Friday Facts #277 - GUI progress update

[golfmiketango]

Not sure how big of a redesign of these internals you're already committed to, but I think when building GUI stuff it's important to think really carefully about mods and lua, since crappy mod GUIs are bad for immersion.

The current practice of flattening everything into the single LuaGuiElement class seems to be resulting in quite a hodge-podge there. Also, mod authors have to design everything procedurally ("please put a button named "okbtn" into the "widgetconfiguration" panel at offset (10,30), kthx!") which is not really a recipe for great design (although it performs well and tends to be jank-free which are important attributes).

I feel like mods would have better GUIs if the element "type"-specific attributes were somehow split into separate Lua classes, and GUI {de,}serialization APIs were provided that yielded some sort of human-friendly representation that mirrored the visual containment hierarchy (maybe that's already available somewhere and I'm just not thinking of it?)

Anyhow just thinking out loud ... so to type

[MirandaStreeter]

Am I the only one who is actually really looking forward to to the UX overhaul? It's been the biggest obstacle I've seen to introducing new players to the game.

As well, SCRUM/Agile development isn't necessarily the best for every project. Games as a service has its drawbacks. And to quite an extent, they're already getting iterative feedback with these yearly/semiannual releases.

Also, suggestions like "Add a toggle for the old GUI" might delay the release even further at this point. Development on the new GUI might have involved a lot of refactoring and underlying rendering tweaks that won't play nice with the old one, and maintaining two separate branches to appease two separate player bases can grind everything to a halt.

Patience is a virtue. Every release so far has been well worth the wait, and the next one should be no exception.

[Serenity]

Am I the only one who is actually really looking forward to to the UX overhaul?

I like the new UI. It looks a lot more polished and professional. Definitely something that should be done before the final release. I just don't think it's a huge priority right now.

Though I guess they ripped out so much of the foundations that they are committed now and can't release anything with the old UI

[Jap2.0]

Am I the only one who is actually really looking forward to to the UX overhaul?

I like the new UI. It looks a lot more polished and professional. Definitely something that should be done before the final release. I just don't think it's a huge priority right now.

I mean, they've been saying that the next release would be the final release for several years now.

[usafphoenix]

Wave Defense games and RTS-style games benefit more from diverging technology trees / CHOICES for upgrade path more than they benefit from variability in map generation. While having procedurally generated maps can be good for replayability, the map generating method is a minor thing, and a well-made preset map (like starcraft) is totally fine. What matters is having variety in biter AI and variability in options for defense and attack equipment.
Games like starcraft, age of empires warcraft have vastly more combat units, some of their technology systems LOCK/block certain techs if a different progression path is chosen. THIS would make wave defense style scenario 100x better. Increases in variability of biter types and AI and a technology / research system that made the player choose: Do i want Shotguns or sniper rifle? Laser turret or biter-seeking landmines? Self-repairing wall or electrified fences? You can only choose one. Which will it be?

the game thusly has not been designed to "limit" use of one sort of item if another is chosen, but for an RTS game, this is a powerful mechanic. the mechanic of choice and consequence. There are mods that add different biter types with different immunities, and mods that add different weapons....but you can eventually get all the items in the game. the map matters very little (with the exception of occassionally finding great sections of mountains that can be used to snipe biters that can't reach the turrets by any direct means.)

AI improvements in biter tactics: How they attack the base, do they always make a b-line, or do they split a larger force into two groups, have one group circle around behind the base and attack from two directions at once? When met with resistance biters don't flee, they go to their death (mods change their behaviour and can have them flee, regroup, or "rally-cry" for reinforcements. they also tend to stick in a swarm. What if they broke through a wall, and decided to fan out? did the player place enough defenses inside their base to handle such a tactic? No? well, better act fast then, fellow human, cause these aliens have come to wreck yo' shit.

TL;DR: choice and consequences for technology and item progression + varying upgrades that lock when a different tier/item path is chosen + improved and varying biter tactics and biter types + different defense and attack weapons/vehicles/defenses would improve a wave defense scenario 100x more than the decision to have procedural map vs consistant map.

[Zool]

Getting 4K screens is affordable indeed. Getting a good 4k screen is terribly expensive. By good, I mean with relatively accurate colors, high contrast, homogenous brightness, no clouding, good viewing angles, ... The sub 4K market still has a good future awaiting it. I think it would be a bad move to focus too much on developing 4k oriented stuff.

I started 4 years ago with an 28“ Dell 4K screen that only could display 4k/30Hz, which was quite annoying in the end ... also a 28“ screen prooved to be slightly small for many applications.
So now, I upgraded to 2 x Samsung u32h850umu, which works absolutely charming. They actually cost around 440€ for 4k/60Hz. 60Hz might be low for Shooter-Style games, but for Factorio its all you ever need! And finding 4K screens with more then 60Hz might be challanging for both you, your budget and your gfx-Card

[Serenity]

I mean, they've been saying that the next release would be the final release for several years now.

By now it's clear that there will be a 0.18

[Jap2.0]

I mean, they've been saying that the next release would be the final release for several years now.

By now it's clear that there will be a 0.18

Well, yes. They generally push 1.0 back to the next version before releasing the current one.

[ske]

choice and consequences

Well said. Choice is more about what you don't choose instead of what you choose. If you can choose everything you really don't need to make any choice at all. Some kind of hard limits are required to enforce choices.

[Mike5000]

While you're revamping the GUI, how about an option to click on an alarm to silence it for 5 minutes?

When my low-oil alarm goes off it's going to take me a while to fix it. The continuous alarm is annoying and masks any new alarms such as a biter invasion.

[Chaoseed]

However I have some indecision within myself, and there are several advantages and disadvantages between a handmade custom maps and randomly generated worlds.

Why not both? Include a few handmade Wave Defense maps, also include the option to randomly generate maps. Every so often, release another carefully made map, to keep player interest high. If you want to get really fancy, you can include achievements for winning in both modes.

[Light]

While you're revamping the GUI, how about an option to click on an alarm to silence it for 5 minutes?

When my low-oil alarm goes off it's going to take me a while to fix it. The continuous alarm is annoying and masks any new alarms such as a biter invasion.

This can be remedied with additional circuit logic.

I'd highly recommend looking up guides on creating clocks and timers so there's a short pause before the alarm goes off again if the condition remains true. It can also greatly assist in other tasks which could be time sensitive, such as running belts for only short bursts of time and so on.

[Unknow0059]

I haven't played many of the scenarios as i haven't finished the campaign yet.
I don't think a random map would be more fun than a hand-crafted one, but i also think there should be variety, in which case i'd definitely play more than if it was only one or a couple maps.

Maybe we could have both.

[DOSorDIE]

Dont wait too long with 0.17 ... on 29.02.2019 "Anno 1800" will be released.

[Oktokolo]

Dont wait too long with 0.17 ... on 29.02.2019 "Anno 1800" will be released.

That shouldn'd be a problem as Factorio is in another genre than the Anno series.

[ske]

Dont wait too long with 0.17 ... on 29.02.2019 "Anno 1800" will be released.

That shouldn'd be a problem as Factorio is in another genre than the Anno series.

They will soon enter the industrial age and we know how it goes from there.

[nljr]

About 25 years ago, I wrote a little webapp (In Perl. Those were the days!) to do something a lot like this:



It let developers check boxes to show progress and stored it on a server where people outside the team could view it.

How would the dev team feel about a tool to expose their progress that way?

[JackGruff]

Could you perhaps tell us what other GUI frameworks lacked? I've been thinking up to this point, why did you not just use Qt and theme it to your liking?

It is cool that you are inventing your own wheel, and doing it well

[CreSerty]

The strange location of the reset button, because there is an empty line. You can do it easier. Here are two examples.


Another thought. Pressing the middle mouse button resets only the selected setting. I find this very useful.

[Koub]

This place is not publicly available