Re: 90/10 rule... source?

Derek M Jones <derek@knosof.co.uk>
31 Jan 2004 00:52:30 -0500

          From comp.compilers

Related articles
[3 earlier articles]
Re: 90/10 rule... source? nkavv@skiathos.physics.auth.gr (2004-01-16)
Re: 90/10 rule... source? gah@ugcs.caltech.edu (glen herrmannsfeldt) (2004-01-17)
Re: 90/10 rule... source? derek@knosof.co.uk (Derek M Jones) (2004-01-17)
Re: 90/10 rule... source? jcownie@etnus.com (James Cownie) (2004-01-22)
Re: 90/10 rule... source? vidar@hokstad.name (2004-01-22)
Re: 90/10 rule... source? ieuk001@attglobal.net (2004-01-22)
Re: 90/10 rule... source? derek@knosof.co.uk (Derek M Jones) (2004-01-31)
| List of all articles for this month |

From: Derek M Jones <derek@knosof.co.uk>
Newsgroups: comp.compilers
Date: 31 Jan 2004 00:52:30 -0500
Organization: Knowledge Software
References: 04-01-038 04-01-059 04-01-077 04-01-108 04-01-124
Keywords: practice
Posted-Date: 31 Jan 2004 00:52:30 EST

Vidar,


>> > Yes and no. I think that the rule is real, but it has been made into
>> > a pseudo law of nature, which it isn't.
>>
>> What is a pseudo law? One that only applies to a small percentage of
>> cases?
>
> The 90/10 rule, like many others such as for instance Moore's law, are


Moore's law has the benefit of facts to back it up. So for the
time being it is true.


> In this case, it is generally the case that a small part of a program
> will turn out to take up a significant part of the execution time of a


My point is that it is not generally the case only that people believe
it to be the case. Where is the experimental evidence? Yes there are
some programs that follow a 90/10, but there are a lot that do not.


I think a 10/90/10 rule (10% of programs spend 90% of their time in
10% of the code) would be more appropriate, while you appear to
believe in a 90/10/90 rule (90% of programs spend 10% of their time in
90% of the of the code).


There must be some MSc. student out there looking for a project. You
will need to select a representative sample of programs (picking SPEC
is the easy option, but you might have you work cut out showing it is
representative).


Of course facts proving the opposite (if that turns out to be the
case, which I think it will) does not appear to be any more hinderance
than no facts at all, for people to believe urban legends (would the
existence of some programs following the 90/10 'rule' disqualify it as
an urban legend?).


http://www.knosof.co.uk/cbook/misart.pdf


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.