Mailing List Archive

Support open source code!


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: value of time [was: system questions]



re:

> Has anybody done any benchmarks on human-relevant multitasking?  What

> I mean is this.
...

Probably not enough.  However, I do recall an article by Peter Norton wrote 
in PC Week or InfoWorld about 10 years ago rationalizing why he upgraded his 
secretaries from regular 8086 pc's to the latest expensive 80286 AT's, even 
though all they did was text entry.

When one notices a perceived difference in reaction time from a machine, 
however slight, production increases in my opinion, and time is money, and I 
like to make money.

I've done all the tweaks you mention, and I still want a 200MHz Pentium 
Plus.  I've programmed for a living also, and I could never imagine settling 
for a 2 hour compile if a faster system could do it in 10 minutes.  Time is 
Money.  Faster is more Money.  Faster is Better.

Ted

*****
Ted Matsumura, Adaptec Japan Ltd. 
ATM Program Manager, InterNetworking Technology (INTO)
phone: 03-5276-8433, Fax (03) 5276-9364 
email: tmatsumu@example.com
http://www.rahul.net/tedm
*****
-------------
Original Text
>From owner-tlug@example.com, on 1/10/96 7:51 PM:
>>>>> "Kise" == Norihide Kise <s100234@example.com> writes:

    Kise> Ken,

    >> You mentioned previously that you don't mind something that is
    >> a bit slower as you have the time. Have you ever put a
    >> dollar-sign value on your time now or what you would like your
    >> time to be worth? By doing so, you might realize that by paying
    >> a bit more, you'll be saving quite a bit of time and the nicer
    >> system might just pay for itself?

I agree with Ken that time management is very important.  I can't see
P90 vs P75 as a crucial issue though.

Investing in a faster machine is generally far less relevant than
reconfiguring your current environment, I've found.  Eg, adding 4DOS
to my 386/33 with a 25ms HDD for $49 made a much bigger difference in
my performance than borrowing the next guy's 486/33 with a 17ms HDD
back in '91.  Switching to Linux + bash + Mule + (real) X on my
current (486/50MHz) system made a much bigger difference than moving
my former DOS + 4DOS + OEmacs + NEmacs + DESQview/X to a colleague's
(P66) system.  But these are time, not $$, investments.

Better algorithms (ie, buy software instead of free/shareware in some
cases) and well-tuned configuration is usually a better investment
than faster CPU or HDD.  (Just reviewing my .procmailrc and .fvwmrc
once a week or so and tweaking them makes a noticeable difference.)

On the other hand, getting hardware you don't already have (a CD-ROM
drive, eg) often makes a big difference.  And improved communications
capability (faster modem, getting linked to an Ethernet, etc) has
always paid for itself in my experience.  I think Jim T is probably
right about monitor real estate if you're using a windowing system.
 
    Kise> Maybe it might be true if your job is very crusial about
    Kise> time, such as stock brokers. However, I'm just a plain
    Kise> student and don't make much money, that's why I said I don't
    Kise> mind getting a bit shower machine (P-75). Please note: if I
    Kise> were not a student and had decent income, I would buy a
    Kise> powerful machine (P-100), but since I'm a poor student,
    Kise> somehow I have to compromise.

Has anybody done any benchmarks on human-relevant multitasking?  What
I mean is this.  The benchmarks say that the P90 is faster than the
P75.  OK.  But is it really worth doing that rather than buying extra
memory?  I rarely get upset about compile times and the like, even
when I'm building Linux or XFree86.  I just run it in the background
on my 2.5-year-old 486/50MHz.  So what if it takes a couple of hours?
Even when I'm developing software, and recompiling/relinking fairly
often, it's not a big deal with a makefile (if the rebuild takes more
than 30 seconds, I use the time to document the change I made, even if
it doesn't work---it's usually just as important to know what didn't
work).

But ... I often find myself using well over 32MB of virtual memory
when I've got Mule, Canna, Netscape, XFree86, the kernel, and maybe a
couple of XTerms running, as well as the HTTPd and so on.  My feeling
is that for a person who is using the machine for anything except
serious iterative math calculations (eg, raytracing graphics for your
billiards simulation :-), it is memory, not processor performance,
that is the bottleneck.
 
    >> You might want to check out some books (or tapes) on Time
    >> Management, for example Time Power by Charles Hobbes or 'self-
    >> improvement' type of books that focus on improving one's
    >> finances, like Charles Givens' 'SuperSelf' or 'Financial
    >> Self-Defense'.

    Kise> Sorry, I've never read any book on the subject. I know I
    Kise> should read books on the subject to optmize my use of time,
    Kise> but I'm too busy doing other things at school and at work.

The main thing as far as time management in my *personal* experience
is learning to say "No".  Especially when they say "it'll only take a
minute".  Faster computers are not very relevant unless you are doing
massive repetitive calculations (like the "rocket scientist" stock
analyst that Kise mentioned).

Once again, all this is IMHO.

-- 
                            Stephen J. Turnbull
Institute of Socio-Economic Planning                         Yaseppochi-Gumi
University of Tsukuba                      http://turnbull.sk.tsukuba.ac.jp/
Tennodai 1-1-1, Tsukuba, 305 JAPAN                 turnbull@example.com


Home | Main Index | Thread Index

Home Page Mailing List Linux and Japan TLUG Members Links