<br><div class="gmail_quote">On Fri, Dec 10, 2010 at 5:08 PM, Heiko Oberdiek <span dir="ltr"><<a href="mailto:heiko.oberdiek@googlemail.com">heiko.oberdiek@googlemail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im">On Fri, Dec 10, 2010 at 03:10:05PM -0500, Peter Davis wrote:<br>
<br>
> Thank you, Heiko. I agree that having both JFIF and EXIF data is<br>
> over-specifying the information, and creates opportunities for such<br>
> inconsistencies. In fact, files with such inconsistencies are *very*<br>
> common. I've run into this situation, and had to code around it, many times<br>
> over the years.<br>
><br>
> There should be some consensus on which overrides which in cases where the<br>
> image is inconsistent.<br>
<br>
</div>I consider it as a bug of the application that creates the image file.<br>
Why should one case more "bug free" than the other?<br></blockquote><div><br></div><div>You are, of course, free to take that position. However, in a very large number of cases, with a very large number of files, users will not get what they want or expect. The Chrysanthemum.jpg image comes from the Corbis library, a pretty large stock house, I think. I think Photoshop and other widely used apps commonly write files like this. So you can either "be right," or "do the right thing." I think it's like saying "HTML is wrong because it allows style information like <b> and <i> to be set explicitly. Therefore, my browser will not support HTML."</div>
<div><br></div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im"><br>
> It seems that much of the more widely used software<br>
> (Photoshop, InDesign, Windows, pdfTeX, the IJG JPEG library, etc.) use the<br>
> JFIF data, while ImageMagick and XeTeX use the EXIF data.<br>
<br>
</div>JFIF data are very easy to parse, but very few information is provided.<br>
EXIF data needs more effort for parsing, but can provide a variety<br>
of information. File formats with redundancy involved have the risks<br>
of inconsistency. Fix the inconsistencies and the file reading<br>
applications are happy.<br></blockquote><div><br></div><div>In the printing and publishing world, at least in my experience, Photoshop is *the* de facto standard image manipulation tool. Any file that comes from Photoshop is, by definition, "correct."</div>
<div><br></div><div>-pd</div><div><br></div></div>