Re: What does ImageJ stand for?

Posted by Curtis Rueden-2 on
URL: http://imagej.273.s1.nabble.com/What-does-ImageJ-stand-for-tp5024714p5024736.html

Hi everyone,

In response to this thread, I have done the following:

* The front page (https://imagej.net/) now says "Welcome to the ImageJ
ecosystem" and shows a disambiguation box explaining about the two flavors
of ImageJ, ImageJ1 and ImageJ2, with links to more information.

* The downloads page (https://imagej.net/downloads) now clearly links to
the download pages of both Fiji and ImageJ1, with a sidebar explaining why
there is not yet a download for plain ImageJ2.

* On that same page, just below the download links, there is an expandable
box with the label "Need help deciding? Click here." that lists some bullet
points in favor of each system. As a software engineer, I dislike
marketing-style comparisons; my goal with this table was to sum up, for new
users, why you might want to use each tool. Contributions to these
respective bullet point lists are welcome, especially if you feel a major
strength of either ImageJ1 or ImageJ2 is missing from the table—please go
ahead and edit the wiki page directly.

* I removed the "© ImageJ" from the imagej.net footer; it was left over
from the original theme. The license for the imagej.net website is CC BY
3.0 (https://github.com/imagej/imagej.github.io/blob/main/LICENSE.txt).

* I moved ImageJ2's top-level GitHub repository from
https://github.com/imagej/imagej to https://github.com/imagej/imagej2. This
hopefully helps clarify that imagej/imagej1 is the repository for ImageJ1,
and imagej/imagej2 is the repository for ImageJ2. We should not have any
repository at https://github.com/imagej/imagej because it would be
ambiguous. Relatedly, it would be nice to eliminate the imagej/ImageJA
repository, but I do not think it is technically feasible unless Wayne
changes the ImageJ1 release workflow.

* A relevant improvement that is in progress, but not yet complete, is
https://github.com/imagej/imagej.github.io/issues/97: restoring the
"project icon" that indicates which particular project area a page belongs
to. For example, the TrackMate plugin is part of Fiji, and should have a
boxed Fiji icon at the top next to the page title, which when moused over,
explains that that plugin is part of Fiji. The old wiki had this feature,
but we haven't finished porting it over to the new wiki yet; once this is
complete, I hope it will help mitigate the confusion surrounding the scope
of each wiki page.

Please note that *ImageJ2 is not Fiji*. The ImageJ2 project is an
initiative, funded by NIH in 2010, to expand ImageJ's utility and scope.
ImageJ2 is a continuation of ImageJ, not a fork—you can think of it as a
superset of ImageJ1, since one of its components is a compatibility layer
with ImageJ1. ImageJ2 also includes the ImageJ Updater, ImageJ Launcher,
foundational SciJava components, and other libraries. For details, see
https://imagej.net/develop/architecture and/or
https://doi.org/10.1186/s12859-017-1934-z.

The Fiji distribution of ImageJ is a download of ImageJ2 + the Fiji
plugins. ImageJ2 is distinct from Fiji, and usable separately as a software
library, with or without ImageJ1 support. It is part of my group's
immediate-term roadmap to enhance the ImageJ Updater, such that it becomes
feasible to offer a download of ImageJ2 only without the Fiji plugins,
which lets you choose individual plugins to install as you need them,
rather than requiring a single monolithic Fiji update site for the Fiji
plugins.

Regards,
Curtis

--
Curtis Rueden
Software architect, LOCI/Eliceiri lab - https://uw-loci.github.io/
ImageJ2 lead, Fiji maintainer - https://imagej.net/people/ctrueden


On Mon, Jun 14, 2021 at 11:42 AM Stein Rørvik <[hidden email]>
wrote:

> I also use ImageJ 1.x almost exclusively as there is little of value for
> me to be found in the Fiji distribution as I don't work with biology. The
> very slow load time and poor Windows compatibility of Fiji makes me avoid
> it whenever possible. When I find useful plugins in Fiji, I usually try to
> figure out the dependencies so that I can get them to work in ImageJ 1.x.
> Yes, this can be a mess, but it works well for my workflow.
>
> I agree with Herbie's thread starting post here, in that when you visit
> the new imagej.net pages, you get the impression that Fiji _is_ ImageJ;
> the pages obscures the fact that Fiji should be considered a _fork_ of the
> ImageJ project which has now evolved into something quite different. Fiji
> is often incompatible with the "classic" ImageJ program since the new
> plugins usually rely on a totally different code base on the bottom level.
> There is nothing wrong with that, and Fiji in itself is very useful, but
> the distinction between the two flavors should have been made more clear on
> the wiki. When you visit https://imagej.net/downloads the only reference
> to ImageJ 1.x is at the bottom, where there is a link to "old versions of
> ImageJ 1.x " which should instead read "the original version of ImageJ".
> Also, this link should have pointed to
> https://imagej.nih.gov/ij/download.html or https://wsr.imagej.net/distros/
> rather than https://imagej.nih.gov/ij/download/ which is now outdated.
>
> When that is said, I see nothing wrong with the term ImageJ1. When this is
> used in context, it makes it much more clear that we are thinking of the
> "original" ImageJ version. The problem with the new wiki at imagej.net is
> that it uses the term ImageJ instead of ImageJ2 in too many contexts.
>
> I also see nothing wrong in using the term "ImageJ" in a broader context
> (including Fiji) when (and only when) you are referring to unrelated
> programming frameworks or image analysis ecosystems such as Amira/Avizo or
> MatLab. When someone says "I did the segmentation in ImageJ and analyzed
> the objects in Avizo" I think this phrase would mean the same both when the
> person used ImageJ 1.x or Fiji.
>
> Stein
>
> -----Original Message-----
> Sent: 13. juni 2021 20:54
> Subject: Re: What does ImageJ stand for?
>
> Greetings,
>
> To be pedantic, there are no Linux distributions called "Linux" and
> generally for those who don't know the difference between a kernel and a
> distribution it would not help the resolution to their issue to know this.
>
> Since I only use ImageJ, i.e., ImageJ1, I get frustrated when searching
> for new things or resolutions to problems to find out the hard /
> embarrassing way that what I found was specifically for FIJI and not
> ImageJ1.  I find quite often, the FIJI documentation referred to itself as
> ImageJ and not blatantly as FIJI.
>
> Fred
>
>
> On Sun, June 13, 2021 9:41 am, Christophe Leterrier wrote:
> > As I wrote, if thousands of developpers can understand what they talk
> > about when they mention "Linux" I think we can manage to use "ImageJ"
> > for several related things without any risk of damage
> >
> > Le dim. 13 juin 2021 à  15:48, Herbie <[hidden email]> a écrit :
> >
> >> Thanks Christophe!
> >>
> >> "Regarding ownership of the ImageJ name, to my knowledge ImageJ1 (the
> >> software created and developed by Wayne) is and has always been
> >> public domain."
> >>
> >> -- First, ImageJ1 is not the name Wayne used for the software he
> >> developed.
> >>
> >> -- Second and legal questions aside, isn't it simply a question of
> >> fairness (especially in science) to not use the name of a renowned
> >> piece of scientific software for a whole group of software that was
> >> in great part
> >> (MB) not written by its author?
> >>
> >> Apart from that and as I wrote before, it is generally not a good
> >> idea to
> >> (later) use a specific term (ImageJ as a single stand-alone software)
> >> _also_ as a generic term (ImageJ as "software ecosystem" comprising
> >> differently named stand-alone software products) in the same field.
> >>
> >> Best
> >>
> >> Herbie
> >>
> >> ::::::::::::::::::::::::::::::::::::::::::::::::::
> >>
> >> lechristophe wrote
> >> > Some elements here from a question you aksed in Image.sc about
> >> > ImageJ trademarking in 2015
> >> > https://forum.image.sc/t/imagej-trademark-of-uw-madison/433
> >> >
> >> > Regarding the logo, the original thread about a new logo and the
> >> contest
> >> > is
> >> > here from the end of 2017:
> >> > https://forum.image.sc/t/contest-new-imagej-logo/8485
> >> >
> >> > there's another thread with the voting process that lead to the
> >> > choice
> >> of
> >> > the current logo.
> >> >
> >> > Christophe
> >> >
> >> >
> >> >
> >> > Le dim. 13 juin 2021 Ã  10:39, Christophe Leterrier <
> >>
> >> > christophe.leterrier@
> >>
> >> >> a écrit :
> >> >
> >> >> Dear all,
> >> >>
> >> >> As a long-time user of ImageJ (I'd say since 2004-ish), I saw
> >> >> ImageJ branching out several times with initiatives such as Fiji
> >> >> (which I
> >> have
> >> >> used for a decade), ImageJ2 (which in my understanding kinda
> >> >> morphed
> >> into
> >> >> an alternative, unfininished back-end within ImageJ1), sometimes
> >> >> corresponding to funding efforts. I know about other "branches"
> >> >> such
> >> as
> >> >> ImageJFX but I don't know them well and they seem less developed.
> >> >>
> >> >> Regarding Fiji, I have read some make the parallel with a Linux
> >> >> distribution, Fiji being an ImageJ "distribution". I think this is
> >> >> an interesting analogy as "Linux" today can mean the kernel or
> >> >> something from the whole ecosystem (whether or not sanctionned by
> >> >> the original
> >> creator)?
> >> >>
> >> >> Regarding ownership of the ImageJ name, to my knowledge ImageJ1
> >> >> (the software created and developed by Wayne) is and has always
> >> >> been
> >> public
> >> >> domain. I don't know enough about licenses to be sure what kind of
> >> >> derivative it allows, but in my understanding it is quite flexible.
> >> >>
> >> >> Christophe
> >> >>
> >> >>
> >> >>
> >> >> Le sam. 12 juin 2021 Ã  13:03, Herbie &lt;
> >>
> >> > l16@
> >>
> >> > &gt; a écrit :
> >> >>
> >> >>> Dear listers,
> >> >>>
> >> >>> in the sequel to my previous question posted to this list,
> >> >>> Kevin Eliceiri immediately provided some information regarding
> >> recent
> >> >>> terminological changes.
> >> >>> Because his reply to my original question didn't make it to the
> >> >>> list yet, and because the further email-exchange happened
> >> >>> off-list, I
> >> should
> >> >>> like to post my understanding of the situation.
> >> >>>
> >> >>> 1.
> >> >>> A longstanding fact is that Wayne Rasband conceptualized,
> >> implemented
> >> >>> and continuously extends and maintains the scientific image
> >> processing
> >> >>> software that he named "ImageJ" nearly 25 years ago.
> >> >>> (I can't remember that he ever termed it differently.)
> >> >>>
> >> >>> 2.
> >> >>> Only recently however, "ImageJ" appears being used in either a
> >> different
> >> >>> or a much wider sense (see the attachments to my previous post).
> >> >>>  From Kevin Eliceiri I now learned that here "ImageJ" stands for
> >> what
> >> he
> >> >>> calls the "ImageJ ecosystem" that includes all "ImageJ elements"
> >> such
> >> as
> >> >>> "IJ1, IJ2 and FIJI".
> >> >>>
> >> >>> ----------------------------------------
> >> >>> I feel that it is not a good idea to use "ImageJ" as both, a
> >> specific
> >> >>> _as well as_ a generic term, and I'm pretty sure that such usage
> >> will
> >> >>> puzzle not only me.
> >> >>>
> >> >>> Did Wayne Rasband agree with the "extended" meaning and usage of
> >> >>> "ImageJ", the name he gave to the software that he developed?
> >> >>> ----------------------------------------
> >> >>>
> >> >>> Apart from that, I'm surprised that IJ1 (that I understand is
> >> >>> Wayne Rasband's ImageJ) or Fiji, both stand-alone software,
> >> >>> should be
> >> regarded
> >> >>> as elements of ImageJ.
> >> >>>
> >> >>> All opinions welcome...
> >> >>>
> >> >>> Herbie
> >> >>> ::::::::::::::::::::
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html