Hi Ann,

Thank you for continuing to engage the ImageJ community in this discussion!

However, please note that this is an issue tracker, not a general discussion forum. It would be much better to discuss these social issues on the ImageJ mailing list.

> I have now discovered that the person who was writing and supporting
> Coloc2 in Fiji is no longer doing so. They have apparently left the
> MPI in Dresden and are now working for a company.

If you refer to Dan White, the project lead: he is still active in the community and tries to help with Coloc 2 issues when he has time (e.g., just today: http://imagej.net/pipermail/imagej-devel/2015-June/002605.html). Dan is one of the most talented, friendly and helpful members of the community with whom I have had the pleasure of working.

If you refer to Tom Kazimiers, the main developer: AFAIK he currently works at HHMI Janelia Research Campus, as part of Albert Cardona's group. He is also an asset to the community and as you can see from his GitHub profile (https://github.com/tomka), he is still very active in development of related software projects.

> This is a major issue for those in the community who need
> colocalisation analysis for their scientific research. Since the
> plugin is effectively half finished and the writer of the plugin has
> little intention of completing the work apparently.

These statements are not constructive. Please understand that ImageJ and Fiji are open source software projects, not commercial software products. The ImageJ community supports the software largely on a volunteer basis, as time and energy permits. As such, it is important for users to be patient when reporting issues; see this FAQ entry for details: http://imagej.net/FAQ#I_reported_an_issue.2C_but_it_is_still_not_fixed.21_Why_not.3F

You will have to be very specific about:
* What exactly you are trying to accomplish;
* What exactly is insufficient about the Coloc 2 plugin;
* How you envision the software working, such that Tom or others could improve things

Your mail to the ImageJ list a few days ago (https://list.nih.gov/cgi-bin/wa.exe?A2=IMAGEJ;956552e2.1506) was a big step in the right direction; we should continue the discussion from there to identify specific ways the colocalization support in ImageJ can be improved for your use cases.

> Likely the MPI-Dresden will be taken up coding plugins for SPIM, such
> as Big data viewer and presumably wouldn't have personnel to finish
> the project.

Rather than make assumptions, why not have a respectful, public discussion which engages all relevant parties appropriately?

> Colocalisation was reliably and robustly implemented in IJ1 and needs
> must IJ2 / FIJI should at least be able to reproduce the functionality
> available in IJ1 in newer versions. 

As I said before (http://fiji.sc/bugzilla/﷒0﷓), Fiji ships both of the old MBF ImageJ colocalization-related plugins: Colocalization_Threshold and Colocalization_Test. You will need to be more specific about exactly what functionality is missing from Fiji that was present in MBF ImageJ and/or ImageJ 1.x.

Fiji tries very hard to be backwards compatible with ImageJ 1.x -- anything missing in that regard is very likely an oversight and readding old components which are still needed is almost always extremely straightforward. But from your description, I honestly do not know which plugins you are talking about.

> I can see if my image informaticist might be interested as a last
> recourse.

I encourage you to engage the technical resources available to you to best achieve your goals. ImageJ and Fiji are available to you as free and open source (FOSS) software, thanks to the efforts of many software developers worldwide over a course of many years. If your group contributes back to the project for the benefit of all, that is not a "last recourse" but rather a wonderful thing, and how Fiji was built. We're all in this together. http://imagej.net/Philosophy

From a technical perspective: I will leave this issue open hoping you will point me to the plugin(s) you want to see made available. Then we can close this issue after they have been added to the Fiji distribution.

Specifically:

Coloc2 is 90% done . The issues are all to do with (i) menu functionality e.g. selecting what to colocalise and which coefficients are needed. This is implemented in colocalisation threshold (set options checkbox), though this is mathematically weaker than Coloc2 and colocalisation threshold can't colocalise ROIs from the ROI manager.

(ii) Results display. In the output results file one needs to know what slice and ROI the colocalisation coefficients correspond to and be able to only chose the ones of interest for the study . 

(iii) This all needs to be scriptable so if one has 100 - 200 images Fiji can munch through them and generate an results file exportable to Excel with:

Columns corresponding to:
 
Slice, ROI, Threshold, 'My favourite colocalisation coefficient', 'Other colocalisation coefficient', 'AN OTHER parameter as desired'

Rows corresponding to:
Slice 1, ROI1
Slice 1, ROI2
Slice 2, ROI1
Slice 2, ROI2 etc

Not dissimilar to the output of multimeasure for analysing particles. Manders Coefficient and ICQ in IJ1 can do all of this and are scriptable easily.

(iv) Another thing is the x and y axis of the plots in Coloc2 are not labeled so it sin't clear which axis corresponds to channel 1 and channel 2 in the output. I assume it's always consistent. I guess this is a easy fix. 

(v) A more delicate matter, These issues have been outstanding for over a year, which is the reason for additional fuss. Generally all issues of 'plugin not playing' are promptly addressed in updates, since Fiji as a group of developers are really good at maintaining tools. So perhaps these Coloc2 coding issues may be easier to resolve than the 'finding a person who has time to do the coding' issue.. 

Thanks for the details, Ann.

> (i) menu functionality e.g. selecting what to colocalise and which
> coefficients are needed. This is implemented in colocalisation
> threshold (set options checkbox), though this is mathematically weaker
> than Coloc2 and colocalisation threshold can't colocalise ROIs from
> the ROI manager.

Filed at:
https://github.com/fiji/Colocalisation_Analysis/issues/7

> (ii) Results display. In the output results file one needs to know
> what slice and ROI the colocalisation coefficients correspond to and
> be able to only chose the ones of interest for the study . 

Filed at:
https://github.com/fiji/Colocalisation_Analysis/issues/8

> (iii) This all needs to be scriptable so if one has 100 - 200 images
> Fiji can munch through them and generate an results file exportable to
> Excel with:
> 
> Columns corresponding to:
> 
> Slice, ROI, Threshold, 'My favourite colocalisation coefficient', 'Other colocalisation coefficient', 'AN OTHER parameter as desired'
>
> Rows corresponding to:
> Slice 1, ROI1
> Slice 1, ROI2
> Slice 2, ROI1
> Slice 2, ROI2 etc
>
> Not dissimilar to the output of multimeasure for analysing particles.
> Manders Coefficient and ICQ in IJ1 can do all of this and are
> scriptable easily.

Filed at:
https://github.com/fiji/Colocalisation_Analysis/issues/9

However, it would be helpful if you could clarify what you mean by "Manders Coefficient" and "ICQ" in IJ1. Are these plugins? If so, where did you get them? Are they part of the MBF ImageJ distribution?

Doing a quick web search, I see that ICQ is a statistic available as part of the old Colocalization Analysis plugin -- is that what you mean?

> (iv) Another thing is the x and y axis of the plots in Coloc2 are not
> labeled so it sin't clear which axis corresponds to channel 1 and
> channel 2 in the output. I assume it's always consistent. I guess this
> is a easy fix. 

Filed at:
https://github.com/fiji/Colocalisation_Analysis/issues/10

> (v) A more delicate matter, These issues have been outstanding for
> over a year, which is the reason for additional fuss. Generally all
> issues of 'plugin not playing' are promptly addressed in updates,
> since Fiji as a group of developers are really good at maintaining
> tools. So perhaps these Coloc2 coding issues may be easier to resolve
> than the 'finding a person who has time to do the coding' issue..

Please understand that your "additional fuss" comes off as an attitude of entitlement, which may negatively motivate those involved in the project from addressing your wishes. It is always best to be polite. The open source software development is radically different from more traditional goods and services; here is a nice summary:

http://jeremyckahn.github.io/blog/2014/10/19/open-source-does-not-mean-free-labor/

The traditional open source mantra is "scratch your own itch" -- because with OSS, you _can_. Like I said before, if you have development resources available to you, and you need the above features to be addressed, this is a perfect opportunity to give back to the OSS community.

If you have additional comments on any of the issues above, you are very welcome to create a GitHub account and comment on the issue(s) in question. Your feedback is very helpful here!

Lastly, it may interest you to know that I'll be visiting the HHMI Janelia campus in a couple of weeks, where we are planning to discuss future development directions of Coloc 2 with one of the main developers. Please understand that AFAIK, no one is actually funded to work on this project right now, so any progress will probably be slow. But obviously we do want it to be a high quality, useful piece of software for the scientific community, so hopefully that discussion will be fruitful for planning and next steps.