Less than a week until the giant collection of robot hacks will be at your metallic fingertips :)
I've played the first few real runs to ever actually use the system and it's been a lot of fun, can't wait to share it with everyone! Here's the very first run of a hacker-to-be after having raided the first garrison you can find (-8):
The Beta 7 release date should be about 9/4-ish.
This week in particular I worked on a variety of little last-minute features and fixes...
You Die, Now Also You and You
"Gunslinging" is now guaranteed. It's the ability to attack multiple targets in succession if using a gun-only volley (and the first target is destroy), and it used to only happen 50% of the time, further modified by utilities.
Guaranteeing this effect was already possible later in a run with some targeting computers, but now it'll be a reliable tactic even early on, and make a difference for builds that start out by adding extra weapon slots. In the background I've always been working on plans to increase the value of non-utility slots, especially weapons, so this is one way to achieve that (although it took a player request to point this one out, thanks Jivya :D). This change also has the nice advantage of simplifying the whole secondary targeting thing.
For consistency (and because giving melee builds extra situational buffs is a long-term trend :P), multiwielding melee builds are also guaranteed to switch to additional targets after taking out the first, assuming there are follow-up attacks.
On the topic of melee combat, it turns out that some knockback code I added earlier made it possible to push robots back multiple spaces while multiwielding impact weapons, which doesn't make much sense so I had to do something about that xD. It was pretty funny, and suddenly "safer" to knock targets back into explosive machines.
One Key to Rule Them All
While technically I think the 'a' key rules them all since it handles lots of part management automatically, you can never have too many optional ways to manage your inventory! Some players would prefer a fully modal approach to item interaction like that found in other roguelikes, rather than requiring the use of Ctrl/Shift/Alt key modifiers. (Yes we're talking keyboard play here--mouse users who can do everything with the almighty cursor you can skip to the next section :P)
The new 'p' key is the answer. It opens a menu that gives access to any part-related action, so
-> [action] -> [letter or number of part]. Here's the reminder that appears on entering the mode:
Along with this update, the menu and other modal state identifiers that have always appeared across the top of the map view now displace the temporarily interface message if there is one, like so:
Earlier any message would temporarily overlap with the mode name, making it harder to tell which mode was active, or if one was active at all. It didn't seem to bother anyone, but then once familiar with the functions you won't be in any given mode for long at all anyway (plus there are sound cues). Still worth polishing up, though.
Autobots, Roll Out!
With everything pretty much in order, this week it was also time for automated testing. This is where bugs (especially weird emergent stuff) can be caught simply by having a large volume of random actions play out, more specifically in this case by pitting random robot factions against one another.
You can divide bugs into two main categories: Those caused by player interaction with the world which can be identified by observation upon doing XYZ, and those caused by some unknown source who-knows-where on the map. Automated testing can catch some of the former, but most important to catch before release are the latter type because inexplicable crashes might result in a lot of misdirection when it comes to players trying to report bugs without having anything to go on (as long as it's a crash the logs will include a trace to shed some light on the issue, but that might not always be enough).
Of course, that's not to mention if something can be caught right here and now via automation it's way better to let the computer do the work than bugging players! (and later me with reports that distract from progress elsewhere xD)
With so many bots and systems out there on any given Cogmind map there are tons of things going on behind the scenes, so the chance for hidden emergent bugs is high. I used to always find lots of stuff via automated testing before each update, although with Beta 7 this is the first time ever that autotesting turned up zero issues. That was nice :). It's likely because a huge chunk of the upcoming release is player-initiated hacks rather than automated systems, and thus outside the scope of autotests. Myself and others have tested all the hacks and resolved any known issues, but there's probably still room for emergent craziness to show itself. If we really need it I'll be releasing a 7.1 to correct anything serious.
For testing I generally just run four instances at a time, one for each physical core on my laptop, but this time I upped the instance count to nine just for fun. That's about 90% CPU usage right there.
I also recently added yet more on-map alerts for Beta 7, so while these tests are going on there sure are a lot of red messages piling up as turns race by.
Be with you in a few days for Beta 7, which has of course been dubbed "Hack the Planet" :D
Cogmind will be 10% off next week for the new release.
A note about SITREPs: These are regular progress updates where I share features coming to the next release, but aren't actually in game yet. Public releases get their own dedicated news announcement and changelogs/release notes, for example Beta 6 here.
There may be other relevant discussion of this SITREP on the GSG forums or r/Cogmind.