Too many choices aren't always good

Came across a blog on Googles' blog blogroll. This man seems to be really involved in lot of technological stuff, talks about IMac, Apple, Linux, KDE, Gnome, Perl, Python, Ruby and on and on, he is just an amazing geek.

Well the topic basically is about TOO MANY CHOICES. His blog had a post on Perl to Python, Sure. But Ruby? very interest point raised. I totally agree with the fact that too many choices and too much internal competition in Open Source world sometimes keep us away from reaching the best possible output/Softwares.
As a user, I sometimes feel good to have many choices, but at other times it pisses me off, when I see my distro just removing some application from the package and putting another alternative one. The commands which I learnt and softwares which I became very comfortable with cease to come packaged in to the Fedora Core CD all of a sudden. All the time the number of Image Viewers keep fighting internally, whichever gets a bigger userbase/developerbase or possibly endorsement by a distro packagers makes it to the Distro. Same happens with Media players, with Window Managers.
Window managers play an important role in software duplication, KDE provide its own set of apps for Multi protocol chat, browsing, File-management, Image Viewing(kuickshow, showimage, showfoto ...), media player and Email Client(Kmail) while Gnome provides its own set of apps for the same tasks. No wonder users of one WM like application developed for that WM because they interact well with the WM. Wait I forgot to add another set of apps, lets call them Xapps, that application which run on both of them equally well, this include Xpdf, Xchat, Xmms and so on. Now there will be three application for each task, you can use whatever ur distro provide by default, or have to go through the pain of installing ur favourite from net, which might not be very simple all the time.

Wouldn't these application's development be faster had they shared development efforts? Wouldn't they have had a better user-base had they been overall good?

As a developer too its always confusing which platform to choose, If I write an app on KDE platform, I miss the Gnome userbase, and vice versa. How good would it be if I can write an app which works equally well in both the WMs. I don't know what kind of standardization could be done, but something need to be done immediately to make sure we can combine our efforts rather than just divide them.

The latest news in regards to standardization is that, X Server seems to be going in to two possible branches. (Xgl and AIGLX). As much as Sound Servers used to suck due to incompatibility, its probable gonna be Xserver times to do so.
On the other front, Beegle's code is now being cleanly interfaced in to UI and NonUI parts so that even KDE can use it. I did miss it earlier and would really love to use it now.

Fuse and SSH-Mount(SSH FS)

Regular users of Samba protocol suite(Linux/Unixes way to access and be part of Microsoft Network Sharing protocol) might understand the importance of a mounted filesystem. Basically after mounting the filesystem you can access the files on the filesystem as if they were on your own system under the mount point.
This concept is also the basic of NFS mount which is very widely used for sharing a Disk across multiple remote systems. Users just mount some directory of the NFS share on their system and use the files on the remote system as local files.

Sometimes its irritating when you are asked root privelleges to mount file Systems on your system. I was myself puzzled by the super user requirement for mounting files, If you are puzzled too, go ahead read the book "Design of the Unix Operating Sytem" by Maurice J Bach.

But there is a solution to the above requirement, its called FUSE, which is acronym for File System in User Space.
I chose to ignore the implementation details of the thing and just use it, Its pretty interesting and useful.
The important idea is that it doesn't require any root privelleges during the operation of mounting and unmounting. A large number of Filesystems based on FUSE have been developed. Among others the some interesting ones are FlickrFS, SSHFS, GmailFS, and SMBNetFS.(Well there are so many of them, you gotta checkout which one is of use to you)

I tried playing around with SSH FS, its pretty cool. I was able to mount a system over SSH protocol, i.e I can just see the files as if they were on my own system after 'mounting' my login's home directory in to local directory. You can play audio files over a SSH server !!! After installation all that you need to do it. (A.B.C.D is my ip address)

[vardhman@localhost ~]$ mkdir ~/labpc
[vardhman@localhost ~]$ sshfs-fuse-1.4/sshfs A.B.C.D: ~/labpc
vardhman@A.B.C.D's password:
[vardhman@localhost ~]$ cd ~/labpc/

I then tried the smbnetfs filesystem, its wonderful to. Here are the command

[vardhman@localhost ~]$ mkdir ~/winmount
[vardhman@localhost src]$ ./smbnetfs ~/winmount
SMBNetFs-0.3.2
[vardhman@localhost src]$ cd winmount/A.B.C.D
[vardhman@localhost A.B.C.D]$ ls
dump #.. other files shared on machine A.B.C.D

Basically all you need after creating a mount point is to give cd /ip address and wooo you see the shares are if you smbmounted them. I would love to know if we can write a script in konqueror to automate this task then smb://ip address is given in the address bar.

I heard FlickrFS and GmailFS are cool too, I do plan to try them soon.

By the way unmounting all of them is simple just give the command

fusermount -u < mountpoint > </mountpoint >

I am told fuse support is coming inbuild in kernel after 2.6.14, In my kernel the one with FC4 it was not working by default. I had to give the --enable-kernel-module option while compiling, make install in to defualt prefix does require root previlleges and so does the modprobe fuse command do. Also you might need to set the LD_LIBRARY_PATH variable. Installation is pretty straight forward and simple. I am sure it will be turnout to be very useful.

Some Felicity 06 Photos...

Got a Kodak Data Cable finally. So now you can see some pics captured from(in?) my camera during felicity.


Wonderfull Bharatnatyam performance. (Wonderful photography too ? )


(same as above)


(wonderfull balance. isn't it?)


Atanu Dey at Felicity



Aarabhi Group's Violin performance (Frankly, I didn't get the full feal of it 'coz the music was totally new to me, they sang all telugu compositions AFAIK, except for the last Sare Jahan se aacha, which was awesome)


This was plain awesome, A wonderful fusion of all the musical intruments, I even recorded the sound on my cell (any way to host them on net?)



Tried photographing a fireworks, looks cool!


Mr. and Ms. Contestents (No particular reason for this photo ;), just that it came quite good)


Mr. and Ms. Felicity again!

Wud luv to host more photos if you have some good photos to share, let me know. (I missed the dance and some other sessions)

Being a Linux Enthusiasts and not a Linux Zealot

The title might suck but then its not always easy to summarize things in title.

Very often Linux enthusiats are called Linux Zealot. I decided to find out the reason behind why ppl like me and many other get so passionate about simple things like an OS. I have figured out at least one big reason for this: the sense of Ownership. Don't think I am gonna write about the Freedom etc stuff, I don't really philosophize that a lot.

By Ownership I mean willingness to accept responsibility and also to change the situation towards betterment. The most direct example I find in my excessive support of KDE compared to GNOME or any such WM whenever there comes a situation for debate. Though I was always impressed with the GUI the KDE guys brought to Linux Desktop, but the obsession comes more due to a second reason : My excite ment of developing something, albeit a small one, for the KDE. (Though when I started my initial bias towards KDE was a major reason for putting my efforts at becoming a KDE contributor rather than trying something Gnome-ish)

The same situation holds true with most Open Source Software (OSS) users. When you ask them any scopes for improvement they can positively show you many and also accept many of your suggestions, but one thing they don't like is OSS being criticized by those one sided people. The normal users do understand that lots of improvement is possible but they also understand that they can contribute and improve equally well as any other person and the software is their own. They see it their responsibility to make the situation better rather then just whine for it being not the best. Being a part of the sucha big community of like minded is a unique feeling in itself.

There is a important thing that I have learnt by experience, Its no use being a pro supporter for any OS or software during a clash of views with ppl who don't really want to change their opinions. Much better way to spend your precious time would be directly get involved in development of the same software (OS level development may be a bit bigger milestone) and improve it. Ultimately thats the the best advantage of OSS and the experience and learning is worth the time being spent in it. Your contribution to the community would also be much more appreciated.

Well for timepass I do enter in debates, promoting my favourites software to ppl who care to listen. So post like why KDE rules and so on will still keep coming. :)

Talk by Atanu dey at felicity

I can't possibly even summarize the great talk in my own words. Those at IIIT can see the video recording of the same which shud be available soon through Felicity organizers. A small introduction to the speaker is that he is the author of one of the most popular blog in the country http://www.deeshaa.org. Here is my attempt to summarize it in brief. Since it was a very general talk its difficult to emphasize every point.
Surplus of information: Example of how dietrician industry(?) has come up so successful with the availabilit y of surplus of food, by logic their will soon be a need of ppl who can provide ways to summarize the huge amounts of information. How nicely our sense and brain filter the use amount of data available to us and filter and store it so efficiently.
Monkey Fish Example: "The monkey wanted to save the fish from drowning so it kept the fish on top of a tree", telling the current state of our leaders (monkeys) who are claiming to be saving us (fishes) from drowning.
Importance of Increasing production power in the country
Rival good/Non Rival good: The difference between food vs information. Some one accessing some information doesn't effects others chances of accessing the information. Also talked about public good and private good.
Marginal Cost of Duplication of Information
Difference between knowledge and information: Knowing is the summary of information received by a human being, analysed and stored in the brain, its a private good.
Increasing Production vs increasing employment: The former one being more important.
Ability to adopt innovation Vs Information for Development: Explained it with the example that its important for end users like farmers to adopt the technology rather than knowing the science behind it in the first place. Once wheelbarrow was invented it was not supposed to be re-invented by each and every person who needed to use it. They can simply use it. Cost of innovation is high and that should not prohibit us from using the latest up2date technologies and growing at the equal pace as others in the world do.
Removing illiteracy: this one was a very interesting. Why could over government never actually tackle the illiteracty problem. How and why they should actually do it now. "A illiterate's children could be literate by a literate's children will never be illiterate" that explains how easy it is to get rid of the literacy if the government indulges in a proper investment at once rather than a slow/partial polity over 50 years which has failed. How goverment spends around 4000/- Rs per year for a child's literacy in the country and what if the government kept an offer to private sector to literate the children at say Rs 20000/- per child for all the 100 Million ppl in the country.

Some discussion went on for command and control vs property rights, but I was not able to get all of it.

Laptops to Buy at felicity

We just had a look at the laptop shops during the felicity, after some discussion with the vendors we arrived at the conclusion that HCL laptops would be the best buy for any one. First those guys provide good configuration, Second they have good service, Third they don't trouble you with OS installation and stuff.

The third point is worth consideration, HCL vendors not only give you freedom to put OS of your own choice during the purchase but also change it later without bothering to inform them. All others vendors have at max single installation system that too is fixed when you buy the laptop. Toshiba will provide option to install Linux in the beginning but their models are very restrictive.

Acer too provide Linux installed machine, but I think they will also not be flexible with the dual OS installation if some one needs. Cost of a HCL laptop will probably the best offering for that configuration, which is what made them so successful at Desktop purchase in IIIT too.

Also note to ask hardware compliance with Linux when purchasing a laptop. The simplest option is to put a Live CD in the laptop's CD/DVD Rom and check if all the hardware is supported.

Bad Bad Luck at felicity competitions ... :(

Yesterday evening, I was informed that our team( different partners) secured good positions in both the contest prelims. Today's final turned out to be not so gr8 after that earlier performance.

In Debug the C-Bug, Ranjith and Me(I?) were doing quite well, until we realized we just didn't have to debug the program but get the desired output which was given. (Ranjith had earlier experience of Debug and has describe to me that we will be given a code and will need to find out what the code is about and make it work. When we were going to show our solved 4 questions, we realized we actually near the correct output in only two of them. Later I spent lot of time in the first question and didn't give much time for Ranjith to do the second question, which was almost solved. After the contest Ranjith described the exact answer to me, and we are still blaming each other for getting the Second Position and not the first. Which we missed by only one question. :(

In Java Jargon, Sachin and me formed the team. it was a total hardluck, we solved so many questions wrong ... Not that we didn't know them but because of no coming back rule maybe. You were suppose to answer the question or leave it before going to next one and there was no coming back. We did the contest too early and didn't utilize the whole team. I do take the blame for that but also blame the organizers for creating quite some confusion with some wrong questions and not keeping enough supply of questions. We didn't get any prize in this contest.

BTW yesterdays' evening programmes were Awesome. I am gonna blog about them when I get more time.