MAMA's Techno-Geek Thread

[GadgetGav]

This is really Techno and geeky isn't it...

[Actual Size]

Is this better?

[bakerjc]

Want to touch my monkey?

[snid]

Quote: Originally Posted by snid

I've gotta pay more attention to what I'm doing when I create jpgs for my webpage...

Sometimes, I is an idiot. Or: Damn PCs!

At some point in the recent past, my work laptop decided to switch itself from 32-bit color to 16-bit color. Most likely when I un-docked it and had to tell it that the second monitor was no longer attached, then re-dock it and tell it the second monitor was back. I never switched the bit depth myself, but it changed. Thats what made me think my jpg was over-compressed. I was seeing ugly bands of different colors in the image. When I noticed some other images looking funky, I found the real problem.

At least I know I'm not going completely insane.

[jwardell]

Quote: Originally Posted by snid

Josh is very slightly famous?

http://www.engadget.com/entry/1234000850027358/

I will withold my comments about the japanese culture and just say go watch Lost in Translation.

[Actual Size]

Quote: Originally Posted by bakerjc

Want to touch my monkey?

No...but...

[snid]

Not quite up to Monkey-boy Ballmer, but...

http://blog.monkeymethods.org/2005/0...teen-beat.html

[jwardell]

Hah! I'm sure he picked up all the ladies by throwing floppies at them.

[snid]

Serious geek post alert! Avert your eyes if you are easily frightened!

I've been wasting time at work the last few days. We're really close to a code release to customers, so making changes is a bad thing. Instead, you wait around for someone to say "something broke!" then rush to fix it. Nothing for me to fix recently, so I've been playing.

I'm still working on web stuff for my image gallery / home page with the thought of having as little work needed as possible in the publishing phase. Right now, all I need to do is throw a jpg in the right directory with the right filename format, and the rest is handled by the JavaScript in the index.html page. I like it that way.

But, I was thining, it would be cool to have some of the EXIF data available (shutter speed, aperture, focal length). "Hmmm", I think, "that info is in the jpg, and the browser has the jpg, so in theory the browser already has the information... how to get at it?" Well, an image loaded by an img tag is a black hole... no cool JavaScript methods to get anywhere with that. But, I've been playing with XMLHTTPRequest objects recently, which allow you to grab any web accesible file you want, like the jpg. Sounded good. I found the EXIF spec and some sample code on how to pull info out of the file, and started away.

Here's the problem... the closest I can get to the binary information in the jpg is a String representation. UTF8 should save me... but, it doesn't. I've found all kinds of wierdness depending on the browser and platform. I simplified things and threw away the EXIF code for now, and instead focused on just generating a hex dump of the jpg file (or the first kb or so of it). No good news. On the PC / firefox, all the bytes are there, but a bunch only contain the UTF8 "unavailable" character FE (or FFFE, or 65533 - yes, that's not a byte, more confusion). On the Mac / safari, I think almost all (or even all) the "bytes" I found came back as the right values, but there were some bytes missing! I think all the missing bytes were 00, but, when you're dealing with offsets to find the right byte of information, a missing byte is a problem.

IE just keeps yelling "undefined" at me.

current test code / hex dump is here http://homepage.mac.com/jtross/exif.html
the jpg it's trying to dump is here http://homepage.mac.com/jtross/photoblog/1.17.2005.jpg
if you want to play / compare for yourselves. Unix hexdump is the standard. "--" in the center part of the dump represents an "unknown" byte. "-" in the right part of the dump represents a byte that's not in the normal ASCII printable range.

So, I guess having the browser use JavaScript to extract the EXIF info isn't going to work.

Next idea, which will have to wait until I get home, is to use an AppleScript folder action to do the EXIF extraction for me and create an XML file with the data for each jpg I add. Then, use an XMLHTTPRequest to get the contents of the XML file, and JavaScript to display it.

[GadgetGav]

Well, I thought I was a geek, but I think I have to turn in my badge... I read it all and I am certain I can be of no help whatsoever..!

On a slightly lower plane of geekiness - why doesn't Apple's new Pages app in iWork publish HTML pages..? Wouldn't that make sense..? When is Apple going to make an easy to use, OS X front end for web page design over and above the .mac templates..??? Do you think that's for iWork2..?

[snid]

Quote: Originally Posted by GadgetGav

Well, I thought I was a geek, but I think I have to turn in my badge... I read it all and I am certain I can be of no help whatsoever..!

Yeah, but ask me about analog circuits, and I'm useless.

I'm probably pretty useless with digital circuits, too, since it's been over 10 years since I worked with them.

[Actual Size]

Jason...have you ever taken a look at Jalbum ? You can take an entire directory of images, dump them in via drag & drop, specify an output directory, and voila...you get a web photo album with the EXIF data for each image (if you use the right template). Yes...it's available for the Mac.

[snid]

Quote: Originally Posted by Actual Size

Jason...have you ever taken a look at Jalbum ? You can take an entire directory of images, dump them in via drag & drop, specify an output directory, and voila...you get a web photo album with the EXIF data for each image (if you use the right template).

Where would the fun be in that?

[Cape Cod MINI]

Quote: Originally Posted by snid

Serious geek post alert! Avert your eyes if you are easily frightened!

I've been wasting time at work the last few days. We're really close to a code release to customers, so making changes is a bad thing. Instead, you wait around for someone to say "something broke!" then rush to fix it. Nothing for me to fix recently, so I've been playing.

I'm still working on web stuff for my image gallery / home page with the thought of having as little work needed as possible in the publishing phase. Right now, all I need to do is throw a jpg in the right directory with the right filename format, and the rest is handled by the JavaScript in the index.html page. I like it that way.

But, I was thining, it would be cool to have some of the EXIF data available (shutter speed, aperture, focal length). "Hmmm", I think, "that info is in the jpg, and the browser has the jpg, so in theory the browser already has the information... how to get at it?" Well, an image loaded by an img tag is a black hole... no cool JavaScript methods to get anywhere with that. But, I've been playing with XMLHTTPRequest objects recently, which allow you to grab any web accesible file you want, like the jpg. Sounded good. I found the EXIF spec and some sample code on how to pull info out of the file, and started away.

Here's the problem... the closest I can get to the binary information in the jpg is a String representation. UTF8 should save me... but, it doesn't. I've found all kinds of wierdness depending on the browser and platform. I simplified things and threw away the EXIF code for now, and instead focused on just generating a hex dump of the jpg file (or the first kb or so of it). No good news. On the PC / firefox, all the bytes are there, but a bunch only contain the UTF8 "unavailable" character FE (or FFFE, or 65533 - yes, that's not a byte, more confusion). On the Mac / safari, I think almost all (or even all) the "bytes" I found came back as the right values, but there were some bytes missing! I think all the missing bytes were 00, but, when you're dealing with offsets to find the right byte of information, a missing byte is a problem.

IE just keeps yelling "undefined" at me.

current test code / hex dump is here http://homepage.mac.com/jtross/exif.html
the jpg it's trying to dump is here http://homepage.mac.com/jtross/photoblog/1.17.2005.jpg
if you want to play / compare for yourselves. Unix hexdump is the standard. "--" in the center part of the dump represents an "unknown" byte. "-" in the right part of the dump represents a byte that's not in the normal ASCII printable range.

So, I guess having the browser use JavaScript to extract the EXIF info isn't going to work.

Next idea, which will have to wait until I get home, is to use an AppleScript folder action to do the EXIF extraction for me and create an XML file with the data for each jpg I add. Then, use an XMLHTTPRequest to get the contents of the XML file, and JavaScript to display it.

[snid]

Quote: Originally Posted by Cape Cod MINI

I have no idea what you are talking about

I put a friggin warning at the top of the message!