Re: [MacPerl] Re: Benchmarking

[Bruce Van Allen <bva@cruzio.com>]

At 12:26 AM 11/15/00, Scott R. Godin wrote:
>on 11/14/2000 10:37 PM, Bruce Van Allen at bva@cruzio.com wrote:
>>  I've lost the reason you're not using reverse. It seems to me that in

>because reverse simply will not do what I need it to do under the context
>I'm using it.

[snip]

>By doing this in advance, I can sort the output list the *same way* (neat
>trick, that) whether I am sorting by rating & map title or sorting by map
>title only, thus obviating the need for multiple blocks of code or custom
>sort routines -- simply by transforming the data to sort into a more
>specialized format. You'll see how it works... It's a very nice trick :>

Sounds like a variant of the Guttman-Rosler sort, which I've started 
using in my data-handling modules -- way cool.

>....but simply taking the reverse of $coercename wouldn't work at all. I
>don't need it backwards, I need every letter transformed into its
>alphanumeric opposite, but kept in the same order.

Ahhh.

>MacPerl cgi's choke on the output of the data however, as the size of this
>file seems to be too large for the built-in websharing control panel to
>handle (but only for the utdm section.. the rest work fine) (if anyone knows
>a fix or workaround for this, I'd be enormously grateful). ((and yes, I've
>tried increasing the memory partition of the cgi to over 150MB without
>success. argh.))

How about having your script write its output to a temp file and then 
returning an http "Location:" (redirect) rather than "Content-type:" 
mime type. You could give the temp file an id so it persists for your 
visitor. How frequently do you update your list of maps? Is that 
automated? You could clean out the temp files when you do that, or 
just have the CGI delete (unlink) any temp file more than 24 hrs 
old...

Cheers
- Bruce

__Bruce_Van_Allen___Santa_Cruz_CA__

# ===== Want to unsubscribe from this list?
# ===== Send mail with body "unsubscribe" to macperl-request@macperl.org