As we muttered about caching, at 09:38 2002-10-08 +0100, Nicholas Clark wrote:
>The actual formatting is the slow part, isn't it? Not the finding or 
>reading of the pod source file?

As far as I've seen, turning the pod into whatever output format is the 
"slow" part, not the part of rendering that output format to 
screen.  Incidentally, perltoc is what I typically pull out to try 
parse/render times, because it's so much bigger and weirder than anything 
else.  On the other hand, ya know, as processors get faster, for Pod of 
typical length and complexity, it's getting so that pod2whatever-ing AND 
the formatting take less than 1 second altogether.

So I guess I can just say I'm not going to bother implementing caching in 
perldoc because a few dozen milliseconds of time saved isn't worth the 
hassle of the cache system acting crazy for some people.  I suppose I can 
say if people want zippy caching, they can just periodically manify/htmlify 
everything and read everything thru man or their browser.
Also, people have gotten along fine so far without perldoc caching, so it's 
not like they'd /miss it/.

--
Sean M. Burke    http://search.cpan.org/author/sburke/

Reply via email to