Archive for the ‘Blogs’ Category
Decentralized Social Data Framework: A Modest Proposal
Three (good) Indian blogs that you probably don’t read
Time to post something different for a change than the regular longwinding ones.
Shashikant’s tiny world: He posts across a wide variety of topics covering technology, the economy and other aspects of current affairs. He has something that is desperately lacking in most of the Indian blogosphere: a perspective that is not a wannabe version of the popular western ones. If only he would change the circa 2001 blogger template to something more contemporary, but I can hardly complain since I read the full feed in an RSS reader. And, no, I am not linking to him because he’s linked to me.
Gopal Vijayaraghavan: Gopal works for Yahoo! and is one the leads for the PHP APC cache. That does not mean he posts only about profiling PHP code and race conditions in it. He also writes about movies (from a very non-critic and normal viewer point of view) and a lot of other non-tech related things. The only minus point is that he does not allow comments on his blog.
Cleartrip blog/Hrush: I know this is a corporate blog, probably disqualifying it from being considered as a normal blog. But most of the content on the blog is penned by Hrush Bhatt, Founder & Director, Product and Strategy for the company. Other than the fact the blog is one of the best and the most open blogs among Indian corporates, he also gets additional brownie points from me for quoting two bloggers in the web data sphere that I follow closely: Danny Ayers and Joe Gregorio. Minus point: Not updated frequently enough.
While on the topic of blogging, I was wondering recently if the only major difference that blogging has brought to the platform is that being biased is no longer uncool? These days, I tend to switch off from any discussion that aims to figure out the biases of mainstream media. The fact of the matter is that everything and every human being is biased and we are conditioned by our biases.
The only difference is that it used to be cool to claim that you were unbiased as a media entity. When you report from the field, you are supposed to stick to the facts and not colour it with your biases. I think this is quite badly misplaced. While, as a blogger, it is cool for you to be biased. In fact, you are encouraged to come clean on your biases than cover it with a veil faux neutrality. Other than that, if you take out the scale and economics of the matter, there is hardly any differentiation: Both sides have bad reporting, band language and myopia to the obvious.
Interesting and hypothetical over-the-top question of the day: Can you imagine a newspaper filled with op-ed writers?
Paging Messers Page and Brin: Please shut down Orkut
Two of Google’s worst products in its product line up are Orkut and Blogger. There are various reasons why those two deserve that label, but when a company worth billions, with more PhDs than you anyone can count on its rolls puts up a notice that says, “Security tip: Never paste a URL or script into your browser while logged into orkut.com, no matter what it claims to do,” it really does not get any worse than that. Google, please do yourself and your users a favour and shut the damn thing down till you fix it.
Apparently there has been a spate of recent Google Account hijackings that don’t follow any particular pattern. There is a fairly high probability that the warning on Orkut has something to do with one of the twin curses of Web 2.0: a CSRF or an XSS attack. Orkut handles its authentication and cookies differently from the rest of the Google framework.
You can log into Orkut and also be logged into other Google products like Google Reader and Gmail without being prompted to authenticate yourself again when you browse to those products. Conversely, if you log into the other two and browse over to Orkut, you will be faced with the authentication prompt.
In all probability, Orkut is using another cookie of its own in addition to the Google account cookie and somewhere in between a malicious script is hijacking the Google account cookie, using the cross domain permissions that are granted to Orkut pages to do the initial authentication on the GLogin.aspx page. In any case, Google should have fixed the problems with Orkut than to expect users not to paste a URL or a script into the browser while they are logged into the website.
Google’s greatest strength is its computing framework (one that even Microsoft will take a lot of time to catch up with its ‘cloud’ initiative), where applications basically plug into Big Table and GFS, requiring relatively smaller teams of developers to sustain and develop the lesser-important products; Orkut and Blogger belong to that category. After all, since when does getting an Ajax button to post a comment or having product blog (OMG! We have a blog now, we are so 2005!) or having dynamic pages on a blog network represent significant advances in the history of humanity?
The trouble is that the same strength works as Google’s major weakness too. Since they don’t need massive teams to deploy and sustain these applications, the products don’t get the attention that’s required and function mostly on autopilot. And unlike what most people think, Google does not really care much about being a segment leader as long as they can mine usage data, do behavioral analysis and use that to improve the advertising cash cow. But that does leave holes like these open, which is just not done and I hope Google fixes the holes soon before someone figures out a Orkut-wide attack.
p.s: Get someone to fix the language in the warning. It almost sounds like they are urging users not to use Orkut irrespective of what the site claims to do.
Dude, who stole my page views?
Why does traditional media, including the online ones, struggle so much against user generated content (yes, as a matter of principle and whatever else you can hate the phrase for all you like) and non-traditional media generated content? Beyond the fact that most of the Youtubes are full of content produced by the traditional monsters (note to traditional media: your content discovery model is toast, find a new one or have one shoved down your throat with excessive force), the nagging little problem for the oldies is the cost structure. It is just way too expensive to produce the content they produce, while the average Joe with an N95 on the street might end up producing something that will blow you out of the water on any given day.
Once up on a time, the only way to produce and distribute content online was to pump in a lot of money. Everything would cost a fortune: the wire copies that are required to pump in the non-unique stories, the reporters who would create the unique content and the editors who would package and publish it all online. Web hosting was expensive, technical support was expensive, there was no Blogger, WordPress or even Django and to publish content online you would probably have to create and maintain a content management system of your own.
Before it became the favourite pond of sploggers, Blogger, along with a host of other publishing websites decimated the high entry cost to publishing, personal or otherwise. Of course, the quality of content on the average blog was not quite at par with something that came from a media house’s stable. Where they scored was in terms of width and variety. 200 bloggers on the same platform covers a much wider area than 20 of the best editors sitting together and the monopoly was broken.
All you need now to get published is a domain, a web hosting account and the ability to use a browser, totaling something less than $130 for a year. And if you go about it in a smart way, you can easily break even and even generate differing degrees of profit. This is why you have slew of new online publications that don’t do much of end-to-end original content. It is much easier and cheaper to let the big guys do the heavy lifting; they will just latch on to it and provide that little bit of insight and background, which is mostly not allowed in a normal reportage. There is really nothing wrong with that model, it is an opportunity that the traditional media model has brought about. Nobody should feel any shame in making a living off it.
The average mainstream online media publication (the ones that publish 24 hours or close to it) employs something in the region of 20 – 30 people in production to keep the show going. Mind you, that number does not guarantee any exclusive content; they can comfortably cover most of the day’s events, but creating exclusive content is an additional effort on top of that. Making things even more difficult for them are recent developments like Google’s refusal to index wire copies from publications anymore, while the agencies themselves charge you extra for online usage rights. It suddenly reduces the footprint these publications have on aggregators and search engines.
So, is it really that a Paidcontent or a Gawker Media will be the New York Times of 2010? I am afraid not. They do use a marginally leaner model of generating content, but they are still based on the flawed old model, which is too costly to run and it just does not scale. This is just an interim where there is a bit of arbitrage in terms of the cost vs revenue equation for the new guys, compared to the old ones, but it is not going to last forever and this is certainly not the future.
Technorati tags: new media, media, user generated content
WordPress.com makes a million, blogs that is
Congrats to the lads at Automattic on their anytime-now millionth sign up on WordPress.com. The company is the dark horse in the entire blog hosting business and they quietly go about doing their thing. And other than the long-forgotten misdemeanor of placing some icky ads on the WordPress homepage, Matt’s hardly put a foot wrong in recent times. And for a ‘virtual’ company (there is no real office these guys work from), they have accomplished some amazing feats:
1) Use cheap hardware and mostly open source software to deliver performance and reliability that far outstrips any expensive solution out there. The list of software used reads something like this: Debian/Ubuntu, PHP, MySQL, Litespeed, Pound, Wackamole, Spread, Nagios, Munin, Monit, NFS, Postfix, MyDNS. How often have you read about an unplanned outage on WordPress.com? You can get an idea of their set up at Barry’s blog.
2) Scaling out PHP and MySQL to support an operation of this scale. Okay, it is not quite the LAMP stack (Litespeed instead of Apache, read Matt’s comments here on the subject), but it does make the case that PHP, when done right can do incredibly well, both in terms of scale and performance, which is only augmented by this post by Steve Grimm on the Memcached list about Facebook’s architecture.
3) WordPress.com also has what could arguably be called the best user contributed text content repository online at this moment. There is such a wide variety of content — from technology to adult — that is of a pretty good quality. Excellent and pro-active policing of illegal content has also ensured that there is little spam within the network, a problem that is a growing problem on Blogspot.
With 7.5 million daily page views and 45 million unique visitors, they look more like an acquisition target by any of the big boys who would be drowning in a pool of drool of envy after seeing those numbers. Any guesses as to who they might be talking to/have already turned down?
In any case, way to go guys and keep the good stuff pouring in!