I coded fancy little server modules that did all kinds of funky things, like adding new special abilities and customization options to ships.
But what I did, too, was run my own private server alongside the "live", (highly) populated server. Before I stuck a newly compiled .so on the server that would more than likely have a few (minor) bugs (because hacking in new content for an ancient game is not the easiest thing to do), I plopped it on my own server. I then tested whether things broke. Things always broke, because the server software had two dozen threads running simultaneously and there always was some unexpected multithreading hiccup. I sometimes even invited over a few players to see if it would break in a less controlled environment.
Then, once I had faith that the module would run smoothly, I would finally drop the new stuff on the main server. Then I'd wait until population was at its lowest point in the early morning, restart the server, and rejoice! New stuff.
Of course, the community would whine at the clearly very poorly balanced or poorly thought out new content, and then after a week or three things would settle down as people slowly realized it actually wasn't the final nail.
A year or two after I quit this old game, I started up Discovery Freelancer. I undocked my Gallic snub.
[Some lines
of random
nonsense
for dramatic
effect]
And it had 10% core.
Maybe testing things (i.e. undocking at least one Gallic ship before you add fancy Gallic engine shipnerfs) before dropping them on the main server is something Disco's developers should consider doing.
Posts: 3,558
Threads: 107
Joined: May 2012
Staff roles: Balance Dev
(01-26-2014, 10:32 PM)Kazinsal Wrote: Technerf bugs regarding incorrect core levels can be fixed literally as soon as an admin logs into the server's remote desktop.
No big deal.
"It's easy to fix if we screw up, so why test it beforehand?"