its way to the 0.8.6-bugfix branch, this same branch that currently gives birth to the stable 0.8.6x versions of vlc.
Running a 0.8.6c version of vlc off the box kept me stuck with an “Illegal Instruction” as soon as I tried to play a file.
I had to analyse finely the vlc development tree to discover that the initial patch, courtesy of the gentoo people, had never been included in the vlc bugfix branch. So I applied the gentoo patch to my 0.8.6c source, bootstraped, configured and compiled.
Configure:
I used the following options to compile vlc:

./configure --prefix=/usr/local --disable-altivec --enable-ogg --enable-flac --enable-vorbis --enable-theora --enable-dvb --enable-freetype --enable-dvdread --disable-caca --enable-mad --enable-a52 --disable-wxwidgets --enable-libmpeg2 --disable-speex --disable-skins2 --enable-ffmpeg --disable-debug --with-ffmpeg-tree=../ffmpeg --with-a52-tree=../a52dec-0.7.4

Download:
I am hosting a bash script that summarizes the procedure. You should not execute it blindly, but tweak it to your needs.
zip archive thumbnail

Posted in home server, kurobox, software | Leave a comment