Nov 8, 2007

OpenSocial disenchants

I am not really a keen user of socaial networking sites, but I believe that they intrinsically carry a lot of potential to alter/compliment most business models on the web.Like I mentioned in an earlier post, the biggest challenge right now is that social graphs are locked inside sites that own them and thus expanding/harnessing them across these various sites is not a possibility.
I was actually hoping OpenSocial to solve this problem but was disappointed to find out that it is nothing more than a means for developers to build apps which are cross-portable on various social networking sites.
So I guess we are still a long way from USN (Unified social network) or to and fro data pollination on the social graphs.

Nov 5, 2007

Code review challenges

Have tried to capture and compile here the objective of code reviews and the approach for fulfilling the same in a tangible and recordable fashion. Apart from other concomitant benefits of reviewing code like minimizing defects, the primary benefits would easily be the following :

* verify that the code is production quality / release-worthy
* ensure that the way code written by different people follows a standard skeleton
* ensure that each individual programmer continually writes better quality and more maintainable code

But it's not easy - because of the inconspicuous yet overwhelming existence of the programmer's ego or Cognitive dissonance.

I came across an interesting article here that claims to take away the pain from the code review process. And probably a simple checklist like this could make the process more structured.

The other challenge is to quantify change, of course on the +ve side of the number line .. I guess the only way to do that is break down the checklist into much more granular components. For instance, capture design, approach (creative, fast, etc) , unit testing, comments, performance, legibility, maintainability, version control, etc and the degree of all these and have either a boolean or scale rating for each aspect and then progressively track an individual's ratings to see if the methodology needs to be tweaked.

Nov 4, 2007

Google signs on as MySQL contributor

The search giant is spreading it's wings far and wide.MySQL AB announced a few days back that the leading open source database will include code contributed by Google inc., along with some security improvements in the development roadmap through 2009.
That's great news. Hopefully MySQL users world over would get to derive the benefits of several customizations Google has been doing on the db engine.Albeit, the extent and areas of contribution from Google would be interesting because if this move is more strategic than philanthropic, then Google obviously would be choosy about releasing contributions to the MySQL code base rather than opening the flood gates.
Also, why is it that the company has chosen to begin contributing to MySQL only and not any other platform that the company uses and customizes to a great degree for better than optimal performance, like Linux or Apache, etc.Why did the company chose to partner with IBM to train university students on cloud computing ?
MySQL had also announced earlier this year that the enterprise source code would not be available for non-clients.So one is forced to question whether contributions made by Google would also be split amongst the two versions ?

Also, interestingly is there a paradigm-shift in the open source world where in contributions from individuals begin getting overshadowed by those from the big guns ?

Nov 3, 2007

Social graph and beyond

Taking the previous thread forward, presuming we can solve the unique yet common identity problem, how can I use and benefit from my social graph apart from just staying in touch with people and sharing community/work specific knowledge/information.
I reckonOpen Social a project undertaken by Google that involves building "a set of common APIs for building social applications across the web", could be the next big thing.Though the description does not reveal much except that it would enable social networking sites to allow influx and outgo of user and graph centric data.This move makes the possibility of a USN (unified social network) if i may call it,coming into existence look real.Moreover, since data flows two ways on graphs, this could mean that one could relate any piece of data/information on the web to one's social graph.Six degrees !

Oct 29, 2007

Ubiquity of the social graph

Networks - professional or otherwise that started off as fads have the possibility of evolving into very potent tools for the internet and it's users. The metamorphosis that some of the networking sites have gone through could not have been planned (people might claim so !).
If you ask me, the idea of social / professional networks must have originated from the old Web 1.0 (if you please) concepts of discussion boards, mailing lists and communities .. sounds more like an evolution path to me.
Well then, what next ? Shall address that in a bit, right now let's understand the current situation : there are plenty of social networking sites right now being used extensively / sparingly depending on the country you are in.Which is a problem, because if as a user you have to expand / cross-pollinate your networks amongst these various sites, you can not do so unless you register with all of them or ask people you'd like to network with, to move in on one of these sites that you are at. Any which way, you would have to maintain different networks across different sites and these networks exist in a Euclidian paradigm.
So, the solution is simple, have an unique identity for each user that all these sites recognize.Well, it actually is not as simple as it sounds.Although Facebook opening up it's network through apps and MySpace vowing to follow suite, the networks may be accessed but they would remain to be locked-in aat these individual sites.
And as a user, I don't like that.
I'd love for my graph to be ubiquitous,accessible and extensible with ease, independent of the platform (read a specific site).