This is a simple Stack Overflow / Stack Exchange chat bot which keeps an eye on a selected list of rooms, and makes sure they are not frozen.
The threshold is currently set to 12 days since the previous message, to leave some leeway for possible accidents (the freeze happens after 14 days of inactivity).
There is no interactive interface to the bot; create a pull request if you would like to add a room to Sloshy's watch list.
Sloshy's operation is driven by a simple YAML file,
sloshy.yaml.
This is basically a human-readable text file, though the YAML format imposes some structuring conventions to also ensure that the file is properly machine-readable. For more information, perhaps see Wikipedia's YAML article and/or https://yaml.org/; but you really don't need to be very familiar with the format to make simple changes.
| Configuration file format changed in PR #23 |
|---|
In the YAML configuration file, add the server's name
to the servers key if it is missing (though that's unlikely)
and create a new rooms entry with information about the room
you want to add;
- The
idis the room's numeric identifier. This (together with the server's name) is the way Sloshy finds the room. - The
namefield is just a display string for Sloshy's status messages. It can be anything, but should describe the room reasonably unambiguously to humans. - The
contactfield indicates who requested the room to be added, in case we would need to touch base with you later on to assess whether the room still needs to be on the list, etc. The format should be your user name, followed by the network account id in round parentheses. The network id number can be found by clicking "Network profile" on any profile page for your account. (For example, Sloshy is user 16115299 on Stack Overflow, but network user 21818820.)
In brief, if your chatroom's URL is https://chat.stackexchange.com/rooms/12345/my-room, and your network account ID is 123456789, you would add
servers:
chat.stackexchange.com:
rooms:
- contact: your name (123456789)
id: 12345
name: "my room's name"
sloshy_id: 514718in the configuration file.
(This is showing the complete YAML structure;
the servers top-level key
and the server chat.stackexchange.com with the corresponding sloshy_id
obviously already exist in the file.
Thus the three lines starting from the - contact: line
are what you would normally add for a new room.)
Perhaps notice also that there are two main chat servers with very similar names, where one contains "overflow" and the other contains "exchange". You want to make sure you add the room in the section belonging to the correct server for this particular room.
The previous section explains what technically needs to go into your pull request. Creating the required code change for a pull request should be easy if you are familiar with the basics of GitHub's PR process and YAML (or at the very least, making changes to a text file. :-)
Before you propose a change,
-
Please make sure you have the approval of the room's users, typically by asking a room owner or a moderator for approval.
-
Please also separately make sure that Sloshy has write access to any privileged room you want to add to the list.
For adding an individual room owned by yourself, the process is pretty relaxed; just say so in the description of the pull request which you fill in when you create it. (As always, including a good rationale for the change is considered good practice; pull requests with no information about the rationale and/or context can and will be rejected summarily, not just by the Sloshy maintainer, but more generally.)
If you add more rooms (in a single pull request, or over time), getting moderator approval for individual rooms may not be sufficient. Then, it's probably better if you can anchor the decision with a discussion about the guiding motivation for adding these rooms to Sloshy's configuration (probably with a moderator; but if you can adequately poll the users of those rooms and other pertinent stakeholders e.g. with a meta post about this topic, and reach a reasonable consensus among them, that would probably be fine, too).
There is an option --migrate which accepts a YAML file argument
and rewrites it from the old schema to the new.
Example:
python3 sloshy.py --migrate sloshy.yamlThe old configuration file format did not have a schema identifier. The migration code simply assumes that your file uses the old schema.
Besides servers, there are some other configuration options
which can be specified in the YAML file.
For the most part, these are undocumented options
for use in internal testing,
but the following are expected to remain stable and supported:
schema: top-level key which identifies the YAML schema version (currently 20211215, corresponding to the date 2021-12-15).local: boolean; set totrueto disable connecting to chat. Sloshy will still need to open network connections to fetch chat transcripts, but will not emit any chat messages.nodename: string; set to the name you would like to display as Sloshy's location in the startup message.threshold: expression indicating after how much inactivity to emit an anti-freeze message to a room. The default is 12 days; set to a different number to override. (See the source for advanced and possibly unstable options.)
There is a simple Github Action which runs nightly.
Pushing a new version to the master branch deploys it.