61edb3eb95
- Integrates Volundr's btime library and associated process scheduler changes. - btime is implemented separately from the process scheduler, as precise time measurement is also useful elsewhere. - `TimeOfHour` is no longer internally throttled; throttling is instead done by `SCHECK`. - If btime's `gettime` cannot be called at world startup, an error will be output and the world will stop. - Retains the change to schedule processes according to game time, rather than real time. - Removes the (now unused) update queue files. - Removes the process scheduler testing files. - These are standalone tests for the process scheduler, completely unrelated to its use in the full codebase. We never used them. - Moves the process scheduler defines into __DEFINES. - Makes the lighting process run once before the round starts. - Renames `scheck` to `sleepCheck`, to ensure any code that tries to use `scheck` will fail to compile. - Adds `SCHECK` and `SCHECK_EVERY` macros that skip calling `sleepCheck` entirely until a specified number of `SCHECK`s (50 by default) have been called. - Makes most processes iterate using their `last_object` variable, allowing hang recovery to show the type that caused the hang. - Makes processes output an error when they filter out a type they refuse to process. - Rolls the recently-added alarm subsystem into the alarm process. - Removes the now unused subsystems code.
ProcessScheduler
A Goonstation release, maintained by Volundr
##SUMMARY
This is a mostly self-contained, fairly well-documented implementation of the main loop process architecture in use in Goonstation.
##LICENSE
This work is released under the following licenses.
This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License. The complete text of this license is included in the file LICENSE-CC-BY-NC.
This work is licensed under the Affero General Public License 3.0. The complete text of the license is included in the file LICENSE-AGPL.
##INSTALLATION
To integrate the process scheduler to your codebase, you will not need anything except the contents of the core/ folder. The rest of the project is simply for testing and to provide an example for the process scheduler code.
Test project setup
To compile and run the test project, you will require:
- node.js
- BYOND
Clone the repository to a path of your choosing, then change directory to it and execute:
npm install -g
bower install -g
Then you can either compile with DM or open the DM environment in DreamMaker and compile/run from there.
##USAGE
###BASICS To use the process scheduler in your SS13 codebase, you'll need:
- core/_defines.dm
- core/_stubs.dm
- core/process.dm
- core/processScheduler.dm
- core/processScheduler.js
- core/updateQueue.dm
- core/updateQueueWorker.dm
To integrate, you can copy the contents of _defines.dm into your global defines file. Most ss13 codebases already have the code from _stubs.dm.
The processScheduler is intended as a replacement for the old master_controller from r4407 and other fork codebases. To implement it, you need only to add the source files to your DM environment, and add the following code into world.New, above where the old master_controller is initialized.
processScheduler = new
processScheduler.setup()
processScheduler.start()
The processScheduler will automatically find all subtypes of process, and begin processing them.
The interface code in test/processSchedulerView.dm is simply an example frontend, and can easily be rebuilt to use other styles, and/or render simple html without using javascript for refreshing the panel and killing processes.
###DETAILS
To implement a process, you have two options:
- Implement a raw loop-style processor
- Implement an updateQueue processor
There are clear examples of both of these paradigms in the code provided. Both styles are valid, but for processes that are just a loop calling an update proc on a bunch of objects, you should use the updateQueue.
The updateQueue works by spawn(0)'ing your specified update proc, but it only puts one instance in the scheduler at a time. Examine the code for more details. The overall effect of this is that it doesn't block, and it lets update loops work concurrently. It enables a much smoother user experience.
##Contributing
I welcome pull requests and issue reports, and will try to merge/fix them as I have time.
Licensing for code submitted via PR:
By submitting a pull request, you agree to release all original code submitted in the pull request under the MIT License. You also agree that any code submitted is either your own work, or is public domain or under a equally or less restrictive license than the MIT license, or that you have the express written permission of the authors of the submitted code to submit the pull request.