[Bioclusters] O'Reilly Clusters Book Review

[Tim Cutts]

On 25 Feb 2005, at 5:43 pm, James Cuff wrote:

> I think we should be a bit more proactive here.  It's a real small 
> world.  There was a mention of a 'bioclusters book' on this list a 
> while ago... maybe we should lead by example?  I'd be happy throw in a 
> chapter...

I'm not sure the market is big enough.  And anyway, I think the 
principles are fairly generic.  I used to think bioinformatics tasks 
were somehow different, and that bioinformatics clusters had different 
rules.  The more I talk to other cluster managers around the world, the 
less I think that is true.  There are plenty of other disciplines with 
similarly embarrassingly parallel, high data/cpu ratio problems, and 
they build clusters in the same way.

> However, all I know about writing books is that takes for ever and 
> ever, it's really hard work, big fat personal egos get in the way of 
> good content, most of it is out of date by the time you finish, and in 
> the end no one ever seems to thank you for it.

How many of us actually really learn anything from "how-to" style 
books?  I don't.  I get that stuff from the web, mailing lists, and 
talking to people.  I generally only find O'Reilly books useful if they 
are a more "reference style".

Tim

-- 
Dr Tim Cutts
Informatics Systems Group, Wellcome Trust Sanger Institute
GPG: 1024D/E3134233 FE3D 6C73 BBD6 726A A3F5  860B 3CDD 3F56 E313 4233