• To improve security, we will soon start forcing password resets for any account that uses a weak password on the next login. If you have a weak password or a defunct email, please update it now to prevent future disruption.

Tool Ren'Py UnRenGUI, UnRen-forall(v9.4), UnRen-Powershell-forall(v9.4), UnRen-old

5.00 star(s) 3 Votes

Twistty

The Happy Beaver
Respected User
Donor
Former Staff
Sep 9, 2016
4,154
37,860
Madeddy
Glad to see your still playing around with Unrpyc - your git page seemed quiet.

" renpy devs fucked just the engine code with some changes up again "
Wonder what they did this time - I read through the changes - but....

Cheers
 

VepsrP

Well-Known Member
Modder
Dec 13, 2017
1,387
1,364
Nah. Simple. You cited it from renpy, i did also from you.
And on renpy.org above the "pre-release" button is a text where clearly stated is for what such a version is and what not to do with it.

Also true for the 8.2.x is, the renpy devs fucked just the engine code with some changes up again. Makes the decompiler incompatible for now.
It's true, what they did with the parameters is just a nightmare. It is necessary to completely redefine the function for their restoration. There will be a decent amount of work.
 

Madeddy

Active Member
Dec 17, 2017
813
464
Madeddy
Glad to see your still playing around with Unrpyc - your git page seemed quiet.
Yeah, that's true.
1. I got a bit annoyed and lost some interest because in original unrpyc is no activity; the dev is busy with RL. This all needs some big revamping.
2. There where also no heavy bugs, just some annoying new displayables and VepsrP got this.
3. We got then warm seasons and this takes me to doing summerly stuff.
4. I have now some pain and movement problems in my left hand and i need to see whats going on.
5. I must admit, building my own Railway Empire did also take a bit time. :eek:
Madeddy
Wonder what they did this time - I read through the changes - but....

A complete rewrite of this part. Hope what i did codewise is enough. But i fear...

It's true, what they did with the parameters is just a nightmare. It is necessary to completely redefine the function for their restoration. There will be a decent amount of work.
Nah. I got this already.
I steal sometimes your code, now you get mine. How did "Jack O`Neil" put it? "Togetherness is gooood." :p
I put it in my GH or post it here.

Greets
 

Madeddy

Active Member
Dec 17, 2017
813
464
Its in my experimental 1.2.0_dev branch.

I don`t really like the first two lines of my changes. Defining the "nameonly" var at start was needed or you get a "`bla` referenced before assignment" error. Maybe there is a better way. Will see.

The branch has also some unresolved old error about unknown AST nodes. Hm, later.
 
Last edited:
  • Like
Reactions: yoyomistro

yoyomistro

Engaged Member
Jan 15, 2017
2,679
3,356
Yeah, that's true.
1. I got a bit annoyed and lost some interest because in original unrpyc is no activity; the dev is busy with RL. This all needs some big revamping.
2. There where also no heavy bugs, just some annoying new displayables and VepsrP got this.
3. We got then warm seasons and this takes me to doing summerly stuff.
4. I have now some pain and movement problems in my left hand and i need to see whats going on.
5. I must admit, building my own Railway Empire did also take a bit time. :eek:

A complete rewrite of this part. Hope what i did codewise is enough. But i fear...


Nah. I got this already.
I steal sometimes your code, now you get mine. How did "Jack O`Neil" put it? "Togetherness is gooood." :p
I put it in my GH or post it here.

Greets
Wow, please get your hand checked out if possible. I hope everything is OK and it's just some temporary pain. Glad you got some time outside.

extrapos and extrakw are now part of the list (actually dict keys) parameters
while that code would not yield exceptions, it added the same names twice which is a waste of time
Yo, been a while since I've dropped in here. I've been having this killer bug from unpacking Our Bright Days with a positional argument for a function. For some reason, when the game is unpacked it doesn't recognize the 'pos' attribute or argument; works fine when it's packed. I refuse to play it packed lol but just wanted to see if any discerning minds here could solve this. I've tried looking to see if there's some sort of replace script running or something, no luck finding it.
You don't have permission to view the spoiler content. Log in or register now.
Do you think this bit here will fix this bug with Our Bright Days decompilation? Not sure from reading the actual code (not super familiar with the Ren'py codebase, as much as I stare at it from time to time lol) but it seems to be pointing in this direction.
 

Madeddy

Active Member
Dec 17, 2017
813
464
Wow, please get your hand checked out if possible. I hope everything is OK and it's just some temporary pain. Glad you got some time outside.
Yeah, i agree. Lets hope.

Do you think this bit here will fix this bug with Our Bright Days decompilation?
Cannot check atm because i dont get even to your bug. I have another error if i use the v1.2.0 decompiled files from this game. And i think it will perhaps soon bring even more "fun": The changelog of renpy shows new revertable classes... o_O
 
Last edited:
  • Like
Reactions: yoyomistro

yoyomistro

Engaged Member
Jan 15, 2017
2,679
3,356
Yeah, i agree. Lets hope.

Cannot check atm because i dont get even to your bug. I have another error if i use the v1.2.0 decompiled files from this game.
There was another issue I ran into with that title decompiled with passing **kwargs into a list; the shopping list code passes **kwargs and that causes errors when running decompiled.

You don't have permission to view the spoiler content. Log in or register now.
 

VepsrP

Well-Known Member
Modder
Dec 13, 2017
1,387
1,364
Support for changes to the "AttackOnSurveyCorps" game engine has been added. The game starts. I suggest that people interested in the game check the rest.:BootyTime:
 

VepsrP

Well-Known Member
Modder
Dec 13, 2017
1,387
1,364
Nah. I got this already.
I steal sometimes your code, now you get mine. How did "Jack O`Neil" put it? "Togetherness is gooood." :p
I put it in my GH or post it here.
I've looked at your changes. It was all cool, of course, but I thought of doing it that way. :BootyTime: But none of this makes any sense at all.:HideThePain: As I said, they completely redesigned the parameter system. They used a different type of dictionary(OrderedDict). Added typing to the parameters, as in python itself. An approximate solution can be found on my GitHub.
 

Twistty

The Happy Beaver
Respected User
Donor
Former Staff
Sep 9, 2016
4,154
37,860
UnRen-forall v9.1
Wow - that was way faster than I expected!
How many past versions of Ren'Py does this v9.1 work til?
7.3.5 or just 8.0.1?

Thanks VepsrP (y)
 
  • Like
Reactions: wm.pghub

Madeddy

Active Member
Dec 17, 2017
813
464
...but I thought of doing it that way. :BootyTime: But none of this makes any sense at all.:HideThePain: As I said, they completely redesigned the parameter system....
That's totally OK with me. If this works better for you. :coffee:

But it makes sense i think. I noticed that is with "positional" the same behavior and code area we had with "extrakw" and "extrapos", so i tested it with the same solution there which works for now without much hassle.

This is how in see it whats going on with all this:
1. Nearly 2 years ago they began to work on this parameter stuff. "extrakw" etc. was either "None" or with a value, but always there. After the change they are allowed to be missing, so it banged with the old code. You cannot check for a class-vars state that's not even there sometimes, so we did write some fixes.
2. Now they did rewrite this part and the same happened to "positional". If i rewrite this part now, like you in part did i think, we run perhaps the next months after small new bugs from it.
3. Momentan unrpyc still works becuse the RenPy devs put the old stuff as legacy in the parameter rewrite. When they remove it later we rewrite this again.
4. The commit to this says it clearly: They are not done with redesigning the AST and ATL stuff. When they do the next part we rewrite this again... We do not know where this goes in the end.
5. I admit, some of the code in "unrpyc" i do not understand or is hard to follow where it goes and what it does. I did not write this program, so another dev's code to understand is horror and its on top of it a "Decompiler". Not the easiest things out there.
6. The "unrpyc" boss said it self: In unrpyc is a lot of old code for py2 and hacks and stuff for older RenPy compatibility etc. So if he has time again, i expect bigger changes in unrpyc. Then is this code for parameters again rewritten.
7. In May this year ends RenPys py2 support and they throw this all out. I don't want to know what this means... So, i expect the 7.7x version the last we see from py2. Maybe this means we rewrite then some stuff... again. o_O

No thanks you, a quick fix does for now IMHO. If not, its not the end of the world. I think unrpyc will anyway need to be in big parts redone. Half of the code out, some rewritten, half of the old code will still be good. A new version for py3 only and a old one for legacy RenPy is IMHO better to maintain.

BTW: Good work VepsrP.

Greets
 
Last edited:

VepsrP

Well-Known Member
Modder
Dec 13, 2017
1,387
1,364
No thanks you, a quick fix does for now IMHO. If not, its not the end of the world. I think unrpyc will anyway need to be in big parts redone. Half of the code out, some rewritten, half of the old code will still be good. A new version for py3 only and a old one for legacy RenPy is IMHO better to maintain.
So you just bypassed the lack of [positional, extrapos, extrakw] properties of the paraminfo object, eventually just losing the function parameters. I was able to add support for a new way of arranging function parameters. By the way, this method is completely consistent with what is used in python itself. As for legacy, it's support for old unrecompiled scripts, which is exactly what I dislike most about RenPy.

If you don't want to waste time working with "temporary" changes, I can suggest the "at transform" component (file - sl2decompiler.py , function - print_keywords_and_children, located at the very bottom. This stuff doesn't work properly in the game Our Bright Days, if I remember correctly. For this reason, I decided to rewrite all unrpyc - to change the way decompiled parts are written to the final file. A partial implementation can be viewed on my GitHub, the unrpyc-2.0 project.
 
  • Like
Reactions: yoyomistro

Madeddy

Active Member
Dec 17, 2017
813
464
Do you think this bit here will fix this bug with Our Bright Days decompilation...
...another issue I ran into with that title decompiled with passing **kwargs into a list; the shopping list code passes **kwargs...
I see. I must say my forks unrpyc version is broken for renpy 7.7.x/8.2.x and probably even for 7.6.x/8.1.x. I am too far behind on fixes and I'm in over my head here ATM. Well shit.
VepsrP did good work with his version. I mean it! Must been a lot of work. :unsure:
 
  • Like
Reactions: yoyomistro
5.00 star(s) 3 Votes