[kepler-dev] The next largest...Re: Memory leak.

[Matthew Brooke]

Kevin

I already had Photoshop fired up for some other task, so I took a quick 
look at those images, and just by  re-optimizing them, it was possible 
to reduce the file sizes (significantly in most cases). The sizes 
"before" weren't exactly huge, but it seems from your investigations 
that Java has some memory overhead associated with displaying these 
images, so it will be interesting to see if this helps (I wonder if the 
overhead is constant or proportional to filesize?). Here are the before 
an after stats:

                       was     now
EOL.jpg               38k     11k
geon_logo2.jpg        40k     9k
kepler-logo2.png      16k     6k
ptolemy-100.jpg       6k      2k
roadnet-icon.jpg      52k     18k
seeklogo.png          6k      4k

I reduced file sizes without reducing visual quality; I didn't change 
any dimensions, since i didn't know if any images were used elsewhere.

m



Kevin Ruland wrote:
> The next largest items which I have questions about are 3 or more
> int[].  They are each >800k and I believe they are the pretty graphic
> images which appear in the root window (is there a name for this
> thing?).  The stack dumps are truncated so I only know they are somehow
> related to JPEG images. I commented out the big table of .jpgs from
> intro.html and they went away.
> 
> I don't know if they are really leaks since the graphics must be
> displayed.  I did check the filesizes and they are all <50k or so. 
> Except for Roadnet and EOL, they are all pretty close to the rendered
> size of 100x50.  The interesting thing, a 24bit 100x50 image is only 15k
> uncompressed so I'm confused why java appears to allocate so much space.
> 
> I don't know where the html is rendered so I can't tell if there is
> something wacky there.
> 
> Kevin
> _______________________________________________
> Kepler-dev mailing list
> Kepler-dev at ecoinformatics.org
> http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-dev

-- 

---------------------------------------------
Matthew Brooke, Ph.D.
Marine Sciences Research Building, Room #3407
University of California
Santa Barbara, CA  93106-6150
ph: (805) 893-7108   fx: 805-893-8062
brooke at nceas.ucsb.edu
---------------------------------------------