Symphony.

Hi,

There’s a problem. I installed an ‘image cropper’ extension. Then I found out that my JIT doesn’t have an /image/4 mode, which was described in some post. It was named jCrop or something, in fact, I don’t understand the idea, because in JIT’s README there are only 0,1,2,3 modes. I got a bit mixed up with all this Modified JIT extension (http://github.com/klaftertief/jitimagemanipulation/tree/jCrop), so I uninstalled my JIT and tried to install a new one (from the link above). I did it via ‘git clone’ command. Now, I have a JIT installed (in extensions) but it totally doesn’t work. Neither 4 mode nor 0,1,2 and 3.

Help me please, because I’m getting frustrated with this :(

Okay, I downloaded JIT from here http://github.com/klaftertief/jitimagemanipulation/tree/jCrop and uploaded it via FTP. Now, the version is not 1.06 but 1.09. I don’t get it at all, but it works :)

Hi rudy,

the Image Cropper extension needs the forked JIT extension to be able to crop images. After cloning the repository from the link above you have to switch to the jCrop branch. I should have made this clearer in the README. But that doesn’t explain why JIT doesn’t work at all. Is the JIT extension listed and activated in your Symphony installation?

Glad you got it working. Seems you replaced an older version of JIT with a newer one. Symphony disables updated extensions and you have to reactivate those to run their (potential) update scripts.

Quick question thats not directly related…

Is there a way to allow all images from external sources to be manipulated? I don’t want to copy and paste every external domain that I link to.

Didn’t try by myself but maybe * works?

I tried that but it seems to have no effect.

Maybe http://* is worth a try?

I’ve got completely mixed up! I’ve removed JIT and uploaded the new one from http://github.com/pointybeard/jitimagemanipulation via FTP (no SSH access):

1. “Normal” JIT functions (let’s say, resizing) does work only for my old images. When I upload new ones, it doesn’t work. Access permissions are the same, my browser chache is clear, I’ve even removed all files from /manifest/cache directory. Old images are resized and displayed nicely. For new images I get 404. What’s wrong? This is so weird…
2. I don’t know how to “merge jCrop branch”. How to do it? I have only FTP access to my server.

“Normal” JIT functions (let’s say, resizing) does work only for my old images.

Sounds like you’re seeing cached images, I might be wrong though. What does /manifest/logs/main say and are the .htaccess rules in place?

I don’t know how to “merge jCrop branch”. How to do it? I have only FTP access to my server.

In this case you simply have to download from http://github.com/klaftertief/jit_image_manipulation/tree/jCrop

This is /logs/main file:

2010/10/05 16:52:22 > WARNING: 2 - filemtime() [function.filemtime]: stat failed for /seiko/workspace/0 in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 143

2010/10/05 16:52:22 > NOTICE: Image class param dump - mode: 0, width: 0, height: 0, position: 0, background: 0, file: 0, external: 0, raw input: 

2010/10/05 16:52:22 > ERROR: 256 - Image /seiko/workspace/0 could not be found. in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 220

2010/10/05 16:52:22 > NOTICE: Image class param dump - mode: 0, width: 0, height: 0, position: 0, background: 0, file: 0, external: 0, raw input: 

2010/10/05 16:52:22 > WARNING: 2 - readfile(/seiko/workspace/0) [function.readfile]: failed to open stream: No such file or directory in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 225

2010/10/05 16:52:22 > NOTICE: Image class param dump - mode: 0, width: 0, height: 0, position: 0, background: 0, file: 0, external: 0, raw input: 

2010/10/05 16:52:53 > WARNING: 2 - filemtime() [function.filemtime]: stat failed for /seiko/workspace/0 in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 143

2010/10/05 16:52:53 > NOTICE: Image class param dump - mode: 0, width: 0, height: 0, position: 0, background: 0, file: 0, external: 0, raw input: 

2010/10/05 16:52:54 > WARNING: 2 - filemtime() [function.filemtime]: stat failed for /seiko/workspace/0 in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 143

2010/10/05 16:52:54 > NOTICE: Image class param dump - mode: 0, width: 0, height: 0, position: 0, background: 0, file: 0, external: 0, raw input: 

In this case you simply have to download from http://github.com/klaftertief/jitimagemanipulation/tree/jCrop

I’ve done it. Works only for old images.

The problem is that probably JIT doesn’t work at all and it just shows cached images. But how is that even possible? Where could my old images be cached?

Ahh! And .htaccess - it is all right but I had to remove this part of it:

### CHECK FOR TRAILING SLASH - Will ignore files
#RewriteCond %{REQUEST_FILENAME} !-f
#RewriteCond %{REQUEST_URI} !/$
#RewriteCond %{REQUEST_URI} !(.*)/$
#RewriteCond %{REQUEST_URI} !.*debug.*
#RewriteRule ^(.*)$ $1/ [L,R=301]

This makes Symphony work at my hosting provider’s server.

Any ideas?

/seiko/workspace/0

Looks very suspicious to me. What URL are you requesting?

I’ve made 3 tries. First is an old image:

http://www.visio.home.pl/seiko/image/1/150/400/obrazki-naglowek/top.example.jpg

No log, nice display.

Now, two new images:

http://www.visio.home.pl/seiko/image/1/150/400/obrazki-naglowek/az1.jpg http://www.visio.home.pl/seiko/image/1/150/400/obrazki-naglowek/az2.jpg

They log:

2010/10/05 19:57:38 > ERROR: 256 - Cannot load CMYK JPG Images in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 236

2010/10/05 19:57:38 > NOTICE: Image class param dump - mode: 1, width: 150, height: 400, position: 0, background: 0, file: obrazki-naglowek/az1.jpg, external: 0, raw input: 1/150/400/obrazki-naglowek/az1.jpg

2010/10/05 19:59:30 > ERROR: 256 - Cannot load CMYK JPG Images in file /seiko/extensions/jit_image_manipulation/lib/image.php on line 236

2010/10/05 19:59:30 > NOTICE: Image class param dump - mode: 1, width: 150, height: 400, position: 0, background: 0, file: obrazki-naglowek/az2.jpg, external: 0, raw input: 1/150/400/obrazki-naglowek/az2.jpg

404 error.

The log sais Cannot load CMYK JPG Images, JIT can not process CMYK images (yet).

This has been discussed before.

EDIT: I think this is a bug, especially since this pops up every now and then, and does not really make sense.

Wow, those pictures were uploaded by a graphics designer. I’ve downloaded one of them, opened in Paint, resaved and reuploaded. The problem has vanished! That must be a problem with a picture format.

Could you take a look at http://www.visio.home.pl/seiko/image/1/150/400/obrazki-naglowek/az1.jpg and tell me why it doesn’t work? What’s wrong with this jpeg?

Creativedutchmen: huge thanks! But it really makes me curious: is that such a big deal to implement processing CMYK files? It should be just a matter of using a proper JPEG Processing library or something, huh?

Rudy: for as far as I know, the gd library supports cmyk only in the later versions (and gd is used in git).

I guess JIT could be extended to do a check for the newer version, and apply cmyk images if available, it just takes coding time..;)

CMYK isn’t mentioned to be used with displays as it’s a subtractive color space (for printers: more paint => less reflected light) whilst RGB is an additive color space (for displays: brighter pixels => more emitted light).

One might think transforming CMYK to RGB is easy but it’s not, there are hundreds of profiles to do so. This means you would end up with one picture looking very differently on different computers.

This means you should always use RGB for displays. I think most people who create stuff in CMYK should be able to convert it to RGB before uploading, you just have to tell them. :-)

Maybe it would be an idea to extend the core upload field to display a short message when CMYK images are uploaded? Like an CMYK tag behind the file name that shows more details on hover.

I think that’s the better idea Nils H, rather that than adding CMYK support.

It’s often a lot easier to convert RGB to CMYK rather than, like Nils W said, CMYK to RGB, so if, for example, the site was there to provide downloadable images, they can be CMYK’d afterwards.

Some browsers are a little weird with CMYK images too…