[prev in list] [next in list] [prev in thread] [next in thread] 

List:       xml-dev
Subject:    Re: [xml-dev] Do long element names impact performance?
From:       Shaun McCance <shaunm () gnome ! org>
Date:       2015-10-15 15:27:39
Message-ID: 1444922859.2239.255.camel () recto
[Download RAW message or body]

On Thu, 2015-10-15 at 15:53 +0100, Michael Kay wrote:
> > Comparing two strings for equality is linear time on the length of the
> > strings.
> 
> Not necessarily. If most of the comparisons return false, it may be a
> lot better than linear. It's likely in many cases that the comparison
> will start by comparing the string lengths, so you might get the
> fastest comparison on average by having the lengths highly variable.

Linear time is the upper bound, which is generally how algorithms are
described. Whether comparing string lengths is a good idea depends on
your string implementation. I'm a C programmer. For me, a string is a
NULL-terminated char array. It costs time to compute a length, so I
wouldn't bother.

> But who says it's string comparison that dominates? It might be the
> effect on network latency, or the cost of doing compression. You need
> to make measurements to find out.

I agree 100% with this statement. Using single character element names
is an absurd premature optimization. Always measure first.

--
Shaun



_______________________________________________________________________

XML-DEV is a publicly archived, unmoderated list hosted by OASIS
to support XML implementation and development. To minimize
spam in the archives, you must subscribe before posting.

[Un]Subscribe/change address: http://www.oasis-open.org/mlmanage/
Or unsubscribe: xml-dev-unsubscribe@lists.xml.org
subscribe: xml-dev-subscribe@lists.xml.org
List archive: http://lists.xml.org/archives/xml-dev/
List Guidelines: http://www.oasis-open.org/maillists/guidelines.php

[prev in list] [next in list] [prev in thread] [next in thread]