Tuesday, August 21, 2007

Kudos WhyFirefoxIsBlocked.com

Why is it hard for people to understand that there is nothing called "Free Lunch"?

Somebody was concerned, that adblock is ruining their revenue by blocking the ads (shown on the website). And at the same time, not providing the website owners, the ability to block adblock users. They created a site (if you are using Firefox you will not be able to see the website; use IE tab extension or any other browser to see the site.) The site tried to put up a genuine concern. They were targeting Firefox because it abetted adblock.

Now today I see a childish blog. David Peterson tries to counter the points put forward by whyfirefoxisblocked.com. He tries to give some apple and orange comparisons as well.

Now, how hard is it for people to ponder on "Free Lunch". Basically there is nothing called a free lunch. You have to pay for it. What you pay or how you pay is a different ball game all together. Its because of people like David Peterson, Richard M. Stallman came up with the phrase "Free as in Freedom and not as in Free Beer". I have a little advice for people who believe in Free Lunch and Free Beer. Why don't you guys open up a small place and provide Free Lunch and Beer to everyone? Or coming to more technical terms; why don't you guys open a site which boasts about ad free internet? I even dare Mr Peterson to open a website, and maintain it without serving any ads and without charging any subscription fees. Mr Peterson, does not himself serve an ad free blog, then how come he objects to something that he himself believes in?

Put your efforts where your mouth is. If you believe in ad free internet, stop serving ads on your blog. And even further, people who say that quality content generation, without incentive, can be sustained in the long term; I would like to point out safari. Safari is the ultimate resource, but with its own set of limitations. Safari does not include all the books published by O'Reilly, it also has a high subscription fees and after all this there is a limited set of books to keep in your shelf. Why doesn't O'Rielly allow free unlimited download of all the books?

How charming the concept of Robin hood might seem, but the concept is not everlasting. Robin hood was there to bring a balance amongst people, once the balance is restored, there is no room for Robin hood. People will make fun of one, who wants to be a Robinhood in a balanced society.

Coming to the freedom part, Freedom is not selective. Freedom is freedom for everyone, for every single human being, for every single living form. If Freedom was selective, i.e. freedom for some and not others, then what is the difference in freedom and tyranny? Adblock provide users, the freedom to see ad free pages, and robs the website owner with his/her freedom. Doesn't it sound like dictatorship?

The problem is not with Firefox, but the problem is with its promotion of adblock. If Firefox and adblock both existed independently there was no problem. And let me remind you, that if M$, instead of arm twisting OEMs, had promoted netscape, the web would have been a completely different place; there would have been no internet explorer as well as no Firefox. We are seeing these things happen because M$ chose to promote internet explorer.

Lets look at the scenario from Firefox's perspective. The default web page for firefox is google. This is because google pays Firefox; what google is actually paying is revenue generated through ads. Why doesn't Firefox make wikipedia as its homepage? It seems to me like Firefox pockets the ad generated revenue and then asks adblock to stop revenue generation for other small players.

The internet is fueled and made interesting be actual living people. And humans have a family to support, they have bills to pay, how can these people live without any incentive. I would call it even, if with the ad revenue I am able to pay the net connection bill, but adblock is not even allowing me to do this. The current status of Artificial Intelligence is not so good that it can keep generating quality contents without human intervention. Even if it was, computer hardware needs to be bought, and somebody needs to pay to the net connectivity and electricity consumed!


AddThis Social Bookmark Button

Tuesday, August 14, 2007

Road of Open Source success - III

This is the third installment for this topic. Here you can find first and second posts.

OoBE
OoBE is as abbreviation for Out Of Box experience. I would like to quote "You never get a second chance to make a good first impression". I don't remember who said this but this is something that I firmly believe in.

OEM is a key player in this field. What goes inside the system is dictated by an OEM, and there are a lot of people who don't bother to look for alternatives and just stick to what the OEM has provided.

If you take windows as an example, Vanila install can not play flash, does not have a JRE installed, can't read pdf files, can't play DVD, can't play formats other that mp3, wma & wmv. Only with windows XP, users were able to view compressed files, and the compression format was limited to ZIP. If you give this vanila installed system to a new user, s/he will certainly say that M$ is worthless.

The problem is not only for home desktop users, its also prominent for corporate users. Corporate users try to avoid this problem with the use of custom images. Someone from the IT department is given the responsibility to install windows along with all the required applications, on a single machine. Once this installation is complete, a full image of hard disk is extracted. This image is then copied to the rest of the machines.

Now coming to the open source world, the same issue comes up. The reason for this is not technical but political. Almost all the major distributions, in order to avoid litigations, stay clear from proprietary code. In order to attract more users towards open source, a better out of box experience is required. And nowadays it means that a user should be able to browse the internet without any issues.

I remember my first encounter with Linux, It was with Redhat 5. It really gave me a hard time to bring up the GUI. This was long ago though, and things have changed since then. But the automatic GUI configuration is still important. All the new displays are LCD and there are more coming in the form of a wide screen. Some advanced users have already started using dual monitors. But, still no distribution correctly configures wide screen or dual screen monitors correctly, even after enabling DPMS .Moreover, in case the installer automatically create an xorg.conf, a lot of distributions tend to prefer 16 bit depth instead of the better 24 bit.

The next point is the network connectivity. For majority of the computer population, it means wireless. iPhone comes with Wifi, even the nikon S6 that I purchased 2 years ago had Wifi. Giving complete driver support for wireless cards is a daunting task as manufacturers do not release the required specifications. Even then, distributions should make sure that users get their WiFi up and running without any hassles. I am not a big fan of Ubuntu, but it really excels in this field. Slackware along with zenwalk, on some machines, are the worst; they even have problems with ethernet.

Last but not the least are applications. Firefox solved a big problem in terms of "easy and robust platform" required to access the wild wild web. There are a lot of good addons and users feel at home with it. Next problem comes with plug ins. JRE, flash, acrobat reader, quicktime are required to provide a flawless surfing experience. A lot of of distributions still have some problems with this.

One more aspect comes to my mind and that concerns aesthetics and ergonomics. Some distribution dont provide a custom theme and when some do provide, your eyes pay the price. Distributions have to also consider themes as part of the initial install.


AddThis Social Bookmark Button

Monday, August 13, 2007

Road of Open Source success - II

In my last post I tried to explain package manager as one of the key ingredients for open source success. Today I would like to discuss about Control Center

Control Center
Control Center is an application used for configuring all the aspects of your operating system along with the user applications installed on the machine.

If we consider windows, user has a Control Panel. But all the configurations are not concentrated over there. Some configurations are accessible by MSC, whereas user applications follow a completely different approach. Sometimes the user applications add a control panel applet; sometimes they store the information in a registry and options are accessible through Tools>Options and sometimes special ini files are used for configuring the applications.

In *nix, all the configuration options are accessible through a single repository /etc/. Different applications have their dedicated configuration (*.conf) files, which are text files and can be easily edited using an editor of choice. This is plain and simple. But for a new linux convert, this might be a daunting task. It is possible for him to mess up the configuration file.

Recently, wizards have set a benchmark in configuration. They are almost foolproof and extremely user friendly. And it's really easy to create a wizard to generate the required configuration files. Redhat has a couple of disjoint applets that allow to configure some applications. Ubuntu is also coming up with a screen configuration applet.

But user needs a unified mechanism for configuring all the aspects of machine. Something similar to yast or drake and recently smart. The Control Center is a small software that will make windows to open source transition easier for a newbie.

Ubuntu is unofficially the #1 but lacks a unified control center.It surprizes me how the #1 distribution can be without a control center. The next two distributions; PCLinuxOS and openSuSE both have it though. There are a lot of distributions out there and majority of them lack a Control Center, some have re branded KDE control center but none of them is complete. We need more and better more Control Centers to excel Open Source.


AddThis Social Bookmark Button

Saturday, August 11, 2007

Road of Open Source success - I

One of the main ingredients for open source success is OEM (Original Equipment Manufacturer) support. A lot of companies are doing really well in this regard. Ubuntu has partnered with Dell to offer preinstalled desktops. SuSE & Redhat offerings are already available with HP and Dell, though its only for the enterprize server segment.

Apart from the OEM support, there are three more ingredients which attribute to the open source success, namely package manager, control center and out of the box experience.

Package Manager
Package management is not so easy in linux. The original packages or applications were distributed in the form of source code. There was a lot of chaos in the way packages were built and then everything converged to three command step; 1. configure 2. make 3. make install. But even this was not enough. People wanted prebuilt packages, thats when three big giants of linux distribution started, Slackware, Debian and Redhat.

Everyone was having its own type of package management; slackware had tgz, debian had dpkg and redhat had rpm. All three were incompatible. All stood the test of time, but in my opinion only dpkg succeded. Apt is another way to install and configure applications and also contributed the success of ubuntu.

RPM was good initially and also became the industry standard for linux packages; all the enterpize applications started distributing their application in RPM format. But updates are really hard in RPM and we have to deal with package dependency like hell. People are recently moving away from fedora and towards Ubuntu. One of the reasons I see is package management.

Gentoo introduced protage, and it seems to be panacia; it offered great promises. The user has the ability to compile the package with the optimization s/he chooses. Its really great to run dedicated servers or personal workstation on old machines. But it did not find any place in enterprize application. Recently there has been good efforts in form of pacman. Pacman seems promising.

When can the open source agree for a single package manager, or even provide interoperability between packages?

Consider a single scenario, I am using applicaiton X on Centos and feel it is really great. My wife comes over and wants the same application on her ubuntu machine. So what do I do, the best shot that I have got is to download it from one of the ubuntu repositories. If I find it there, its good, but if I don't, I have a bit of trouble. Considering Debian as Ubuntu's godfather and knowing the fact that Debian has the biggest repository of applications, I search it there and find one. But now I am unable the install the same application as that it seems to be incompatible with ubuntu.

So, considering my geek status, I would download the source package and try to install it. This option might not be available for average joe. Now Ubuntu does not provide build tools and kernel headers by default, so I have to install these in order to install the application X. I will feel a little bit frustrated but consider that a linux newbie comes to me and I explain all this to him, he will be petrified by the complexity.



AddThis Social Bookmark Button

Thursday, August 9, 2007

Stategy - Battle vs War


A month ago I had to convince my wife regarding a hair shampoo (imagine a geek discussing shampoo). So, the problem was that my wife was suffering from an acute hair loss ; she consulted the doctors, changed her eating habits, applied a herbal conditioner, and did all sorts of stuff ,but, all in vain. One thing that she did not do was "changing her shampoo". It struck me, and I discussed this with her. But she had been using the same shampoo for several years, hence she was in no mood to change .

I tried to switch her to shampoos that are known to help hair loss. I even forced her to aisle containing shampoos and showed her 'X shampoo stops hair fall'. I made it a point to make her notice the shampoo ads, whenever we watched TV . But everything went down the drain. She was "My shampoo is the best and nothing else matters".

So I had to rethink my strategy, I had to try something different; something that will force her to change her shampoo. After thinking a lot, I figured out that when I was persuading her for a different shampoo, I was pointing here to a shampoo that was of my choice. That might be the reason for her not switching. So I changed my strategy and asked her to just try any other shampoo except for the one she is currently using.Though there was a risk that "just any other shampoo" might make the hair loss worse; but I was ready to take the risk.

On our next trip to the grocery store, she picked up a different shampoo, and things were never same again. The new shampoo reduced the hair loss and she concluded that her long time favorite was actually causing the problem.

So what can we derive from this experience?

I am a long time BSD fan and can take the flame wars to any extent. On the other hand, I really really hate M$. I am fed up with their monopoly and their disregard for standards. I want people to jump off the M$ brand wagon and try open source software. Thus, if someone comes to me asking for Linux help I will provide all the aid I can, just to make sure that s/he converts.

I have seen a lot of flame wars, a characteristic of open source; forking different project because of zeal/ego of developers. The mergers also actually happen, but they are rare. There are distro zealots who go to any extent proving that their distro is the best and nothing else comes even close to it. To some extent I agree, different distros were made with different visions and there is no such thing as "One size fits all". And recently there has been a lot of distro bashing for distributions who joined M$ patent agreement.

But amid all this, the open source community should stand united against M$. It should make sure that people leave M$ bandwagon and join the open source revolution. Whether it is Freespire or SuSE , it should not matter. The community should be focusing its efforts to spread awareness amongst people.

To all the purist who say that distributions that made agreements with M$ betrayed them, I would like to ask them about GNU/Linux. Everybody knows that linux is just the kernel, and there are other kernels as welll; solaris, bsd, hurd etc. In making a distribution a lot more than a kernel is required, and that "lot more" is actually what GNU has provided. Even the compiler used for building the kernel is provided by GNU. Linux cant exist in isolation, it requires GNU support. Then why is it that Linux is named in isolation; why don't we accept "GNU/Linux" as suggested by RMS and supported by Debian. Don't you feel that Linux betrayed GNU by not putting their name?

It is most important that we "the community" should stand united in our war against M$ . Ubuntu, PCLinuxOS and OpenSuSE are the top three listed at distrowatch, but I have my own criticism for each of them. The criticism is so extreme that I literally hate them. But that does not matter, if they are able to pull people out of M$ bandwagon.



AddThis Social Bookmark Button