<div dir="ltr"><div><div><div><div><div>Hi guys, and Jonathan more specifically,<br></div>I'm pulling out all the j3m retrieval and parsing logic out of the JS framework on the front end, and moving it to python. With the goal of unloading the client side and having the option of persisting the separate j3m artifacts.<br><br></div>so far it's all going in here <span class=""><a href="https://github.com/harlo/InformaFrontend/blob/master/v2j3m.py" class="" id="447f82e2b682e603dfce6e2aad587749-6fae41219bb0273944b9af473349a6575524efdc" title="v2j3m.py">v2j3m.py</a></span> for testing, but will be broken up into more modular structure.<br><br></div>Question: the original plan was to make a lightweight REST api making each j3m artifact (mapping to an icon or a value or a graph) available via a GET, so it can be called asyncronously by the presenttinng page.<br><br></div>but while making a test pageĆ <a href="http://localhost:8888/v2j3mview/">http://localhost:8888/v2j3mview/</a><doc -id> ( e,g. <a href="http://localhost:8888/v2j3mview/22f9261d2bee2058379095d8d9ea73b1">http://localhost:8888/v2j3mview/22f9261d2bee2058379095d8d9ea73b1</a>), i realised the other option i to have tornado run all the functions in parallel and render the parts to the page when each becomes ready (since it has all this cool async and multithread stuff built in to it). <br><br></div>So Jonanthan - which would you rather? i can wrap each part in what ever html template is necessary. Or do the requests yourself?<br></div>