amosolov.bsky.social
Developer of upcoming indie sandbox rpg Starsector.
fractalsoftworks.com
317 posts
2,348 followers
19 following
Getting Started
Active Commenter
comment in response to
post
Not really something I've been thinking about! I'm not sure how well the game would play on a smaller screen anyway, but I know people have gotten it to work on the Steam Deck and that has support for re-mapping controls
comment in response to
post
Hahah, better late than never? <3
Glad you're enjoying it!
comment in response to
post
Well *I* didn't, but my understanding is @dgbaumgart.bsky.social draws them at double the size but not detailed, then shrinks them at some point and adds in some detail.
comment in response to
post
A lot of the game steers the player towards conflict/battles in one way or another; this is just another example of that.
comment in response to
post
Hi! Those don't actually exist, except for the "Hound in hangar" one. The amount of work to draw a full-res one is massive!
comment in response to
post
Not something I was thinking about, no! (I think in-fiction you could easily make arguments either way, in all honesty.)
comment in response to
post
The size isn't the main thing here, there are efficient and inefficient ways to do it - what triggers and conditions are used and how. I'm not worried about size as a limiting factor, but on the other hand a single rule with a poorly performing condition could make it slow.
comment in response to
post
Gotcha, thank you for expanding on it!
I'm assuming "caravans" are something you can join and get into some fights alongside of, along the way? If so, this has actually come up as an idea for Starsector at various times; some tricky details implementation-wise though!
comment in response to
post
When I take a step back from being so "in it" that's honestly wild to me and I appreciate it so much! <3
comment in response to
post
Thank you, glad we're on the same page there, haha!
comment in response to
post
Massively positive!
Regarding "piracy" specifically, though: that option is always there and easily available and the video didn't really alter that aspect of the equation too much, imo.
And, thank you for your support in buying the game :)
comment in response to
post
The main point is to make sure they can get all of their bombs out during a run; they often weren't able to before!
comment in response to
post
That sounds about right, yeah. The biggest problem is going to be with the Codex and certain tooltips (e.g. for hullmods or industries) crashing there if they're not updated to be aware of what context they're being created for.
comment in response to
post
It shouldn't!
comment in response to
post
Yeah, it's still a thing!
comment in response to
post
Probably? I mean, they use that AI, so they'll be affected by it, but it's all down to details and what one considers "benefits" :)
comment in response to
post
Would be curious to hear what they are!
comment in response to
post
Literally never, actually, so that's interesting!
comment in response to
post
Yeah, I could see stuff like that! Though I'd still want to avoid repeatable ones, but "more", perhaps triggered by events, is a possibility.
And an "escalate" option is coming in the next release, to speed things along if desired.
comment in response to
post
Working on it :) "When it's ready" and all that!
comment in response to
post
Yep! The tooltip actually says this but yeah, easy to miss.
comment in response to
post
Hey, one out of two isn't bad :)
comment in response to
post
It only *really* matters for your fleet! It does have some impact on how aggressive the patrols are on the campaign layer but that's veeeery marginal.
comment in response to
post
Ah - they're all permanently resolve-able very intentionally, otherwise it's the sort of thing that, while it can be fun, it can also easily be a chore to keep protecting your colonies against. So it's not meant to be something that just keeps going forever!
comment in response to
post
Hmm, I think officers mostly cover that! Un-officered ships are a good deal less common (depending on the strategy of course) especially as one's fleet develops.
I do get what you mean, though; it's something I'm keeping in mind.
comment in response to
post
Always has been!
comment in response to
post
Since forever :)
comment in response to
post
Yes - or rather, it's the same control for both.
comment in response to
post
That's a possibility, yeah! I'd like to keep the AI more general-purpose than having hardcoded stuff for just one ship, though, and as soon as you do that, there are going to be types of threats it misses. Might even be with just Monitor-specific "what is threatening, actually" code!
comment in response to
post
I definitely replied, but I wouldn't necessarily call this an "answer" :)
comment in response to
post
[REDACTED BY COMSEC]!
comment in response to
post
Hmm, you should be good, still a few things left to do here :)
comment in response to
post
There's definitely the RP aspect of it too, yeah!
comment in response to
post
Oh, I definitely hear you on that!
comment in response to
post
I want to stay away from stuff that tweaks the AI in ways like this; the "try it out and get feedback on how the behavior works" gameplay loop is WAY too long for that to be viable, imo. (See: how long properly testing out AI changes takes, and how easy it is to run afoul of confirmation bias)
comment in response to
post
That's very fair! Pathfinding generally means, well, finding some kind of path. This stuff is more about "finding a direction in which to go" - which, now that I write it out, is totally splitting hairs, haha
comment in response to
post
Nothing good that's for sure :)
comment in response to
post
Hmm - I think once a colony hits size 6 (faster because of LM), choosing whether to keep the bonuses or go for an extra industry feels alright. At any rate, I don't think I'd want to add industries *because* of this, that feels like too much "solution" for the "problem", if that makes sense?
comment in response to
post
Yeah, they'll still shoot and use shields against it etc, it just won't ever be a primary target.
comment in response to
post
What you're calling "pathfinding", yeah - I'm not sure I'd *quite* call it that, but close enough :)
comment in response to
post
It's an interesting idea, but that might just lead to Ion Cannon Monitor builds :)
comment in response to
post
And making an AI that handles these kinds of conflicting priorities well is very difficult, so imo it's just a no-go for those reasons. It sounds cool! But in practice I think it would change combat into something less interesting
comment in response to
post
The thing is, if you add appreciable hazards like this to the battlefield, you really need to make sure the AI can handle them *well*. Otherwise it turns into a game of "friendly ship blew up for no good reason" and "cheesing the enemy AI to blow up for no good reason"
comment in response to
post
Hmm, that's a maybe!
comment in response to
post
(Ahem,)
In every case, it'd be better to see what the human put into the prompt box so I could know their intention with the text rather than trying to fish it out of the generated slop.
Often it's just "hype this garbage up", to be fair; and it'd be nice to just get that done with and move on.