Feature summary
In Wikisource Proofreading page editing view, we can no longer scroll the djvu page image being proofread. This feature used to be available, then removed because of technical issues, but never installed/activated in Seadragon.
Can you please fix this feature, as promised.
Thank you
@Ineuw: For future reference, please use the feature request form (linked from the top of the task creation page) to create feature requests, and fill in the sections in the template. Thanks.
This feature used to be available, then removed because of technical issues,
Reference welcome.
but never installed/activated in Seadragon.
What is "Seadragon"?
Can you please fix this feature, as promised.
Where to find some promise?
@Ineuw The current interface implemented using Openseadragon binds the scrolling function to zooming the image instead of the panning up and down (which could be toggled into scroll to zoom by clicking on the image) in the old implementation. However, the old system was very hard to use in terms of touchpad/accidental clicking which was one of the reasons we switched to the newer system. It is technically possible to implement this behaviour using Openseadragon, but I feel like implementing it would actually be a step backwards in terms of usability.
Can you provide specific examples where scrolling to pan up and down would be beneficial compared to our current system?
@Soda, The new system is very good for text only pages, BUT, I can give you a 1,000 examples I am working on currently. Please attempt and edit a page with images and text. https://en.wikisource.org/wiki/Page:Africa_by_%C3%89lis%C3%A9e_Reclus,_Volume_3.djvu/590. Images are surrounded by text and navigating around an image without scrolling doubles my time of proofreading a page. This may sound crazy but I see results in my recent contributions on WS.
I claim experience on the basis of the number of images I contributed to the commons which ended up used in Wikisource proofreading projects. One does not make such an arbitrary decision of dumping a feature that existed since the beginning of Wikipedia and then abandon it. Inductiveload was kind enough demonstrate the Openseadragon and I can live with the extra clicks. But if you are inclined to consider, perhaps you can make this as an option the users' preferences, Edit settings.
Hi. I studied both, and selected the correct template but removed the lines that were irrelevant. At the time of writing I only remembered the word seadragon name but found it very odd that Wikipedia would use a Microsoft software, but then it was close enough, am I right?
The djvu image in the proofreading page was ALWAYS scrollable at least since I started in 2009. I hope that this is enough reference.
I assumed the "promise" in posts with Inductiveload when he guided me to the demo page. No one ever used the words "I promise"
I intend to take a look at some of the Openseadragon patches and try to shepherd them along and fix outstanding bugs wrt to Openseadragon if possible during this summer. I honestly feel like implementing exactly the same interface/interactions as were present in the old zooming panning library would be a step back in terms of usability. That being said, Inductiveload's patch for adding the Ctrl + Scroll modifier is a good starting point of adding the more and more desirable features to our Openseadragon implementation.
One of the things we discussed but wasn't implemented was the ability to remember zoom and pan position across pages. Implementing scroll and pan remembering across pages would effectively nullify any of the current issues wrt to scrolling to specific areas of pages since you should be able to pan the page once per every Index: page and then not have to adjust pages until you encountered (say) a text page.
But if you are inclined to consider, perhaps you can make this as an option the users' preferences, Edit settings.
We do already expose a lot of customization options via the mw.proofreadpage.Viewer endpoint. Almost anything that we implement inside ProofreadPage can be implemented clientside via scripts/gadgets using that API. If you want to do some specific/custom interaction, you can use the API to effectively make any kind of changes and customize the Openseadragon viewer in any way that you want.
One of the things we discussed but wasn't implemented was the ability to remember zoom and pan position across pages. Implementing scroll and pan remembering across pages would effectively nullify any of the current issues wrt to scrolling to specific areas of pages since you should be able to pan the page once per every Index: page and then not have to adjust pages until you encountered (say) a text page.
I believe that you are under a misconception. Remembering zoom position and size between pages is irrelevant. Editing is better served by the current default reset of each new page opened for editing. Default magnification and the top of the page where editing always begins. Magnification is great for a closeup look, scrolling is far easier than sliding a djvu image 4 - 5 times to get to the text below an image (even when the djvu is minimized). There is no vertical real-estate to speed this up. Doing this thousands of times ends with me complaining here.
I hope you have a Wikisource account to edit a couple of pages with the current pages. Try my current project Africa Volume 4.