[jdom-interest] Internal DTD subset verification
Dennis Sosnoski
dms at sosnoski.com
Wed May 8 20:02:42 PDT 2002
Elliotte Rusty Harold wrote:
> Your tests haven't convinced me. They're a lot of problems with them..
Some specifics here would be really useful. Otherwise it just comes
across as carping...
> In other words, your tests deliberately do not include the cost of
> I/O, which makes sense for what you're doing because I/O would indeed
> swamp what you're trying to test. However, the fact is there's not a
> huge amount of point to us optimizing input that's going to be swamped
> by I/O in any real work scenario.
The issue is not that I/O would "swamp" what I'm testing, but rather
that it's irrelevant. The I/O overhead is going to be the same
regardless of what model is used. Is it greater than the build time?
From a quick sanity check (wc the full set of mid and large xml files,
from disk not in cache) it looks like reading the files is at least as
fast as building the JDOM representations on my system. That's obviously
going to depend a lot on how you're getting the documents, though -
which is the other reason I don't want it as a factor in my tests.
> How are you actually measuring memory usage? I did not find any
> details on your site. Based on the following, it's not obvious to me
> that you're getting accurate counts:
>
> Testing the memory usage of the representations works a little
> differently, in that the program keeps all the constructed copies of
> the document and pauses between relevant tests to encourage garbage
> collection. Memory usage per copy of the representation is found by
> dividing the total memory used by the number of copies.
The memory measurement issue has been discussed previously on the JDOM
list. See
http://www.servlets.com/archive/servlet/ReadMsg?msgId=149992&listName=jdom-interest
for a detailed response to Jason on this issue (and the follow-up at
http://www.servlets.com/archive/servlet/ReadMsg?msgId=149993&listName=jdom-interest
for why the size actually increased when the JDOM representation was
accessed).
- Dennis
More information about the jdom-interest
mailing list