Factorio Facts #31

[kovarex]

Good news this week: https://www.factorio.com/blog/post/fff-31

[BootieTrap]

Sweet! Can't wait! Big badaboom! (c;

[BoinKlasik]

That spike really is nice to see.

[yareczek]

I was reading about your problem on stackexchange and point about moving from 2d to 3d isn't so bad.
Well I am not so good at graphics so I am probably wrong about this but if you mix this 2 together it may be good solution.
Terrain generation is pretty good so let it be 2D and all objects make 3D render. That may look pretty well.

I imagine this can be painful but better now then later. Number of entities will rise exponentially so how far 3d optimized graphic cards go with 2d graphic?

I'd love to help you just for fun (and programming like old times) but my work is 24/7 stand-by with 8-12h normal work I'll try with some hardcore mods in free time

[TNSe]

Sadly I do not know enough about the internals of allegro, but if you are able to change the rendering pipeline at all, I would look into the following solutions:

- Quick solution: Texture compression. Either in form of DXT1/S3TC, or in the form of going to RGBA5551/RGBA4444. Add a color quality slider in options, which basically converts textures from RGBA8888 to RGBA4444/DXT5 (Or DXT1 if Alpha is not important?), this is a quick operation on modern PC's, and will atleast halve your memory requirements, at a loss of rendering quality. It would be a lot more work to introduce an 8bit option, which converts textures from RGBA8888 to a palettized 8 bit value. Each slot in the texture would have its own 256 RGBA lookup table that the shader has to convert. This means 2 texture lookups per rendered fragment. It would reduce memory usage from 4x8192x8192 to 1x8192x8192 + 4x256 * items in the texture atlas. (If they are 64x64 each, from 256meg to 80meg.)

- Cheaty solution: Simply render a placeholder if the texture atlas is not yet updated for current frame... Since this will probably not kick in before late in the game, people should more or less know what that textureless sprite is, and it will eventually display correctly. Probably best implemented as an option for a tradeoff in framerate smoothness... You would of course need to implement some kind of routine for detecting what to toss out and into the texture.

- Correct solution: But this may be too slow, or what you are already doing, or considering doing: glTexSubImage2D, that is, if you have access to this function under the requirements to OpenGL you have set.

- Irrelevant solution(?): Send multiple textures into the shader, and picking the correct texture at vertex shader level. This should increase the available amount of textures (atlases?) by the number of texture units, most modern cards (those that are relevant to play the game), support multitexturing. (I guess this solution is out the window since you are probably fighting the 512meg limit?)

As said earlier, don't know the internals of allegro, but this is mostly general 2D/3D-programming stuff.

[Drury]

Just to be sure - this also means Greenlight campaign starting May 1st, right?

[kovarex]

Just to be sure - this also means Greenlight campaign starting May 1st, right?

About that ... we decided to give it some more time

[Drury]

Oh, cool.

What else needs to be done, then? I was under impression it's all ready except for trailer.

[kovarex]

The more we discuss it, the more we feel the game should be finished first for the steam. I personally felt, that with every update, the previous one was much worse experience.

• 0.9 - How could anyone play it without the blueprints automated deconstruction, the terrain decoratives and oil industry before?
• 0.8 - How could anyone play it without automated repair robots, roboports, with the ugly old terrain and the old editor?
• 0.7 - How could anyone play it without the pollution system old enemies without evolving and no fight possibilities?
• 0.6 - How could anyone play it without the the map, personal requests and modules?
• 0.5 - How could anyone play it with the ugly old gui style and without item groups?
• 0.4 - How could anyone play it without the automated train transportation?
• 0.3 - They had no walls, ugly trees, funny laser turret graphics and solar panel pictures, it is looking like picture of small child really.
• 0.2 - Hey how could anyone play it when there is no research, logistic robots and underground belts?
• 0.1 - How could anyone play it when it was not even released before?

Did you get my point? If we continue with this few more iterations, there will be plenty of people, that would never play the game in the current state, but would with the future changes.

[Drury]

Hmm, that's kind of weird reasoning... Not to imply I disrespect you, but I think you're overestimating general public's standards here.

They will play anything. Billions of people will tap the screen of their tablets and smartphones to make a stolen sprite of a bird boost up in such a way that it doesn't hit a stolen Super Mario pipe sprite and enjoy themselves. Factorio is already the exact opposite of that. I know, it used to look worse, it used to be barely presentable, people used to avoid it because of the placeholder graphics and lack of content. But those times are gone, Factorio is now a solid game, not yet finished, but already very playable. Much more so than most Early Access stuff, even if Valve starts moderating it I don't think you'd have much trouble getting your game through. It is unique, it is a huge time sink, and it's being developed very professionally. In fact, as a developer team, I can say without hesitation that you guys do a better job than the Team Fortress 2 team is doing. Steam support would just give you more room for realization in every possible way.

Your game has reached that stage where it only needed that "first nudge" Like a snowball that is sent rolling downhill by someone who has the power to make it happen. Minecraft reached this point in the summer of 2010. Quill18 had the honor of doing it to Factorio a few days ago. The floodgates are open and it's a matter of time before the whole thing blows up and becomes the next big indie hit, and don't you try telling me opposite. It won't get as big as Minecraft, but it's already getting widely recognized and it won't stop at that. You don't really have anyone saying "wow this is crap, make the game good and then come back" (yet you act as such...). Not even a silent minority. There's no reason to hesitate, you can set it off.

Rev up that engine and fly. Or slow down, return to your house and let your Ferrari sit in your garage for 10 years. I know the video is about ideas for a game and you already have a game, but you're acting exactly same.

[Nova]

I understand the thoughs of the developers. Hearing players say they don't like the game because of shortcomings or bugs is not nice. Even less when the player just has no luck and discovers nearly every bug the game has. Every one of this comments is much worse than one converse comment about how good the game is.
Did someone else read the comments about "Banished"?

As Simon says, please don’t let all the overwhelming praise fool you. I ALMOST LOVE your game, but as it is, I find it almost unplayable. farmers leave a bushel of corn on the field at the end of harvest, nobody will touch it, and next season nobody will plant there. builders ignore buildings they are assigned to and gather resources instead. Please let me love your game! Teach the people how to prioritize.

There are several bugs that should have been hammered out before charging $20 a pop for the game. More so – hiring someone becomes “I don’t wanna but I don’t wanna work either” is really, really poor messaging/branding. For one, you could easily hire an independent contractor. This game isn’t so complicated in coding that you have to “bring up to speed” anyone beyond what you want from the game. A contractor also means no expanding. The fact you made (at least) $260,000 from an unfinished game and make excuses for not working on it is beyond me.

That are comments about the game "Banished", a nice medieval colonisation game (without any battles). You start with a few families and have to create a big town of settlers. (Getting food and firewood is the most important task.) The game is finished. There are a few bugs and some inbalanced things, but all in all it's nicely playable. Only one person has created that game in 3 years, and i'm pretty sure this comments hurt.

"Oh, someone works 8 hours a day on a game? Nope, I want my patches, here and now! What? Solving many bugs just takes more time than a few hours for one person? EXCUSES! Unfinished Game! Greedy douchebag!"
Come on, the game has far less bugs than nearly any new title. Battle Field 4 (called Bugfield), Sim City (well, the newest) did have much more bugs and were nearly unplayable in the beginning. (Well, didn't test that myself.) Huge companies with big testing departments didn't find all bugs, but you expect it from one person? The game is not huge, but pretty nice and solid. (The best part: Most of the bugs did not appear on my side. Food left on the farm? Never happened. Game Crash? Never heard of it. The only problem for me were the big stockpiles in the houses and the death marches.) (Yes, i dind't either have the "houses burn but nobody extingushes" bug, but I nearly never had a fire, so this problem doesn't count.)

[ssilk]

In my company I'm called the bug-finder. This game has no bugs, which makes it unplayable. And the only time I had a problem with bugs, it was solved within a week. What else can be awaited?

[just_dont]

The only thing the game could do before hitting Steam EA -- are some accessibility improvements. I.e. prettify the GUI a bit; and ensure that the tutorials explain everything properly and in full (the "help" subforum provides a great feedback to catch every no-so-clear moment). More often than anything else, people tend to whine a lot if they don't understand something crucial about the game.

Everything else is already great enough.

[muzzy]

I still believe the game could use some more endgame and replayability value before the full release. I keep starting new games every now and then, only to realize it always follows the exact same formula and every game feels the same, the exact same way of playing works every time. It gets depressing.

[Drury]

One more thing I forgot to mention - Steam =/= full release.

All Steam is, is basically a way to open floodgates. A flood is coming. It's high time. Your dam is already breaking and you need a solid distribution to handle it.

[Florian]

Just to be sure - this also means Greenlight campaign starting May 1st, right?

About that ... we decided to give it some more time

I really appreciate it. The attention of a early access game on steam drops after a while. When you release it now on steam and have months later a stable multiplayer, less people will notice. So finishing MP will enhance the appreciat of the Game significant.

On the other hand, i'm verry happy about, that i played it so far. i didn't played 0.7 and 0.8
So the comeback was hard (pollution system, oil industry, ...). How hard would it be for a newbie?

[Lkgv]

I keep starting new games every now and then, only to realize it always follows the exact same formula and every game feels the same, the exact same way of playing works every time. It gets depressing.

I feel the same way.

I reached a point where I could make trains on my first game but thought my whole base was too shitty so I started a new game.

Then I reached a point where I could make robots... but I felt the game was too easy so I restarted a new game with a lot more aliens.

Then I started 2 new games and every time the beginning is really boring. I don't really know how to fix it, Tropico, Sim City and Dwarf Fortress have the same issue... after playing the same game for 50 hours you run out of content to discover. The only solution would be to add more more ressources and items to the game... sand, diamonds, clay, new types of rocks... I really need to look into modding, I wonder if it's possible to make a food mod

[Wardiaper]

Game is looking great.

I have no idea what I'm talking about with your computer memory problem because I'm a gamer and not a programmer, but what about designing your machines by reusing some parts? Like have a library of 100 parts +/- and you create new machines from those parts - like making something new in minecraft by reusing the same blocks in different configurations. Use some fancy programming to change the color of parts without having to make whole new sprites and you have virtually unlimited machines you can build with that combination.

Can't wait until Factorio comes to steam. So many bad games have been flooding steam lately that it's time for something great. Really good games like Space Engineers and Planet Explorers are priming gamers for this kind of gameplay.

[badminton]

The more we discuss it, the more we feel the game should be finished first for the steam. I personally felt, that with every update, the previous one was much worse experience.
...
Did you get my point? If we continue with this few more iterations, there will be plenty of people, that would never play the game in the current state, but would with the future changes.

On the other hand, i'm verry happy about, that i played it so far. i didn't played 0.7 and 0.8
So the comeback was hard (pollution system, oil industry, ...). How hard would it be for a newbie?

These two issues need to be balanced carefully I think. Factorio has the potential to become very complicated, rather than very complex. At the moment the balance is pretty good, but if it becomes too complicated (ie Dwarf Fortress) before you do a big release then you're going to force the game down a niche that will cost you sales. As awesome (and free) as Dwarf Fortress is, it's not easy to recommend to a lot of people, games like Gnomoria and Towns have a broader appeal.

My housemate is a perfect example: He's seen me playing Dwarf Fortress quite a lot but he's never felt the need to try it out for himself because of all the screens and hotkeys and 3rd party programs you need to really get the most out of it. When I got Factorio (after watching SplatterCat's spotlight vid <3) I didn't bother to tell him about it because it's a similar kind of game. When he saw me playing it he asked about it and I talked him though my objectively crappy factory and he was hooked. Literally within 10 minutes he'd bought his copy and started playing it. And now, even a couple of months later, every second conversation we have is about Factorio

The biggest pro to Greenlighting the game in its current form is that it isn't too difficult for a new player to get their head around. Also if you're able to do a content update on a regular basis it will keep people playing and talking about it for a long time. The biggest con is that Greenlight has been exploited for quick cash a little too much, so you're probably going to get a lot of very vocal, very immature people constantly pushing you to show how much work you're putting in. Thankfully most of this will be confined to the Steam forums Also it might make development take a little longer if you're always updating two versions of the game (direct dl and Steam), I don't know how that works.


tl;dr
Do you have a line drawn where you can say that Factorio is fully finished? You still have a lot of great ideas that may take years to implement. That big American market is warming up with a few YouTubers putting the word out over the last couple of months, that won't last forever.
As scary as it might be, I think you should hose it off, give it a brush and Greenlight it Keep in mind that it will still take time to go through the Greenlight process.

[ssilk]

I have no idea what I'm talking about with your computer memory problem because I'm a gamer and not a programmer, but what about designing your machines by reusing some parts? Like have a library of 100 parts +/- and you create new machines from those parts - like making something new in minecraft by reusing the same blocks in different configurations. Use some fancy programming to change the color of parts without having to make whole new sprites and you have virtually unlimited machines you can build with that combination.

You know what: I thought also to that, but then I thought, he already must have considered that.
But then I thought a bit more and when you think about making buildings out of such "blocks" you can make some kind of pseudo-3D; you can stretch and distort the layers and build then one over the other, so that you have in the end buildings, which look really 3D, but are just a set of many layers.