General Discussion

[Xela]

Going to look into predicting girlslist screen. It's kinda jumpy and there is a chance that it can improved. Maybe look into girlsrofile screen as well but that cannot be properly predicted because chr is an unknown variable before it's run, maybe there is a way to predict it partly.

Next on the list is looking into loading Dark's fns for loading tags.

[DarkTl]

Origin "Random Girl" is kinda out of place. There is a list of random places in Evolved Help Wanted thread, might as well use them.

[Xela]

Origin "Random Girl" is kinda out of place. There is a list of random places in Evolved Help Wanted thread, might as well use them.

Never bothered me...

I got the first page of the girlslist to predict and it's working well... Otherwise I am not going to continue with this atm, instead I'll look into putting a new tags dict together and loading new tags  from it.

[Xela]

... Question:

How much RAM would you conciser to be acceptable for PyTFall to use?

Right now it's about 200MB (for caching graphical elements). I am still toying with making the game run faster and the problem I've encountered is that it seems to be not enough. Because we can jump to so many screens from main menu, all the images cannot be preloaded to RAM before they are rendered. Ren'Py allows to increase the use of RAM for graphic elements without limits but what do you think a good limit might be?

[DarkTl]

Hm. I never wondered before, but can it use more than 3gb, ie can it be a 64 bit application?
Becasue if it can't, 3 gb is what we have.

I'd say today pc should have at least 4 gb anyway, at least 8 if you use it for gaming and about 16 if you use it for work with special ram demanding software.

[Xela]

Hm. I never wondered before, but can it use more than 3gb, ie can it be a 64 bit application?
Becasue if it can't, 3 gb is what we have.

The largest I've seen for a Ren'Py game was 1.8 GB. It was a commercial post-apocalyptic VN, they now moved to Unity because they want to release for consoles and Ren'Py doesn't support that. Native resolution was 1980x1080 and it had a lot of effects and animations.

Basically I wanted to know if everyone was ok with about 500 - 600MB. The problem is that some parts of the game are very "jumpy", like girlslist or map, transitions skip and even without transitions there is a visible delay. The reason for the is that the images that we use in those screens are read of the harddrive. Today I've tried preloading them into memory but the default setting we've used up to today was too low.

I've increased it and preloaded the map and first page of girlslist screen so they work instantaneous without transitions (when those are disabled in preferences) and that the transitions actually work.

[DarkTl]

That's very low. The more ram the game can use instead of reading from hdd, the better.

[Xela]

That's very low. The more ram the game can use instead of reading from hdd, the better.

I'll look into that. The best case is preloading all graphical elements in RAM and preloading (when possible from previous screens) girlspics on demand but at some point we may want to release for Android and that will have to be recoded. A lot of stuff is being predicted really well by Ren'Py in advance (especially in VN mode like events/girlsmeets) so some things are not required for forced preloading.

Optimization will be required, but right now I am busy with loading new tagging system into the game.

If you want to know what's being loaded correctly, you can click Shift + D and enable: Show Image Load Log. Whatever shows up in red is a f*ckup that we'll have to fix...

[DarkTl]

Btw did you fixed traits application? I recall you mentioned that I forgot to change it during trait application and removal. I changed it from 10% to 5% per level.

Also, I'm not sure how should I deal with skills. You mentioned that

Items now accept a new field:

mod_skills: {"strip": [0, 1, 2, 3, 4]}

Where:
0: Multiplier to action counter
1: Multiplier to training counter
2: Multiplier to skill (when getting a total skill value)
3: Adding to action counter
4: Adding to training counter

Traits  now accept a new field:

mod_skills: {"strip": [0, 1, 2]}

Where:
0: Multiplier to action counter
1: Multiplier to training counter
2: Multiplier to skill (when getting a total skill value)

Multiplier to counters means faster learning when learning occurs. Multiplier to skill means better performance. I get it.

Does adding to counter mean learning process itself? Like, read a book about cleaning means add to training counter?

[Xela]

Btw did you fixed traits application? I recall you mentioned that I forgot to change it during trait application and removal. I changed it from 10% to 5% per level.

I am not sure. What happened was you changing it in one spot but forgetting (or not being aware of another). I made sure that both spots worked off the same logic, that I remember. Just not sure if it's 10 or 5 percent.

Also, I'm not sure how should I deal with skills. You mentioned thatMultiplier to counters means faster learning when learning occurs. Multiplier to skill means better performance. I get it.

Does adding to counter mean learning process itself? Like, read a book about cleaning means add to training counter?

Yeap, that's exactly how it works. Basically the way you should think about it is that every skill has two parts:

Practical skill (Actions)
Theoretical knowledge (Learning)

A character that would consider him/herself "Too Cool For School" under the same circumstances will loose out to a wiser characters that read a book or take a class once in a while. However if a character is just learning all the time, he/she will never make any money without taking an action.

Practical Skill is never useless either, it's just (A LOT) more useful with theoretical knowledge of the skill.

Multipliers increase/decrease the rate at which Practical and Theoretical knowledge is gained and the final multiplier increases/decreases the final (in the game code and comments it is referred to as "True" knowledge) value is being calculated.

Counters just increase P/T knowledge before the multipliers are applied.


***
Now that I think about it, we should prolly add new fields to items that caps the skill increments for items (so some shitty book cannot increase skill indefinitely.)

[DarkTl]

Alright. I'm not sure though how to show in gui something like "cleaning practical learning+5 cleaning theoretical learning+10 cleaning performance+15%".

Btw when you set a skill in girls data files, like cleaning = 30, what are actual parameters in this case?

Now that I think about it, we should prolly add new fields to items that caps the skill increments for items (so some shitty book cannot increase skill indefinitely.)

- Can changes to multipliers be permanents? Like, you read a book and now you always learn something faster. I'm not even sure if we should allow it or not. It could be too powerful.
- But I think that any modification should be permanent. Since it's not charisma or something, it's practice and knowledge, it's hard to imagine how you can obtain it temporary and then lose somehow. So this part

3: Adding to action counter
4: Adding to training counter

should be permanent only, thus available only for miscs and consumables. And without negative values (I believe you mentioned it already).

[DarkTl]

Never bothered me...

It depends on what are we going to do with origin. If it will be used to make girls know each other (and this is perfectly understandable in case of unique ones), then all random ones will know each other too 

[Xela]

Alright. I'm not sure though how to show in gui something like "cleaning practical learning+5 cleaning theoretical learning+10 cleaning performance+15%".

I didn't want to display them at all... Just the descriptions for items. My plan was to make skills illusive and hidden. Whenever the skill is not enough for something, game will report it as such. If a girl is lacking theoretical skill, game will report that as well (maybe for a price in school or at the higher level of the said skill of MC). Basically I wanted to prevent another calculator thing as with stats where almost everything is shown in plain numbers.

Otherwise we could display them either as you described or just as true skill value. Or everything in the format similar to the item files.

Btw when you set a skill in girls data files, like cleaning = 30, what are actual parameters in this case?

1/3 goes to learning, 2/3 to action.

- Can changes to multipliers be permanents? Like, you read a book and now you always learn something faster. I'm not even sure if we should allow it or not. It could be too powerful.
- But I think that any modification should be permanent. Since it's not charisma or something, it's practice and knowledge, it's hard to imagine how you can obtain it temporary and then lose somehow. So this partshould be permanent only, thus available only for miscs and consumables. And without negative values (I believe you mentioned it already).

Multipliers are all capped at 1.5 (50% extra). I think items like that will prolly mess the game up but you can always use temp consumables that reverse the effects after some time has passed. *It would be very easy to fix the minimum at 0 as well.

You loose knowledge and practice skill if not in use for a while but that is a possible mechanic for the future.

I am not 100% sure how the game will behave on negative values. Prolly nothing bad will happen.

It depends on what are we going to do with origin. If it will be used to make girls know each other (and this is perfectly understandable in case of unique ones), then all random ones will know each other too 

We're pretty far away from this on the game level.

[DarkTl]

Multipliers are all capped at 1.5 (50% extra). I think items like that will prolly mess the game up but you can always use temp consumables that reverse the effects after some time has passed.

Let's say you equip maid dress (1.1) and drink some maid potion (1.1). What's the result? 1.1+1.1, 1.1*1.1, just 1.1?

You loose knowledge and practice skill if not in use for a while but that is a possible mechanic for the future.

That requires years. Items work for days.

[Xela]

Let's say you equip maid dress (1.1) and drink some maid potion (1.1). What's the result? 1.1+1.1, 1.1*1.1, just 1.1?

Result would be 3.2 (it would be capped at 1.5).

I wrote before that it's added effect:

default = 1

Modifiers are just added/subtracted from that.

What I wrote was that 0.01 (1% extra) is already a good option for trait or an item.

***Assuming we were talking multipliers.