.jpg Properties In Linux Vs Windows - Puzzle Caches!

[+suemac]

I've worked on several puzzle caches lately that have coord info hidden in the properties of various .jpg files on the cache page (or elsewhere). You have to save the image to your machine and then view the image properties to find the puzzle info. My problem is this: I've been confused 'cause I sometimes find info in an image that I was sure I'd already checked and didn't find anything in.

 

I figured out this evening that when I'm on my Windows box everything works fine. I can use Windows explorer to view image properties and I see several fields that sometimes hold puzzle info, such as the 'subject' and 'comments' fields. However, when I view the image properties for those same images from my Linux box, I don't see those fields listed, thus can't see puzzle info if it's kept there. I also don't see the info if I just display the file contents on Linux.

 

Has anyone else noticed this behavior or know why it happens?

 

thanks,

suemac

[+ejnewman]

It will depend on which program you are using. Apparently the linux image viewer you are using can't read those extra fields. Try a different image viewer in linux.

[+Mary&Dave]

Try using Gimp on your Linux machine to view the photos. Also, what sometimes works is to view the image in a web browser and select view source. You'll get a lot of garbage, but the parts in plain text (such as coordinates) will still show up that way. They'll generally be near the start of the document.

[+Vinny & Sue Team]

Try using Gimp on your Linux machine to view the photos. Also, what sometimes works is to view the image in a web browser and select view source. You'll get a lot of garbage, but the parts in plain text (such as coordinates) will still show up that way. They'll generally be near the start of the document.

 

I agree with this analysis and the recommendations. And, you can also download the file to your PC, and then use an HTML editing tool, such as NVU (which is available in versions for Linux, Mac and Windows OS) to open the file and then click on the tab which allows you to view source code. I feel that you have a bit more control along with less sloppiness in the display format when using NVU, versus using the "view source code" feature in most browsers. However, if you must us a web browser, well, I use only "view source" for a true unexpurgated look at the source code.

[+ejnewman]

They wanted to look at the EXIF fields in an image file though, not anything in the html source. Doing a "view source" in a web browser or html editor won't do that.

 

Just try different image viewers in linux until you find one that lets you see that stuff.

[+Munin]

To search for plaintext embedded in an image, might I also suggest something like:

 

strings foo.jpg | more

 

(Man, I feel so old! Kids these days, with their fancy-pants "browsers" and "gimps" - bah! )

 

For extracting EXIF info without resorting to such a brute-force approach, jhead should do the trick - it's been ported to pretty much all the distros, so check your package manager, RPMFind, or just grab the source from the jhead site.

 

(Okay, okay, I actually use gimp quite frequently. I might even use a browser from time to time. A real live graphical one! )

[+suemac]

Ok, ejnewman was correct in that the info wasn't in the html source, but was in the EXIF data fields. I tried several different image viewers unsuccessfully; I haven't tried GIMP yet 'cause I don't have it installed after a dist change to Gentoo, but will install and try it out.

 

Turns out the easiest option of all works fine -

more foo.jpg

 

For some reason I didn't see the same info when I used Munin's strings foo.jpg | more, though I'm not sure why. I think I'll have to do some more prodding at the file to see what else I can determine.

 

Thanks for the help. I think my basic problem was that I didn't realized that I wasn't seeing all of the EXIF data using my Linux file manager. Totally messin' with my head !!

[+Vinny & Sue Team]

They wanted to look at the EXIF fields in an image file though, not anything in the html source. Doing a "view source" in a web browser or html editor won't do that.

 

Just try different image viewers in linux until you find one that lets you see that stuff.

 

Aha! Got it now! Thanks!

[+Munin]

Okay, so I took a look at one of the problematic images, and without going into much detail I'll just toss out a couple of additional Linux-oriented tools that might be handy for banging away on images:

 

The first is another brute-force/Neanderthal technique - just dump the data and see what you get:

 

hexdump -C foo.img | more

 

Unlike "strings", which might omit useful information if it's mixed in with non-printable characters, hexdump just shows everything in a format that's at least semi-readable by humans and lets you try to make heads or tails of it.

 

The other handy-dandy utility would be ExifTool, which understands a wider range of tags than most other Linux image programs (including a number of vendor-specific tags), and appears to be available in many mainstream Linux distros. It can be useful if you need to see info contained in one or more of these vendor-specific EXIF tags. (Say, just as a totally hypothetical example, a tag added by Microsoft that displays wonderfully under Windows but isn't understood by most Linux-based EXIF utilities. )