Revision as of 03:41, 29 October 2014 by Schindelin (talk | contribs) (ImageJ1: reorder in the priority order Wayne *seems* to prefer (he never offered any explicit statement about it))

The community encourages discussion about proposed changes on the ImageJ and imagej-devel mailing lists. Submit your ideas!

Start on the mailing lists, searching for discussions related to your contribution to get some context & background. It can also be helpful to search for applicable issues and plans in ImageJ's Trac issue tracker or on the GitHub issue list for ImageJ2, ImgLib2, and SCIFIO. The ImageJ community believes that public discussion is important so that ideas are not lost and exposed to healthy alternate points of view.


Submit patches to ImageJ2 via pull requests against ImageJ2's source on GitHub.

Note that since ImageJ2 has a modular architecture, it is possible that your change would be more applicable to one of the supporting technologies such as the ImgLib2 or SCIFIO projects.


Changes to ImageJ1 are made by Wayne Rasband, who is the sole ImageJ1 developer. He takes patch submissions and then reworks them to fit within the project's development model and style before merging them. Attribution for the changes is noted in the release notes (see ImageJ1's Release Notes/News).

Methods of getting the patch to Wayne include:

  • Send the modified code in a private mail to Wayne (he does not like to discuss code on the mailing list)
  • Send a patch to Wayne via private mail (it should apply to the latest revision and must not use any Git extension because Wayne uses patch(1) to apply the patch)
  • Send a patch to the imagej-devel mailing list (but note that traditionally, Wayne does not discuss code there)
  • Submit a pull request on GitHub against Wayne's ImageJ1 repository (please note, though, that none of the past Pull Requests were merged using the standard Git workflow)

The important part is that Wayne receive the code/patch, since he is the only one with the authority to merge it.

(See also this nice response from Curtis Rueden on the ImageJ mailing list.)