Difference between revisions of "Code Updates"
Jump to navigation
Jump to search
(Created page with "Code Updates for 2020 are listed below. = January 2020 = * Memory leak(s) in accounts fixed. ** This should allow us to not have to reboot weekly any more. Ancient memory le...") |
|||
Line 1: | Line 1: | ||
− | Code Updates for 2020 are listed below. | + | Code Updates for 2020 are listed below. The full list for 2020 can be found at [[2020_Articles|2020 Articles]]. |
= January 2020 = | = January 2020 = |
Revision as of 19:40, 23 April 2020
Code Updates for 2020 are listed below. The full list for 2020 can be found at 2020 Articles.
January 2020
- Memory leak(s) in accounts fixed.
- This should allow us to not have to reboot weekly any more. Ancient memory leaks revealed by the 'you're always your own friend' code have been squashed.
- Group Display
- In group display, @n will always display the name of a character. It will not display descriptions, pretitles or post-titles
- In group display and group commands, name will always be displayed without respect to visibility status
- The end-of-line coloring issue has been fixed with the 'what skills can you now get' code
- A target marked with Mark Target will skip their juggle check 30% of the time
- Some score displays will now have more information.
- A previous field indicator was not pulling data properly.
- Are these used in current score displays? good question, but fixing bugs that might not be bugs is one of my scientific specialties.
- Builders have some increased control over how mobs can set their armor and hitroll values. No one's using these yet.
- Legend now compiles cleanly under both clang and gcc. Squee!
- Builders now have more control over how many attacks a mob has, and can even set a mob to have 0 attacks! Also no one is using this yet.
- Builders have the ability to inherit mob stats in real time.
- This means that they can adapt mobs to specific levels after the mob has been loaded without a massive slew of commands.
- Also no one is using this yet. But expect it to be used in instances eventually for difficulty scaling.
- Much churning has happened trying to get code to compile in a more optimized fashion and still... what's the word for it... work. That's it.
- I'm not entirely sure how we ended up breaking code compiling with optimizations turned on, but there's an effort to fix it.
- Tons and tons of warnings identified by a new code analysis tool have been fixed.