What's new in Annotorious 3.x

[vidya-talapady-percipient-ai]

Copying @rsimon 's answer from a previous question

Annotorious 3 is a complete rewrite which includes many breaking changes from v2.7. (It was driven by my own client projects, where requirements had been shifting. Hence the breaking changes, in favor of other priorities such as improved React integration, realtime-capable state management, a custom popup mechanism that actually works etc.) But v2.7 will stay around, for those who prefer the old feature set.

The main breaking difference is that v3 no longer provides a built-in editor popup. You now have to build one yourself. That seems like a backwards step at first glance. But the benefit is that you now have all the freedom to build things the way you want & need, without being held down by Annotorious' awkward and (during the past years...) completely broken "extension mechanism".

Please also note that in v3, support for images is incomplete. My own client projects are all built around OpenSeadragon currently.

Some notes specifically to your properties:

• disableEditor: no longer applies, since v3 is "headless-only", and doesn't come with an editor popup out of the box.
• allowEmpty: no longer applies either. Because there's no more editor empty annotations are always allowed.
• readOnly: roughly corresponds to the setDrawingEnabled method on the annotator. (But, again, differs a little because there is no more editor popup.)
• handleRadius : is not a config option (yet)
• formatter: is (roughly) replaced with the custom styling. Note that there is no replacement for labels yet. May come in the future. (But it's not a requirement for any of my current clients.)
• fragmentUnit is not yet supported.

I have a few questions on what how Annotorious 3.0 works.

[Vidyadayini]

1. From the docs I see now fragmentUnit is by default in pixels. Anno 3.x does not support percent format ? Just confirming my assumption.
2. What does v3, support for images is incomplete mean ? Trying to understand what support we had in 2.7 that we do not yet have in 3.x .

[rsimon]

1. That's right. Shape unit is pixels. The fragmentUnit option (percentage) is not yet supported in Annotorious v3.
2. To understand what's supported for images, please see the documentation on the ImageAnnotator and its React counterpart. The issues list might also list some recently reported gaps or bugs. The ImageAnnotator is fairly untested and provided as-is. As I said: I'm not working on it myself at the moment, since none of my client projects are using it.

[vidya-talapady-percipient-ai]

So my understanding is that your clients use Annotorious with OpenSeadragon and the Annonotorious OpenSeadragon plugin. OpenSeadragon has it's own pan and zoom. Can a custom pan and zoom component be easily integrated with Annotorious with OpenSeadragon ?
With OpenSeadragon, can we still leverage all the label methods and customizations provided by Annotorious ?

[rsimon]

I'm not sure what you mean by "custom pan and zoom". The purpose of OSD is to provide a pan- and zoom-able viewer (for "normal" images as well as various deep-zoomable formats). Do you mean you want to disable that, and hack your way around an additional custom pan/zoom feature? I'm not aware that anyone has ever done this. It's almost certainly going to be a pain, in a million ways. You'd have much less work in just fixing gaps in the Annotorious image code.

[vidya-talapady-percipient-ai]

Thank you so much. Do you have an estimate of when fragmentUnit could be implemented?

[rsimon]

Not yet, I'm afraid.