Re: Why are images automatically renamed?

Posted by vischer on Oct 08, 2010; 7:48am
URL: http://imagej.273.s1.nabble.com/Why-are-images-automatically-renamed-tp3686678p3686684.html

Hi Gabriel

To illustrate the perfect confusion, consider this case:

You have 50 images in a folder, named:
pic_01
pic_02
..
pic_50

Today, you want to process pic_23.
Unfortunately, you are not aware that pic_23 already resides in the dock since yesterday.
You drag file "pic_23" onto the ImageJ main window, but it opens as pic_24.

You don't notice that the index has increased, process the image and save it.
You are asked "pic_24 already exists, replace?"
Yes, you want to save the processed image.

What has happened? For renaming, ImageJ has chosen the name of an existing file -
the original pic_24 is lost because it was overwritten by processed pic_23.
You don't notice this immediately (especially in time lapses), and are not aware
that you spoilt your analysis results.

>
> How can one differentiate between cases a) and b)?
> How will plugins (and users) resolve the exact duplicate names in b)?

The unique ID (as Frederick mentioned) and retrieving the path via "GetInfo"
gives complete information about an image. As mentioned earlier, "GetInfo" currently
gives incorrect information about images that are silently duplicated or renamed.

Generally it is a bad idea to silently create copies in order to solve some other
problem like a name conflict - copies create confusion.

regards, Norbert