You are correct. I made the mistake of saying it was Bycast when I actually meant Spinnaker. Thanks for pointing it out. I will have this corrected in the blog.

That is a nice perspective.

I didn’t mention other smart things being done like Parallel NFS. Maybe that is a subject for another post. But yes, we need to do smart things to both store data and also to retrieve business relevant information from the stored data. Smartness will never go out of fashion for sure!!!

Just like bandwidth, storage is going to be another fundamental dimension.
(1) MIPS (2) Memory/storage (3) Bandwidth
Each application would have to make judicious design decisions based on the above 3.
Hence storage play would be very critical.

Database demigod Jim Gray has offered many approaches to this nagging problem (http://en.wikipedia.org/wiki/Jim_Gray_(computer_scientist)).

Net net, we are moving from an era of math models to data driven models which runs to many peta byes – in which case, you don’t need models – correlation would be enough. So, smart and flexible storage would be welcome in the near future.

cheers,madhu

What you say is very correct. There are only a certain number of ways you can speed things up and in an abstract sense most domains use similar concepts. And you correctly said it, there will never be ‘one shoe fits all’ kind of solution. Most of the companies would love to get ‘one shoe fits lot of people’ kind of solution. Therefore at any point in time there are going to be niche players who would be providing optimal solution for a particular problem. That way if you understand some of the important problems and have the capability of providing a solution, you have a good chance of winning in the market.

So, one fundamental principle seems to be to not try a one size fits all approach and customize to a reasonable sub-class of problems..

Best Regards,
N.R. Ramesh.

It does look like we are from a different planet when we talk about good programmers, isn’t it

What you express is something which lot of senior folks in India and other places express routinely. My friends like Ramesh, who had commented here, are trying their best to work with various universities and rectify this issue. As you very rightly point out, the prestige of being a ‘manager’ outweighs the joy of writing good programs for many people. The situation is definitely different in Canada / US / Europe, where the ‘manager’ tag doesn’t carry too much weight. Infact a contact of mine, who was once my client, has been doing just bug fixing for the last 25 yrs. There is no way such deep knowledge can be replaced with people with couple of years experience but it routinely happens.

I don’t think you are being skeptical. If we need to correct a situation, we need to first accept that such a situation exists.

My understanding of the great Indian programmer is that, by the time they really get a hang of their first computing language, they have already become a manager!

He is now into a different sort of pipelining If he is not a programmer doing something useful (the only true computing profession is a programmer – everything else is frills) he calls himself an architect. I spoke to one of my programmers who used to work for me some 15 years ago recently. He told me that he spends most of his time drawing pictures in Visio – that’s life of an architect!

The programming position is still alive in the West. Being a manager does not have the social status here as it is in India. Unfortunately, such committed souls are getting replaced quickly by Indian programmers who have spent less than 2 years as full-time programmers. Most Indian programmers, working in India I have worked with, do not have a deep understanding of data structures and programming discipline though they all claim to have a CS degree.

Maybe I am too skeptical. The only true product that has come out recently from India is the zoho products written by high school passed guys. I think that is the best model. Get talent out of high school and resurrect the pride of being a programmer. Forget the engg kids from any institution.

My 2 cents.

Ravi Natarajan

I definitely expected you to comment since you have straddled both the worlds. As you rightly say, it is imperative for many programmers to understand what happens at a low level.

Do borrow the book from me, read it and let me know you views. As you said, if you find it good we should go ahead and suggest to colleges to use this as a reference.