Wednesday, September 29, 2004

30 Years from now

Interesting article - http://www.builderau.com.au/webdev/sitedesign/0,39024698,39129283,00.htm

Takes guts to predict anything more than 5 years in IT field - IF you want to be serious about it that is.... considering the miserable track record of similar predictions ! (remember Bill Gates , Thomas J. Watson , etc ?!)
Some nice facts in the article : puts current technologies into perspective - from a evolutionary point of view - I wonder what it would be like 10 years from now !!


-----
*UPDATE* (Primarily triggered by Manoj's comment :-D )

Moore's law - I never believed much in it , but you must admit that it has been pretty consistent !
Also - w.r.t possible going down the drain - remember the tech jump from vacuum tubes to semiconductors ??
Compare ENIAC, to say your laptop - or even the MAC from 20 years back :P

If you go back 30 years and tell a guy then that desktops will be running at 4 gig 30 years from then - (well first explain to him what a desktop is ;) ) now you can appreciate the tech growth !
Heck , 10- years from now Rao's latest PDA would be much faster than this comp of mine !

What I liked were his article were the broad set of thrusts at the future - indexing , places were mem/proc power is going to be utilised , bandwidth , etc.

I know of people who reorg code and optimise like crazy in assembly for past 20+ years - while I code in C and now java.
Notice all the mem requirements going up and processor power leaking away ?

If you give the current desktop to a guy from 25 yeas back , his programs will be blazingly fast - since his whole attitude towards coding is different : CPU is expensive and precious resource and memory is like a godsend boon not to be used for trivial purposes !
And then we see memory leaks and OOM in current apps - sick.

This interview http://www.builderau.com.au/webdev/0,39024680,39130602,00.htm by the same guy got into lot of rough weather - though his salient point wrt open source interfaces is quiet valid , some of his others views I cannot agree with.

**UPDATE**

Was surprised to know that lot of people did not know of this article : http://www.wired.com/wired/archive/8.04/joy.html

0 Comments:

Post a Comment

<< Home

Ideas flicked !!

One of the ideas that we were toying with has been 'flicked' :) Read more here.
This was one half of what we wanted to do , the other half is , thankfully , not in their stated plans !
Call the VC's fast :)

Well to be fair , we wont have the time and energy to do this on our own ;)

0 Comments:

Post a Comment

<< Home

Wednesday, September 01, 2004

Solaris 10

Will try to do a series of blogs on my findings of solaris 10.
This is hot hot hot people !!!!
And I am going to do it topic by topic - taking upto (and maybe more !) than a week for what I can find about each 'feature'.
Though I am pretty much sure I am not going to be even decently as comprehensive as the kernal engineers from Sun :) - so keep checking out http://blogs.sun.com for their blog entries too !!

Topic for this week - ZFS : should be done by friday !

In case you want to choose - please leave comments on topics for next week : choose from zones , dtrace , N1 - choose more from here (or other places where you can get more info from :) )

-- Update ---

Sorry , cant post much about zfs - will have to wait for official release.
Seems like , it is still not released to external world and so no doc other than the released docs are supposed to be out ....
Some relevent externally accessible links I could find are these : a zfs kernel engineers blog !, overview preso.
Have fun - but only thing I can say from my readings from internal sites is - it is just too good to be true !
If it even works upto 10% of what they say , it will simply blow everything else away (features , storage capacity , administratability , etc) !!!

0 Comments:

Post a Comment

<< Home