Photos
Implementation Stories, in priority order
I can have a better workflow in newer browsers: Chrome supports a very different tech standard than the other browsers, the new getUserMedia API. That needs an entirely separate code path. The results are probably slicker, but it's not critical.
I should be able to make my photos upright: I'm going to call this a usability bug. It appears that each mobile device has its own notion of the orientation of photos; sadly, this doesn't appear to always be the way you
took the photo. So pictures often come out sideways.
I should be able to keep my Photos properly private: At the moment, photos are purely security-through-obscurity, with long random filenames. That isn't really secure in any meaningful sense. We should be able to set them to be genuinely secure, using Amazon's Secure Token Service.
I should be able to use GIF or PNG images: The internals of Photos are written more neutrally as Images, but there are a few places where we are assuming that the MIME type is JPEG. Figure out how to detect the actual MIME type of the source image, and use that.
Also referred to by
After I Add a Photo, button should change to "Replace Photo": In general, this is probably a symptom of the weakness of the Photo-editing UI. If I'm in the Instance Editor and have an empty Optional Photo, I click the "Add Photo" button to fill it. But after the upload is complete, it still says "Add Photo" -- indeed, there is no feedback that it has worked until I say Done.
I can edit a photo that I have uploaded into Querki: From a conversation about Kate's Craft Gallery: one friend remarks that she likes to do some simple editing of her photos, in iPhoto. Ideally, it would be good if we could do something similar with Querki.
I can have a better workflow in newer browsers: Chrome supports a very different tech standard than the other browsers, the new getUserMedia API. That needs an entirely separate code path. The results are probably slicker, but it's not critical.
I should be able to keep my Photos properly private: At the moment, photos are purely security-through-obscurity, with long random filenames. That isn't really secure in any meaningful sense. We should be able to set them to be genuinely secure, using Amazon's Secure Token Service.
I should be able to make my photos upright: I'm going to call this a usability bug. It appears that each mobile device has its own notion of the orientation of photos; sadly, this doesn't appear to always be the way you
took the photo. So pictures often come out sideways.
I should be able to use GIF or PNG images: The internals of Photos are written more neutrally as Images, but there are a few places where we are assuming that the MIME type is JPEG. Figure out how to detect the actual MIME type of the source image, and use that.
Closed Issues
I can specify the desired size of my photos: Initially, we'll be clamping all photos to 1000 pixels on a side, and that should remain the maximum. But we should provide a way to specify a smaller size, to save space and screen real estate.
I can specify a Caption for a photograph: This should probably be possible both at upload time and editing later, which suggests that the Upload Image and View Image dialogs should be unified in some fashion.
I can specify a Target for showing full-size Images: A better way to deal with clicking on Thumbnails. I can say "My List of Photos -> _imageTarget", and this will display the first item in the List, full-sized. Moreover, it will be tagged with the Property info so that, when I click on a thumbnail, the full-sized version of that Thumbnail
replaces this Target.
Photo upload feedback is failing?: Kate has also reported this. I don't know if it's consistent, but I've observed the upload process getting going, and then appearing to hang in the UI.
Photos are too big: Even if I put a photo inside, eg, col-md-6, it is still shown full-sized.
Uploading photos frequently failing: Anecdotally, it seems like when I am creating photos in my Vacation Photos Space, half the time it is uploading, getting to the "Processing..." stage, and then hanging. Notably, it seems to
always succeed the second time I try. If things are deterministic, this might suggest that it is uploading the first time, but not reporting back?
I can specify the desired size of my photos: Initially, we'll be clamping all photos to 1000 pixels on a side, and that should remain the maximum. But we should provide a way to specify a smaller size, to save space and screen real estate.
I can specify a Caption for a photograph: This should probably be possible both at upload time and editing later, which suggests that the Upload Image and View Image dialogs should be unified in some fashion.
I can specify a Target for showing full-size Images: A better way to deal with clicking on Thumbnails. I can say "My List of Photos -> _imageTarget", and this will display the first item in the List, full-sized. Moreover, it will be tagged with the Property info so that, when I click on a thumbnail, the full-sized version of that Thumbnail
replaces this Target.
Add Photo button should activate on Enter: This is slowly driving me nuts as I build the documentation. In the Instance Editor, the Add Photo button doesn't react to the Enter key -- I have to take my hands off the keyboard and use the mouse to activate it. That's just dumb.
Photo controls don't encapsulate right: See the Lochleven inventory space. If I do "Create another Item" from an Item, and take a photo in that editor, the photo gets attached to the originating Thing, not the one being edited. That is, we're using the page identity, not the editor one, and they might be different.