Planning interface architecture about image-sequencer HOT 14 CLOSED

I actually think we may want to create an extremely minimal "simple" html interface first to prove out our architectural decisions and also make it easy for people to understand how an interface can be implemented. Then circle back for a more awesome interface as a second step. What do you think?

from image-sequencer.

Like maybe the simplest just displays the images in div, no more.

Also, let's think carefully about designing this so that the interface code may eventually be spun out into a separate module.

from image-sequencer.

Okay! All this sounds great! Just a few thoughts here:

• Shouldn't the modules be independant of the implementation (i.e. the UI/FrontEnd Code)?
• We also have support for NodeJS. So what will the above methods as mentioned by you do in NodeJS?
• What I had in mind was a module called "UserInterface.js" (Just like Image Sequencer had, initially.). whenever an image is drawn, UserInterface is informed. Then It produces the necessary HTML Elements, or the NodeJS output. Basically all the methods above as you said, just that they connect to UserInterface and that modules themselves don't handle any HTML.
• Also, users should be able to make their own UI as per their requirements, so they Can implement Image Sequencer in their projects, right? So shouldn't image sequencer act like an API? Taking instructions from the user and returning the output so they can use it the way they want.

What do you think about all this?

from image-sequencer.

I do agree to the fact that we should have a basic, but rigid UI in the beginning, and then develop it furtyher.

I actually think we may want to create an extremely minimal "simple" html interface first to prove out our architectural decisions and also make it easy for people to understand how an interface can be implemented. Then circle back for a more awesome interface as a second step. What do you think?

from image-sequencer.

from image-sequencer.

That sounds great! 👍

from image-sequencer.

Had another thought on this -- the most basic version of the demo interface could just insert a default options object into a textarea, and allow people to just edit it directly in text. This isn't a friendly UI, but it very clearly explains what the UI allows you to do with regard to options. What do you think?

So a demo UI could:

1. display the options parameter object for editing
2. display the image after the step is run
3. show a button to re-run the whole sequence at the bottom of the sequence

from image-sequencer.

By options, do you mean the options in this syntax:

sequencer.addSteps('nodule-name',options)

If yes, then I agree with all you've said!

from image-sequencer.

from image-sequencer.

Please see #67

from image-sequencer.

Ok, I think we should implement an example (maybe all examples?) Using the new UI system. Then link to the examples to show how the UI may be hooked up. What do you think?

from image-sequencer.

Yes I think that is a good way to show how to use the UI!

from image-sequencer.

Made the FisheyeGl Demo with the new interface.
It can be seen at https://ccpandhare.github.io/image-sequencer/examples/fisheye

from image-sequencer.

I think we are done with this, closing the issue now! :-)

from image-sequencer.