Currently the Slack client in use is smallwins slack one. Just wondering why not switch to the more official slackapi node sdk.
We like how thin the smallwins client is. At the time at least (and I think is still the case), the official node client was pretty heavy including a bunch of built in caching options and things we think are better left outside of the core library.
It could be worth another look soon, but would be a pretty big breaking change for anyone who relies on slapp.client.
I do agree that the smallwins client is lighter but since this is all used on the server side, size shouldn't be too big of an issue. But, in my own projects, I don't rely on the smallwins client exposed through slapp - I instead pull in the slack one and just query through that. This is especially because the slack one is more up to date in slack features exposed (chat.postEphemeral for example) - with the last update being 5 days ago vs 2 months ago for smallwins. There's clearly a more vested interest now in the slack node sdk being up to date and maintained as opposed to where it was when this project decided on smallwins.
I do agree that it would be a breaking change in switching over - it would be entirely new syntax and everything but I would be willing to take a stab at it if this is something that's in the direction that slapp decides to go.
馃憢 I think there are a few benefits to using the slackapi/node-slack-sdk package in the context of slapp that haven't been mentioned yet:
However, it should be acknowledged that it's still the case that the official client is larger than the smallwins/slack client and it includes additional RTM and Incoming Webhook clients that most developers won't have a reason to use.
I don't anticipate the migration cost being too high for developers since the method calling conventions are the same in the smallwins/slack and slackapi/node-slack-sdk clients, but curious to hear other's thoughts.
Most helpful comment
I do agree that the smallwins client is lighter but since this is all used on the server side, size shouldn't be too big of an issue. But, in my own projects, I don't rely on the smallwins client exposed through slapp - I instead pull in the slack one and just query through that. This is especially because the slack one is more up to date in slack features exposed (chat.postEphemeral for example) - with the last update being 5 days ago vs 2 months ago for smallwins. There's clearly a more vested interest now in the slack node sdk being up to date and maintained as opposed to where it was when this project decided on smallwins.
I do agree that it would be a breaking change in switching over - it would be entirely new syntax and everything but I would be willing to take a stab at it if this is something that's in the direction that slapp decides to go.