Triligon / Projects / Xinha and ezxhtml / ezxhtml wiki / InsertezPictures




InsertezPictures (InsertezPictures)

Objective

To create a custom ImageInsertionModule[Create] to link Xinha and ezpublish.

Desired features

  • Easy selection of images from ezpublish tree
  • Support for size="[imagealias]"
  • Support for align

The output must be of the form:

<img src="realurl" id="ezlink://id" size=[imagealias] align=[left|middle|right] [...]>

This can be achieved by using the current HTML form which is halfway understandable.

Issues

  • HOW should the selection of the images look like???
    • specifically, should we just fetch all images and display small thumbnails?
    • or somehow group them? If so, how?
    • something entirely different? What?

Karel's suggestions

Selection of images (objects in general).

There should be several things configurable (in ini or however): subnodes of what class to show, subobjects of what specific nodes to show, class filter or abovementioned subobjects and show-object-at-this-node-id.

Do you think you could do a mockup graphic of that? I'm not quite sure I understand what it's supposed to do ;)

Criticism about rest of current dialog.First, internal path should NOT show in image url field. Instead it should be emptied and greyed out if internal object is selected. Also, ID field should be removed.

ID Field I'll likely make hidden (it is however very useful for debugging, so for the time being it will stay. URL I'll keep, it's not top priority to fix that to me (I have an idea, mostly related to a switch between internal and external images but that requires some more JS).

Image upload dialog should be in same popup - the fields may be hidden and only appear when upload new is clicked.

That's hard. Current ezpublish versions dont come with the edit before create patch AFAIK and thus the object NEEDS to be existing before I edit it which is only possible with using content/action.

That would need a lot of JS and I'm not quite sure I want to go that road. The usability improvement seems doubtful to me. One coudl even argue that it lowers usability for more experienced users as it means another page change.

So first step: three options of selection, picking one grays out or hides others somehow. Step 2 - appearance options, common to all.The selection from three options above could even be implemented separately as step 0, then next one will contain appropriate fields only. Bit lenghty for us experienced users, but common people will absolutely love this!

Yeah it might make sense. Willl have to think about it. And look how hard it would be, really ;)

What I absolutely want to do is a filter by image keyword thingy.


Comments


Name

zrg

Subject

thoughts about image upload

Message

Selection of images (objects in general).

There should be several things configurable (in ini or however): subnodes of what class to show, subobjects of what specific nodes to show, class filter for abovementioned subobjects and show-object-at-this-node-id.

So settings could be something like this:
# show all galleries in site
XeditInsertParentClass[]=gallery
# show images that are in Media/Images
XeditInsertParentNode[]=51
# show only nodes of type image
XeditInsertClassFilter[]=image
# also show this one image which is in strange place, not gallery
XeditInsertNodeId[]=124

The result of above settings in image selection area would be that in the left part of it would be menu with names of all galleries in system, plus one item for Media/Images, plus one for image at node 124. When clicked on name, in the bigger area at side of it images from under selected gallery or node are showed, in a very condensed format (like now), with prev/next page shortcuts etc.

Criticism about rest of current dialog.
First, internal path should NOT show in image url field. Instead it should be emptied and greyed out if internal object is selected. Also, ID field should be removed.

Image upload dialog should be in same popup - the fields may be hidden and only appear when upload new is clicked.

The alignment and size selection should be done as second step, after submitting the object selection page. This makes it much easier on unexperienced users, trust me. So, if user clicks on existing image, popup is submitted and he gets straight to second step: size and alignment selection with a nice preview. ("back to selection" buton is needed here of course). If user enters external url he also goes straight to step 2. And after filling up upload form, to step 2.

So first step: three options of selection, picking one grays out or hides others somehow. Step 2 - appearance options, common to all.

The selection from three options above could even be implemented separately as step 0, then next one will contain appropriate fields only. Bit lenghty for us experienced users, but common people will absolutely love this!

Heh, i just now tried out the online editor first time - it has all the options for object insert crammed in one popup too. If you take it to three steps like above, ez will be out of pocket money in no time :)

Hope the comment is helpful in some way. Easy to wish, hard to do, appreciate your efforts, man!