Autoplacement

[ssilk]

For poles:
When pressing shift while placing a pole: Auto-connecting is turned OFF. I don't know what to do with the cable, but maybe the pole creation should use no copper-wire and use the copper only when cabled?
When pressing CTRL while placing: Auto-Connecting to ALL poles which are reachable. Of course more copper wire is needed for that.

Factories:
Pressing shift, when placed: It autoconnects the factory to the next pole.

Inside the facctory: When selecting the item it should produce, you can press a button "Auto", tries to build the needed inserters for in- and output. See down!
When you press shift while pressing this button (or there is a second button?), it places also the poles for the inserters and/or factory if not connected yet.

Weapons, Lamps, furnaces, burners ...:
Pressing shift when built: Tries to autoconnect to a pole (even if the item doesn't need power, but maybe you want to use that further?).
Pressing Ctrl: Tries also to build inserters, like for the factory. It tries to build the poles, so that the inserters are also connected.


Algorithm for building poles:
- if already connected: Nothing is build
- if no poles available, no poles can be used. No cable available: Pole may be set, but the cables are not placed. In such cases a hint should be displayed (need to build copper wire first).
- most cheapest poles are used, if possible, if not it tries to use more expensive poles.
- To spare algorithm power: It tries randomly building poles on a line between the next powered pole and the targets, until everything is satisfied. Not more than 2-3 poles are build. If the next pole is too far away no poles are built.

Algorithm for building inserters:
- inserters are build before the poles!
- if already connected: Nothing is build
- if no inserters available, no inserters can be used.
- tries to use the most cheapest inserters first.
- For input inserter(s): Goes around the the factory (or whatever) and looks for a belt, a chest, an inserter (that has already placed one item) or a miner (already placed item) - in this order - if an inserter can grep one of the item-type which is needed inside. Continue, until every item is inserted. Filter-/smart-inserters are pre-programmed!
- For output: Looks for belts, which looks emptier than the input belts. If no emptier belt: Output on input belt possibly after input-inserter. If no belt available looks for chests which are emptier than input chests. Or looks for inserters which can take over output.

[ssilk]

I add a simple idea for the poles: You can tell the autoconnector to which pole it should prefer autoconnection by moving the pole you want to set now from the pole you want to autoconnect to the point you want to set it.

Or in other words: Currently autoconnect chooses the pole which is nearest (?). But with this idea it tries to keep the connection from the pole you came from as long as possible, before it searches the next nearest pole.

Additional: When moving the mouse fast back and forward, or pressing short shift, the autoconnector forgets the the preferred pole.

[suaig]

I like the idea of automating more things, but I think that it should be something that you have to research. Early on you have to manually place things, but afterwards you can do it automatically after you research it and possibly craft automatic construction equipment, just like how you have to manually craft everything before you develop crafting machines.

[MF-]

I add a simple idea for the poles: You can tell the autoconnector to which pole it should prefer autoconnection by moving the pole you want to set now from the pole you want to autoconnect to the point you want to set it.

Or in other words: Currently autoconnect chooses the pole which is nearest (?). But with this idea it tries to keep the connection from the pole you came from as long as possible, before it searches the next nearest pole.

Additional: When moving the mouse fast back and forward, or pressing short shift, the autoconnector forgets the the preferred pole.

I like this suggestion a lot, it would work intuitively and require no additional controls.
I wonder how could this "automatic wiring change" be moved into separate suggestion.

[ssilk]

I like the idea of automating more things, but I think that it should be something that you have to research. Early on you have to manually place things, but afterwards you can do it automatically after you research it and possibly craft automatic construction equipment, just like how you have to manually craft everything before you develop crafting machines.

https://forums.factorio.com/forum/vie ... 5211#p5213
https://forums.factorio.com/forum/vie ... 5181#p5213 here point 4!!
https://forums.factorio.com/forum/vie ... 5182#p5182

I'm one of those, who has suggested to "research better handling". But now I'm not sure anymore.

Things which are important
- features, which extend behavior and those which change it. Changes are bad (the second post here). You can't change the handling of - even small - details during a game, unless you have a very good reason for it.
Extensions are bad, if they brought to late in the game (like the first post), they must be all available within the first 5 minutes! Believe me, this is very important. And they must be limited, not more than 5 extensions at all!

- logic. Factorio is a very logic game. I really like that. There is not too much magic in it. But when the player suddenly can do more thing by "researching" it, this isn't very logic. Maybe he can "train" his abilities, but not research them. I would accept the following logics (for example!): the player built x poles: he can place poles with shift. He placed x inserters: he can auto place inserters with control. Some very easy things early in the game. Open for better ideas, this is just an example.

- speed. I always said, that the gameplay need speed up. It has many repetitions. Has some length. This hasn't changed much. For example, building a track for the train is in my eyes horrible complicated.
Everything, which speeds the game up is in my eyes good. This means also, a better control should be available from beginning, or at least early in the game. See above.

[kovarex]

I'm afraid to add complicated logic hidden behind another controls.
Shouldn't most of the trouble (of repetitions and tedious building) be solvable by blueprints?
I also believe, that there should be some better way to build tracks (using some kind of builder robots for that as well), The player would just specify the starting and ending point, and the track would automatically build the shortest way somehow.

[MF-]

I'm afraid to add complicated logic hidden behind another controls.
Shouldn't most of the trouble (of repetitions and tedious building) be solvable by blueprints?
I also believe, that there should be some better way to build tracks (using some kind of builder robots for that as well), The player would just specify the starting and ending point, and the track would automatically build the shortest way somehow.

+1

@ tracks - don't forget about track removing, it takes so much time.
So far I guess I would be happy with a straight-track building robot and any-track removing robot.
My tracks usually don't have many curved points mid-track, even though those would be way cheaper (straight costs as much as curved, while curved is 5x longer)

[ssilk]

See this: https://forums.factorio.com/forum/vie ... 6958#p6958

I have suggestions from that:

1. I want to click and drag cables over many poles. Select the wire you want (all three types), select start, drag to target and the game makes the more or less most optimal cable-routing in between. While you are dragging you can see, which cables would be built. If the routing is not possible, you see it blinking or a message displays...

2. It's not so clear, that you need to select the base of a pole and not the top. Maybe both should be possible? Or if not this, when hovering over the top of the pole, the base is highlighted to make it clear, that the sensible are is down.

[MF-]

Both sound very good.

[ssilk]

Auto-rotate: rotate a non-self-rotating item

Before it is placed - instead of pressing R several times - press the LMB on the field you want to place it, or any empty field (it should be empty, to place it, right? Exception only if you want to replace an item) and drag some pixels into the direction you want to rotate the item. The item will be rotated into the direction you drag.
This works, even if you hover any other item; the mode doesn't switch to delete/dig.

Edit: could work also on resource-fields...

[MF-]

It happens to me quite often, that I place something and then immediately decide to remove it.

Would you be happy if that was done using RMB?

[ssilk]

Hm. In detail: you have an item in the "hand" (selected). Now you want to rotate it.
When you place it over an empty tile, and now press LMB, then drag into a direction, this is interpreted as rotation of the item in this direction. Now you can place it with RMB.

Or maybe the following: item selected. You stand still. Now you place the item with RMB, but before you lease the button and after it is placed, you move it. The direction of movement is interpreted as rotation of this single item.

[MF-]

Ah, that explains it.
I would never prefer #1 to hitting R, but some indeed might...

I am afraid I might trigger #2 on accident / building walls/rows of something. Now I do that by RMB-dragging the thing around.

[ssilk]

Edit: rewritten due to bad English.

I would interpret #2 only then as a rotation command, if moved some pixels away from the point you begun to press RMB, but keep within the tile where it should be placed. First is, because it should not be rotated just by pressing RMB and accidentally move it. Second is to avoid confusion with stating tiles in a row. (This means also, you will have only very less pixels for rotate, before it switches the mode to place a second/third tile... but this is wanted, because this method is not very useable for zooming out wide from the map.)

I thought, the direction is not remembered; it could be used for the case that you want to rotate only one item of many.

I think to find the right thresholds for every zoom-level needs some tuning, but I think also this will find friends. Should work only when zoomed more or less in, because it doesn't make sense to rotate a single item with this method.

[MF-]

That last post looks like it was auto-translated... I don't really understand it.

[ssilk]

That last post looks like it was auto-translated... I don't really understand it.

Lol. I will correct it. Maybe the autocorrection was too bored.