General Discussion

[Xela]

Is there a list of special characters that use a different font? Or on the contrary, anything but normal, common characters uses a different font?

I've took a shortcut by simply made all characters that are not Basic Latin to use the default Ren'Py font which has the best support. But it can be customized almost on per-character basis, although coding that would be time-consuming. So it's basically the latter where anything but normal is using the default font. Right now it is just for the TisaOTM which we use the most in the game and obviously only those parts that inherit from styles system (which should be all our cases).

[DarkTl]

Based on my experience of writing tags algorithms, in order to handle some cases perfectly we need to calculate the amount of available tags combinations. For example when character has 1 anal and 10 after sex tags, it makes sense to throw after sex now and then instead of showing one anal all the time. Although it's not a priority atm.

[Xela]

Although it's not a priority atm.

Yeah, we cannot pollute default  methods with stuff like that any more, I'll write new modules for this and a lot of other concepts in the future.

==>>
- Bestiary is restored.
- SM fixed.
- Loads of other fixes/improvements.

Ginna kill one more issue before I take a well deserved break... got a lot done today.

[DarkTl]

I increased the x size of interactions buttons to use longer names there more freely. Initially I wanted to try reducing text size a bit, but didn't found anything like that in pyt-screens-interactions rpy 

[Xela]

I increased the x size of interactions buttons to use longer names there more freely. Initially I wanted to try reducing text size a bit, but didn't found anything like that in pyt-screens-interactions rpy 

I don't really like those buttons there but don't have an alternative at the moment... the buttons are styled, that's why you haven't found text size, they are prolly on a different screen as well.

Ok, it is now possible to review next day after the screen was closed by clicking right mouse button instead of left. That was issue #6, time for a break

[DarkTl]

I changed interactions mechanics a bit. There is now a separate limit for every single interaction option based on randint and some other things, like in AA games basically. No more spamming is allowed.
The amount of gained disposition is not just a randint anymore. It becomes higher with high joy, lower with every next try if you use the same option, and lower when you already have pretty high disposition.
The chance of successful interaction option is based not just on current disposition, but also on hero charisma and amount of tries.

[Xela]

Great, we'll get it just right with time... maybe a special module to handle all the options. We just need to setup a working system first and see if it will be worth generalizing it later. I'll see if I can put a few more hours into it tonight but headache is killing me

[lamoli]

Yes interaction gets better and better but something is missing.. you can improve disposition and need to reach a certain number to be able for some ask/result to be working or get a better chance for it to work.. this is the ( normal good approach.. )

What about ( and probably repeat myself ) the Evil type one.. like corrupting the character using blackmail or some other corruption method ( just played with HHS+ and liked their blackmail/hypnosis/drugs method lol ) well would have to ask to a investigator like character to find the dirt on that person if any to get the blackmail working or train that hypnosis skill like that would reduce disposition if it fails ( so at some point if you fail too much the target will really hate you thats where the blackmail would be even more useful as if she has it you could make her accept against her will or if not just balance her disposition with the ( normal good interaction ) so it does not fall too far

I think corruption/blackmail would be really great for unique girls as it would feel better to make that girl yours more than in just 1 way ( when ever she likes you because you do/say good things or be able to exploit the dirt she has / control her hypnotizing her making her likes you more ( dispo up ) or some other things ) for strong personality girls evil option seems more right to me lol as for now i find it a little boring to max any girl dispo then nothing left to do..

Just imagine if at start we do not know the girl personality ( witch sound right by me as youv just met her ) you could reveal it talking to her or know about it asking other who know about her or with a investigator like npc type.. finding it out should helps you know the interaction that would work better on her ( better chance of it working as thats what archetypes are for ) and for the evil aproach finding it out could also give better options to corrupt her..

[Xela]

Yes interaction gets better and better but something is missing.. you can improve disposition and need to reach a certain number to be able for some ask/result to be working or get a better chance for it to work.. this is the ( normal good approach.. )

What about ( and probably repeat myself ) the Evil type one.. like corrupting the character using blackmail or some other corruption method ( just played with HHS+ and liked their blackmail/hypnosis/drugs method lol ) well would have to ask to a investigator like character to find the dirt on that person if any to get the blackmail working or train that hypnosis skill like that would reduce disposition if it fails ( so at some point if you fail too much the target will really hate you thats where the blackmail would be even more useful as if she has it you could make her accept against her will or if not just balance her disposition with the ( normal good interaction ) so it does not fall too far

I think corruption/blackmail would be really great for unique girls as it would feel better to make that girl yours more than in just 1 way ( when ever she likes you because you do/say good things or be able to exploit the dirt she has / control her hypnotizing her making her likes you more ( dispo up ) or some other things ) for strong personality girls evil option seems more right to me lol as for now i find it a little boring to max any girl dispo then nothing left to do..

Just imagine if at start we do not know the girl personality ( witch sound right by me as youv just met her ) you could reveal it talking to her or know about it asking other who know about her or with a investigator like npc type.. finding it out should helps you know the interaction that would work better on her ( better chance of it working as thats what archetypes are for ) and for the evil aproach finding it out could also give better options to corrupt her..

I agree with all of it but not for the next release. I think that even in case of HSS+, they write concepts. We don't, we prototype with Python/Ren'Py script and then improve upon that

For now it's more important to rip lines and test concepts because at one point I'll write Char UI module that will help us to track interactions, mood, dispositions and etc. on per character basis. Interactions module in general will also need some luv. So for now we need lines/simple concepts that will serve for the next release and then we'll generalize/improve the system through a class that can be used everywhere (throughout all modules).

===>
Prolly gonna have some good time for coding again today (finally managed to put in almost 9 hours yesterday).

[Xela]

...

I noticed in your last push that you do this:

Code: [Select]

hero.exp += randint(2, 4)
It is almost meaningless at higher levels, so better:

Code: [Select]

adjust_exp(hero, randint(2, 4))
or just:

Code: [Select]

hero.adjust_exp(randint(2, 4))
It's prolly only temporary as well... but it'll adjust for levels inflation best it can.

[DarkTl]

Yeah, I know that "evil" options are needed too. The problem is, I cannot rip lines for such options because I don't know any games with them. In Illusion games they are either very, very basic and limited or not translated from Japanese.
So lines should be written entirely by us, ie me 

I write lines much better now thanks to my experience in developing, but still it will require a lot of time to do it from the scratch without any base.

I noticed in your last push that you do this:

Code: [Select]

hero.exp += randint(2, 4)
It is almost meaningless at higher levels, so better:

I know. I don't think that interactions should be a reliable way to level up during the whole game. At least basic ones. Maybe advanced ones like sex or dating should be scaled.

[Xela]

Update ND screen GUI #24

I'll prolly do this one next... gonna think about for a bit cause this is not an easy issue to take out (loads of logic/sorting involved).

[DarkTl]

If you preparing for the next release, maybe it's time to fix #107?

[Xela]

If you preparing for the next release, maybe it's time to fix #107?

Not really... I just want to take out some of the persistent issues that are beginning to piss me off. Jobs are in a good place but we're still missing a couple of those. I also want SE to work before next release, although writing logic to fill Gismo's new interface might prove really, really time consuming (ALT is to just fall back to old, working design).

#107 is a good candidate for taking out as well, gonna take a look at that after the ND thing is done. I think that the way to do that is presort the data using a dedicated function and then use that data in the ND screen.

The reason I dropped jobs for a while and started hacking at other things is that there are too many distraction all over the game and the code and there is a constant need to study/circumvent that. We changed the main concept from a simple SimBro design to WM and then to something a lot more complicated than even that. Traces off older logic are left all over the code and a lot of that is not compatible with the new design. To make it worse, there are still older coding approaches left all over the place that would have served SimBro and sometimes even WM concepts perfectly but are no longer acceptable because it plainly requires more complex and confusing code to do simple stuff.

Codebase is getting a lot, lot better and easier to work with, but we're not there yet We will not be there for the next release either but after all the bugs in the release are fixed and I do the next major refactoring effort, that's where I believe the code base really become "PyTFall" instead of that merged with "SimBro"/"WM2" designs.

[Xela]

Damn.. this ND thing is a lot of sorting

I am not sure how to do it anymore, we can no longer just check for class because now classes do not specifically determine what the worker is doing. Rest cannot be checked like it was before at all... we basically checked which of the class sorted chars had rest/auto-rest as a job, that cannot be done anymore either.

Maybe:

There are two "ok" ways I can think of without rewriting the design:

1)
Service (SIW + Servers) / Warriors / Managers / IDLE

Where IDLE means None action + normal rest. Then for Resting row we add all auto-rest action under the classes and normal rest count for IDLE.

2)
Service / SIW / Warriors / Managers

Where resting row is auto-rest only... and normal rest is not reported at all.


In both cases top row are active actions taken from jobs, no longer classes.

I like option one cause it provides more data but may not be immediately understandable so I'll be coding that one unless someone comes up with a better idea or makes a good argument for 1).