The XORBlog

This post is not a comparison or technical analysis of why YouTube is better than Google Videos or vice versa. The title of this post could be more general: Is fragmented Internet traffic better than centrally redirected? But I thought an example is probably better as a title. What I am trying to figure out is whether it's better to have a new brand for each specialized service or a mega-brand like Yahoo!, Google, and MSN etc. as an umbrella brand for these services. Whether the whole really is greater than the sum of the parts in internet branding. Here is what I think are the advantages of both approaches:

The umbrella brand approach:

* The marketing (or brand building) budget is perhaps better utilized. However, brand building on Internet is mostly about creating a great product and word-of-mouth. I have yet to see an ad for Google. I once read somewhere: your product is the best brand building tool you have. This is even more true for Internet businesses.

* Brand is often associated with an expectation of quality. When you launch a new product or service under an existing brand, people automatically expect it to be the same or better than the existing ones.

* Cross selling (referencing): You can cross link a new service from a popular service hence increasing its chances of getting noticed.

The new brand per service approach:

* Privacy concerns. One company may be knowing too much about me, perhaps I am better off using services from many providers. Remember brand is mostly about perception, which may or may not be the same as reality. So, even if del.icio.us is owned by Yahoo!, users don?t really associate it with them unless they see a large Yahoo! logo.

* A failed attempt will not impact your existing brand(s) negatively.

* Your new brand could be bigger than the old one. Arguably Palm is a bigger brand than 3M. Imagine if it was called 3M Computer (perhaps it was ).

* If you have only one brand, users might have difficulty associating brand with services other than the core service. Often you can read posts on Internet about how Google is a "one trick pony". Google is still largely associated with web search. Once you go to Google, you end up searching, it diverts your attention. Once you go to YouTube you can?t do anything other than watch videos.

The Internet services seems to be going the way of Unix, where you have services that do one thing and do it well. You can then combine (just like pipes and redirects) these services into Mashups.

Let's now have a look at current trends in brand building. Intel decided to dump Pentium and focus on the Intel brand only. They subscribe to the one brand fits all philosophy. This seems to be the strategy followed by elite brands like BMW. There is no BMW Fiesta, it's one brand and everything else is a part number. Only time will tell whether Intel will be successful with this approach. Google seems to be following this strategy too with their gXXXX naming convention.

On the other side, we have businesses that have a whole bunch of brands. We already talked about 3M and Palm, though Palm was not a core 3M business. But lets see Apple, it's not an Apple MP3 player but an iPod, not an Apple PC but a Mac. Though Apple is a very strong brand, their "sub brands" seems to be doing pretty well too. If you have seen latest Yahoo! acquisitions, they are mostly opting for keeping the services under their existing brands e.g. Upcoming.org, del.icio.us, Flickr etc. Though in the olden days everything was marked Yahoo! XXXX, e.g. Yahoo! Stores, Yahoo! Auctions etc.

Now the question to ask is: Whether brand or naming really is what makes a product successful? The answer is, of course, not. The most important thing is to make a product that people need (though sometimes they don?t know they do) or want. Then go out of your way to make it easy/pleasure to use. But brand does plays some role in how people perceive your service and this could impact their decision to use your service.

So, will Google Videos ever be YouTube?

The wait is over, SLED 10 is finally here. I switched to Ubuntu when first RC of Dapper Drake was announced and I can safely say it has exceeded my expectations in every way. This is the best desktop experience I have ever had. It looks good, it feels right, it updates itself and I have always found packages of software I needed. Also, what I particularly like about Ubuntu is that there is no community edition and a real version. Ubuntu 6.06 LTS is all that everyone gets. I guess, this saves Canonical a lot of trouble and confusion.

Having said all the above, I have always been following SuSE Linux and I believe Novell is being very innovative in desktop space. So, I eagerly awaited release of SLED. Here are my first impressions and major differences that I see:

* They both install very easily. Everything just works out-of-the-box. I think, SLED has slightly better network configurator, but not a big difference.

* They both recognized my other HW and gadgets in a plug-and-play mode, this includes digital camera, Creative Zen Micro and printer.

* SLED comes preinstalled with Compiz and XGL. I have also tried the 3D effects in Ubuntu but SLED feels faster and more responsive. The reason could simply be that Novell has done more integration testing than Ubuntu.

* Yast vs Ubuntu tools (synaptic et al): They are both easy to use and navigate through. Though I have become used to the apt-get command line so much that I hardly use the graphical tools, but for users who do, they both do the job very well.

* I liked the new "Start bar/box" in Suse. It's after a long time that someone tried to innovate in that area and broke the "start menu" metaphor. It's also possible to have this menu on Ubuntu, here is how: http://angelicpenguins.blogspot.com/2006/07/sled-menu-for-ubuntu-uslab-now-in-repo_14.html.

* Ubuntu gives you a better choice of desktop-applets than SLED but this is a minor issue.

* It's slightly more work to set up Ubuntu as a development box than SLED. Ubuntu doesn't offer install time possibility to select the development tools.

The screenshots

In conclusion, SLED 10 brings everything that Ubuntu does and feels slightly faster. But Ubuntu has bigger and nicer community and .deb is simply better than .rpm. Out of 10, I give both of them 10. Linux desktop has never been so good.

BTW, this was the first time I resized a partition in use with the help of parted. It's been reliable so far.

I managed to get my hands on Freakonomics by Steven and Stephen over a week ago. I must say it's a pretty amusing book. While reading the chapter on cheating teachers it struck me that the kind of cheating described in the book doesn't seem possible with the Indian Education System.

Now please don't misunderstand this post for an endorsement of the Indian System. It's not even a comparison, and perhaps there are other ways in which teachers cheat in Indian schools, but I am just trying to explore that the cheating as described in the book, don't seem likely.

In the US "high-stakes" tests students are tested by the same teachers who teach them. The tests are standardized by the state, but still the same teachers evaluate the answers and can possibly correct the answers for their students (please someone confirm: the teachers also have the key to questions?). Of course, all I know about this system is from this book, so there could be a lot more to it. Here is how it works in the Indian System. State is responsible for the standardized tests as well, but it produces many sets of tests which are on the same level but not identical. Let's say there are 4 sets of math examination papers A, B, C and D. Each school is sent all the sets, but students sitting next to each other never get identical papers. The teacher who supervises the test is generally from a different subject/area e.g. a history teacher will supervise during math examination. The students are anonymized, they are given a "roll-number", their names are not on their answer-sheets. The answer-sheets after collection are immediately sealed and sent to an another school, which doesn't know which school it came from and obviously doesn't know student names. The teachers receive their keys to the exam papers only when they are evaluating (answer-sheets from another school). So, the teachers cannot correct answers and they cannot help during the tests. Also, the students cannot cheat among themselves as they have different papers.

I can't find a flaw in this system. The only way a teacher (or school) can cheat is by giving students more time, but that's not very likely either as the whole school starts and finishes the exams at the same time (say 8:30 to 12:00). And there is, of course, fear of surprise state examiners visiting your school. It's a complicated system, but it seems to work at least against cheating teachers.

Also, according to the theory put forward by the book, everyone responds to incentives. The teachers in US are highly motivated to show great results for their students as their career and in some states even their bonuses depend on that. In the Indian education system no one really cares. The only way teachers can make additional money is by teaching (tutoring) on the side. So, again the cheating in school will not help them in anyway.

(Note to myself: A post later about exploring incentives for participation in Web 2.0 social networks)

This is in reference to: http://blogs.sun.com/roller/page/jonathan?entry=the_rise_of_the_general

Sun blogs don't seem to allow posting comments directly. It goes through some sort of approval, so here are my thoughts anyway:

It seems Sun is closely cooperating with Google. When Jonathan says: "how do I replace a dead drive" in Thumper - the answer is you don't, you let Solaris and ZFS simply remove it from use. He is just following the Google principle: Let Software take care of HW failures. Way to go! It's good to see Sun responding to the real market demand and not cooking up new dishes that no one seems to care about.

IMO, Sun stands a very good chance of a rebound with following these strategies and continued R&D spending (and not listening to Wall Street ).

Abhishek

Blogosphere is full of stories about how the blogs are some day going to replace the mainstream media. Well, what is mainstream media anyway. The Wikipedia defines mainstream media as media that is designed to reach very large audiences. But I think that‘s not what people mean when they say that blogs will replace mainstream media. What they more specifically mean is that the large existing media outlets will become irrelevant over time. In other words, aggregators like delicious and digg are going to make the new york times or washington post irrelevant. That‘s an interesting hypothesis that looks about right if you see the popularity of the aggregators against the old media websites. But let‘s see what exactly does an old school media outlet does:

• It creates the content or sources it from news agencies like AFP, Reuters etc.
• Provides a channel for this content in the form of newspapers, magazines, TV channels etc.

In blogs, the content is created by anyone who has anything to say and there is lots of it. As Paul Graham in one of his speeches correctly observed, that most of the writing/blogs online are of poor quality but the mainstream media is not competing with the average but the best content online and the aggregators help you rank the content. And aggregators are also how most of this content will be accessed so they are the channels. This means that blogs are in fact a major threat to the current media outlets.

This seems to be also the major difference between the strategies of Yahoo! and Google. Yahoo! wants to be a modern media outlet that creates it‘s own content and the Yahoo portal is the channel. Google on the other hand sees itself as more of an aggregator. It doesn‘t produce the content but helps you find it by organizing it. Only time will tell which is the correct strategy. Perhaps they both are. BTW, what about Microsoft? They seem to be doing everything: MSNBC (old media), MSN (like Yahoo!, AOL), Live.com (like Google), XBOX (Gaming platform) and of course Windows. Whatever happened to focus?

Anyway, the purpose of this post was not to show that blogs will replace mainstream media (or better said: Blogs and Internet will become the mainstream media), but to find out in what ways the mainstream media can integrate blogs and benefit from it. A new service called BlogBurst.com tries to do just that, which IMO is a brilliant idea. However, there are many other ways in which the major media outlets are using blogs:

• The existing columnists have blogs in addition to their regular columns e.g. Gartner, CNET, Washington post etc.
• They are buying the blogging infrastructure provider or aggregators e.g. News corp buying MySpace.com
• They are becoming the blogging service providers and featuring blog content alongside regular content e.g. Times of India

Of all the above strategies I liked the Times of India (www.timesofindia.com) approach the most. They help you get started about blogging and if your blog is popular they make it available right next to their other news stories. I see that top blog of today has a whooping 208 comments. And for every news story I read there, I atleast read one blog entry, so it seems to be working. They get the content, and the content author gets visibility and gets paid for it (google ads). Next few years are going to be very interesting, as always the ones who adapt will survive and thrive.

There is a lot of talk on leadership going on these days. Everyone wants to be a leader, but I wonder in a world full of leaders who will be the follower? And then it dawned on me: you cannot be a good leader unless you are also a good follower. A good follower doesn‘t mean someone who just does whatever you tell him, but is also observant, fast learner and contributes his ideas back to the group and leadership. A good follower also learns from a good leader. I guess, only once you have been led by a good leader you can become a good leader yourself. And just as it‘s not easy to be a good leader, it‘s also quite difficult to be a good follower.

Now back to leadership. As everyone will tell you a good leader has a vision and is very good at communicating this vision, breaking it down to manageable tasks and then follows up hard to make sure it gets implemented. Is a good leader just a follower of a vision? As I think more about it, it seems that a good leader needs to have almost all the qualities of a good follower but also has the big picture thinking, can communicate more effectively and is good at breaking down tasks into manageable chunks. Perhaps in other words a good leader is a visionary (his own visions or inherited or stolen?), a salesman and a manager all rolled into one.

So, in conclusion it seems that leaders and followers share a lot of traits. Don‘t be worried if no one has called you a leader yet, if you are a good follower you are already on your way to become a good leader. :-)

Update: Let‘s see how far can we take the analogy.

• Being a good reader makes you a better writer: YES
• Being a good listener makes you a better speaker: YES
• Being a good shopper makes you a good seller: NO
• ... I will think some more :-)

I started using Ruby on Rails a few months ago and I can‘t remember I have ever been so happy while coding in past. But I have also been wondering what are the ways in which Enterprises (large ones) can start using RoR. So, I started thinking about the strengths and weakness of RoR and how it can basically fit into the Enterprise Software Development just now. Of course, introducing anything new is not just a technical but also a political battle. But let‘s see what are the strengths of RoR compared to other software development technologies like J2EE or .NET used in the Enterprise:

- Rapid development: You can‘t deny this. And it‘s not the old drag and drop RAD bulls**t, but we are talking real gain in productivity due to a smart framework. RAD tools were supposed to help with the plumbing, but with RoR there is no plumbing. No configurations, no mappings, no wrappers, just pure business logic in a pure MVC framework written in a pure OO language. It doesn‘t get better than this.

- Happy programmers: As I said earlier I have never been happier while coding. When you start with RoR, on an average there are about 2 WOW moments per hour and not a single hair pulling moment for months (I am yet to encounter one).

- Built in Web 2.0 capabilities like AJAX: This is important if you want to convince some folks that a browser interface could in fact be user friendly and sexy.

As for the weaknesses:

- Scalability: Though I haven‘t encountered it, some folks say there are scalability issues with RoR. I think, it is possible to scale RoR applications, but enterprises are paranoid, so let‘s keep it.

- No major vendor: Technologically this is a non issue. But enterprises seem to pay very big attention to this. The main reason being that they don‘t want to end up with dead ends. It‘s a topic for another debate whether there should be so much focus on vendor stability, but for now it‘s a fact of life so let‘s keep this too.

Now, where can this lead? We have a tool that helps us build something real fast, makes us happy in the process and has some really powerful features, but apparently it can‘t scale and there is no major vendor backing it. What if we use it for prototyping? I am talking about throw-away prototyping, where you build a prototype to gather requirements. After the prototype has served it‘s purpose you throw it away and can start afresh in your Enterprise tools. This way you can also train your existing programmers into new technologies and when RoR or similar framework becomes mainstream you will have a head start. Arguably, this will also make your Java programmers better Java programmers. Sounds like a pretty good deal to me.

One major issue I have had in some projects is that the prototype ends up becoming the system. If you use different (right?) tools for prototype and the system then you are forced to throw the prototype away, which could be a good side effect. :-)

I am having an interesting discussion with Mr. Ramesh Loganathan on Web 2.0 vs. SOA. It seems from the architecture point of view they are the same thing: content and services made available, managed, orcastrated and consumed over many channels including a rich interface. In Web 2.0 the underlying technologies and concepts are RSS, REST/WebServices, AJAX, Mashups, Portals etc. whereas in SOA it‘s an integration backbone like ESB, Web Services, Portals, also some form of Rich interface. Of course, some implementation technologies will overlap.

More @: http://jroller.com/page/rameshl?entry=while_web_2_0_vs

Update: The debate is getting more interesting :-) more @ http://jasonkolb.typepad.com/weblog/2006/04/soa_vs_web_20_t.html

In a previous life I had studies some advertising and PR. And I was wondering how much of the old marketing theory is still applicable in a Web 2.0 world. Philip Kotler defines marketing as: “Marketing is human activity directed at satisfying needs and wants through exchange processes.“ That really is as direct as it can get. So, the first thing is, you have to “create” something that people want, and this applies to any field including Web 2.0 services. But once you have (or believe you have) that something, you need to get the word out. Let‘s look at the old school 4 Ps of marketing and how they apply to Web 2.0:

1. Product: So, you build something that people want or need. This is a tricky one as everyone thinks he was got that killer idea waiting to take on the world and Web 2.0 makes it remarkably easy to “create” things. But in Web 2.0, how can we be certain that there is a market for something. Can market research help? I don‘t know. Perhaps the easiest way is, to build something and put it out there (online). Ask for the feedback and let you service evolve in the way users want to use it. In the end, the users will decide what they want to do with your service. There is no silver bullet.

2. Price: Zero dollars. Now how do you make money on something that you have to give away for free? Is Advertising the correct answer? Only time will tell. First of all the price doesn‘t have to be zero dollars, some sites do charge for content or usage of their services (SAAS model). But most Web 2.0 companies haven‘t yet figured out how are they actually going to price their products. Pricing will play even a bigger role. Example: before iTunes no one really paid for music online, but by pricing a song at 99 cents and making it stupidly easy to get your money, Apple succeeded.

3. Place: The web. Everywhere, anytime. So, I guess Web 2.0 does take care of this part. The days of controlling the distribution channels are mostly over. Though for content you still need channels like Yahoo, Google, MSN, AOL etc. but for the most part the distribution channels are irrelevant. And as the search technologies continue to make progress it will only get more flat.

4. Promotion: This is what it comes down to in the end. How do you get the word out. Also, part of promotion is the BRAND. And I believe the Brand plays even a bigger role in online world than the “offline” world (I don‘t like to call it the physical world because Internet is now mostly part of the physical world too). Of course, the old way is to use banner advertising on other websites to promote your website but that‘s not what smart companies do. Have you seen any ad for Google? So, in the Web 2.0 world it‘s all about word-of-mouth promotion. In fact, that‘s how I learnt about most services. This will become increasingly important as the cost of communication goes down and people can more freely communicate their ideas. The latest trend is to use blogs to promote your ideas or products. Or build communities around your products or services, but how many communities can a person really manage at once? This will become an issue, perhaps we will see tools that help manage your participation into the Web 2.0. Aggregators are partly an answer to this problem, but it‘s still only one way.

So, it seems that the old theories can pretty much be applied also to the new ways of doing things (new economy anyone?).

It is said that it was the toolmakers that made the most money in the gold rush. It was also true for the Web 1.0, but this will be not be easy in Web 2.0 gold rush, as the basic tools are now essentially free. In fact, it‘s the availability of free tools (LAMP, Ruby, Python, Firefox, JavaScript etc.) that has led to such a fast growth of Web 2.0. Is there any money for the tools makers in Web 2.0? I guess, that‘s a topic for another post :-)

It was a rather interesting evening. While reading an article about Web 2.0, I came across project Oh No Robot. Which BTW is an interesting concept. A search on ONR somehow lead to Scott Adam‘s Dilbert website with a banner for the free e-book God‘s Debris. I had no clue what the book is going to be about and never thought I will finish it in one go. But it is so compelling that I just kept reading and lost all the sense of time. Which is ironic because it‘s one of the ideas in the book that the time doesn‘t exist.

The book tries to define the God in an interesting way as being omnipotent (or not?). The book builds on the premise that we (humans) and everything around us is in fact part of the God. Which is very similar to the Hindu beliefs. There is also the concept of Karma explained in a non-religious way.

I specially liked the ideas about pattern recognition, as I believe that pattern recognition plays a very big role in our personal and professional lives. In a way, increasing your level of awareness is all about getting better at pattern recognition. I guess the old school name for this phenomenon is “Experience”. :-)

Overall, a great read and if it wasn‘t for web 2.0, I would probably not read it.