Where Tackling the jira ticket ELEMENTS-1169 / NXP-28812 we found ourselves with different things, that should be improved:
- The nuxeo-uploader-behavior.js acts as a Proxy and delegates the behaviour to the JS object (see the notion of instance). currently we have two JS objects that can hold the upload behaviour the default one nuxeo-uploader-behavior.js and the S3 direct uploader here. each new uploader should register itself to be available in web-ui.
The problem is the behaviour exposes the BatchId, the underline JS Object also have the batchId. and sometimes the batchId is not available as soon as we initiate the uploader (this was fix by ELEMENTS-1169) but we still manage two batchId (the one of the behaviour and the second of the underline JS Object) . We should simplify things:
- Why we keep the outer batch id (the one of the behaviour), we should use the underline once.
- This batchId should be available asap, for the default one, perhaps we should update the Nuxeo JS Client to get the batchID once we initiate it (add new parameters....). Already the Nuxeo JS client has a private method here
- is related to
-
NXP-28812 Terminate multipart upload to S3 when user cancels the blob upload
- Resolved
-
ELEMENTS-1185 Set file error when upload fails
- Resolved
-
ELEMENTS-1169 Terminate multipart upload when user cancels the blob upload
- Resolved