Poll for modding interfaces requests

[thelordodin]

Happy New Year everybody!

Wouldn't that be great if all the modding interfaces requests be put to a big poll, where all the modders could vote for them?
As an example the number of downloads (or uses) of mods would give mod authors priviliged rights to vote.
Then you could prioritize your work on implementing the suggestions.

Also that would be great if you put an estimation on the requsts and only requests with high vote to estimated implementation time would be implemented the first?

Workflow example:
1. Any modder puts writes a suggestion
2. Other modders see the suggestion and vote for it if they need it too
3. At some point suggestion accumulates some votes (say 10 votes)
4. A developer looks on the suggestion and estimates how long will it take to implement it in workhours. Also if there are some arguments why shouldn't it be implemented - he declines the suggestions or accepts it.
6. Modders continue to vote for accepted suggestions and correct (or abandon) the declined ones.
7. Suggestions with highest votes/workhours ratio are scheduled to next releases with highest priority among modding interfaces improvements.
- All modders have (their mods downloads)/1000 votes. The ones who have no downloads have 1 vote.
- Each modder have a queue of suggestions he voted for. Only first 3 of all suggestions in queue really get his votes. Others wait until this ones are implemented.

This way you'd have much better feedback from modders, also you could focus on most important suggestions and the progress would be much more transparent to the community, so they would be able to make even more good mods.
What do you think about it?

[Bilka]

The devs implement basically every suggestion that is feasible. There's no reason to vote for suggestions if there aren't so many that they can't be all considered due to time limitations. They already have the best feedback from modders: The modders can make a request and know that it will be considered, regardless of how popular their mods are or how many people saw the request or how many people voted for it.

By adding polls (weighted by mod popularity) you actually make less suggestions likely to be implemented if you expect the devs to then only add things that have many votes.

[Rseding91]

Is there something you've wanted that hasn't been implemented in the mod API?

[thelordodin]

Sure:
1. Interfaces for ghosts: getting/setting circuit connections, red bar for inventories for ghost chests/
2. Ability to save arbitary data (lua-table) with entity - for sake of saving custom entites' state into blueprints (examples of use: any mods' additional properties)
3. Ability to save arbitary data into GUI elements "tag".
4. Lua debugger (viewtopic.php?f=25&t=55625&p=327187&hil ... er#p327187), for example this one https://github.com/slembcke/debugger.lua if would be perfect to have its CLI API available from outside the game. It would be ok if that require to set an option in factorio to enable it.


In fact for 2-3 it's just one pointer on LuaEntity and GUI element...
I do understand about cache locality and other perfomance questions,... but what's sizeof(LuaEntity)? Do really adding 8 bytes to it really matter?

Use cases of (2):
- Alarm combinator - this mod was very convinient before you introduced "Programmable speaker" - that mod also needed to save a string with entity
- Sticky notes - same magic to save the message string with entity to blueprint
- I abandoned my Complex combinator mod - because there is no sane way to save it's progragram anywhere
- Wireless signals Lite - radio channel should be saved with entity

I think many other mods don't implement some functionality because (2) isn't possible.