General Discussion

[DarkTl]

When in doubt, use indifferent portraits, they are always suitable no matter the situation 

[picobyte]

But I am not going over new code after every push.

Fine with me. Just reporting changes here.

If there is a viewport there already or you put one in, this post is not relevant.

It was already in place there.

I've added a group equip implementation. Items are only shown, retrieved or given if all members of the group have an item or can be provided.

[Xela]

I've added a group equip implementation. Items are only shown, retrieved or given if all members of the group have an item or can be provided.

- Trying to equip items to a group of characters at least one of which doesn't want or cannot equip it causes ctd.
- Transfer button causes ctd. It should either be disabled or preferably lead to transfer screen with all of the chosen chars and mc.
- Some GUI stuff is still misaligned.

And most importantly:
- ALL item operations must go through their respective functions defined in item functions file. You should not pass items between inventories directly or equip them directly. There are cases where this doesn't matter but normal transfer/equipment is never one of those cases.

Reason for this is that in PyTFall a free character may refuse to receive, equip or give away an item. There are also counters in place that attempt to track how many items were transferred between characters (for character to which that matters) so they would give back an item that MC gave them but not one of their own items.

There may also be rules that prevent character transferring or equipping specific items.

All of those functions where such things are relevant take a "silent" argument, when set to True, characters do not try to explain or complain why they don't want to do some specific thing which is useful when dealing with them in bulk.

[picobyte]

Thanks for review. Will go through it later - got to work now.

[Xela]

If you're gonna test new Ren'Py prerelease, pull from github. They are slowly prepping migration to Python 3 so isinstance(..., str) which Thewlis used a lot has changed it's meaning due to new imports.

[DarkTl]

...meaning schools and interactions gonna get wrecked?

[Xela]

...meaning schools and interactions gonna get wrecked?

They got a little bit wrecked but my last push fixed them (in python 3 all strings are unicode so there is no more splitting of "basestring" to "str" and "unicode" as there is in Python 2). When they migrate to Python 3, a lot of stuff will break... for now they just imported some stuff that partially mimics Python 3 behavior. It's not that difficult to migrate but there will be some difficult cases, like for example, in Python 2, dividing two ints returns an int, in Python 3 it returns a float...

[DarkTl]

Ah yes, I even had a testing menu where I checked python divisions. Because in delphi I was taught that division result is always float -_-
Let's see...
- xrange does not exist in python 3, for some reason you use it a lot in the code.
- .next() method does not exist in p3, we use it in quests system and calendar
- zip(), filter() and dictionary methods don't return lists any longer. We don't use zip and filter much, but dictionaries are used a lot. Also has_key method is gone, which is used in a few places.

[Xela]

- xrange does not exist in python 3, for some reason you use it a lot in the code.

Because it's what you're supposed to use (returns generator instead of a list). It's renamed to range in Python 3 and original range is no more.

- .next() method does not exist in p3, we use it in quests system and calendar

Not relevant, these methods are defined by us so they are not subject to this change.

- zip(), filter() and dictionary methods don't return lists any longer. We don't use zip and filter much, but dictionaries are used a lot. Also has_key method is gone, which is used in a few places.

It's an easy fix.

It may be a while till they actually switch so there is no urgency.

[Xela]

They don't have basestring in Python 3 either so this code would have to be adjusted but I don't want to try and write compatible code like importing string types for six. I don't expect this project to have serious issues with migration. SimPy would have to be adjusted a little bit (it always needs to be for Ren'Py) but it's not hard either. I am not sure about the json validator thing but they should have Python 3 support as well.

[DarkTl]

I know what we missing 
We missing Continue option in the main menu which loads the most recent save, and is disabled if there are no saves. It's de facto standard in modern games. I wonder how hard it would be for renpy.

[Xela]

We had one in alpha. It is not difficult at all, I think there is a builtin func for that or i wrote one. Gismo got rid of I think.

[picobyte]

- Trying to equip items to a group of characters at least one of which doesn't want or cannot equip it causes ctd.

this seems fixed now

- Transfer button causes ctd. It should either be disabled or preferably lead to transfer screen with all of the chosen chars and mc.

Some of it fixed, but there are some issues still.

- Some GUI stuff is still misaligned.

better now?

And most importantly:
- ALL item operations must go through their respective functions defined in item functions file. You should not pass items between inventories directly or equip them directly. There are cases where this doesn't matter but normal transfer/equipment is never one of those cases.

Reason for this is that in PyTFall a free character may refuse to receive, equip or give away an item. There are also counters in place that attempt to track how many items were transferred between characters (for character to which that matters) so they would give back an item that MC gave them but not one of their own items.

There may also be rules that prevent character transferring or equipping specific items.

All of those functions where such things are relevant take a "silent" argument, when set to True, characters do not try to explain or complain why they don't want to do some specific thing which is useful when dealing with them in bulk.

I think the checks should be in place. Are you aware of a path where checks do not happen?

[Xela]

this seems fixed now
Some of it fixed, but there are some issues still.

ctds are gone but transfer still doesn't go through functions or something is still off.

better now?

Little bit, I can do this myself.

I think the checks should be in place. Are you aware of a path where checks do not happen?

SE is one such an examples, because we deal with consumables and it's a life and death situations so the idea is that they just share.

[DarkTl]

Speaking of consumables, don't forget that  ideally cblock field should work in SE too.