Closed jmathai closed 12 years ago
I wouldn't want the web address for the image that gets forever stored inside the publisher's web content to be the create
web address, so the "generated=true" approach seems to work well enough.
I think the ideal scenario would be for the OpenPhoto API to return the web address where the image will be. If someone requests the image but it doesn't exist, then have OpenPhoto generate it as part of the 404. After that initial request and image creation, the image would exist for all subsequent requests. Is something like that possible?
@randyhoyt That makes sense. I didn't realize it was stored in the db, generate=true
definitely applies here.
WRT how the API handles it I believe the create URL is less "disruptive" than returning a 404. As a developer you're either interested in
Randomly (from the dev/user perspective) giving a 404 seems hard to program around. The create URL + generate=true
feels to be the best combination.
Issue openphoto/frontend#343 will make the URL to a created photo which was previously throwing a 500 error redirect to the static asset. This might alleviate the need to pass
generated=true