57 | | * The server tries to create a running instance of that party |
58 | | * There are two possible responses from the party server: |
59 | | * If the party was launched sucessfully: the http get returns a websocket address (ws:...) as plain text on which the party can be contacted. |
60 | | * If the party could not be launched (eg, not enough resources) the server returns error code 503 "Service Unavailable". The 503 includes a message like {{{retry later at 1556631286760}}}. The message is kept human readable but the last element of the message must be a unix timestamp containing the time at which to retry. To create this message, the server can either create a default (e.g. let the client retry after a fixed set time) or actually use the estimated termination times of the parties. The current available server does the latter. |
| 57 | * If the party was launched sucessfully: the http get returns a websocket address (ws:...) as plain text on which the party can be contacted. |
| 58 | * If the party could not be launched (eg, not enough resources) the server returns error code 503 "Service Unavailable". The 503 includes a message like {{{retry later at 1556631286760}}}. The message is kept human readable but the last element of the message must be a unix timestamp containing the time at which to retry. To create this message, the server can either create a default (e.g. let the client retry after a fixed set time) or actually use the estimated termination times of the parties. The current available server does the latter. |