11.0 bugs

[pgimeno]

No idea where your -pthread comes from. Running 'rgrep -il pthread' on the source shows only three matches and all of them are C++ files.

(Note I meant -lpthread, not -pthread)

[bartbes]

It's definitely in the generated makefile for me:

Code: Select all

SDL_LIBS = -pthread -lSDL2

(Of course relying on SDL2 to add it isn't a good thing, but it works.)

And -pthread implies -lpthread.

[pgimeno]

I get this in my system:

Code: Select all

$ pkg-config --cflags sdl2
-D_REENTRANT -I/usr/include/SDL2

$ pkg-config --libs sdl2
-lSDL2

$ dpkg -l libsdl2-\* | grep ^i
ii  libsdl2-2.0-0:amd64       2.0.5+dfsg1-2   amd64        Simple DirectMedia Layer
ii  libsdl2-2.0-0:i386        2.0.5+dfsg1-2   i386         Simple DirectMedia Layer
ii  libsdl2-dev               2.0.5+dfsg1-2   amd64        Simple DirectMedia Layer development files
<snip>

I guess that's where the SDL_LIBS variable is set from. I've included my SDL2 versions for additional info. It seems not all systems include pthread as part of the SDL2 flags.

[pgimeno]

Could we have the embedded boot.lua include the empty lines in the original source? The space difference is negligible, and it helps a lot understanding errors when they happen in boot.lua, because otherwise the line numbers don't match the source.

Code: Select all

diff -r 04b3d1aa9e34 src/scripts/auto.lua
--- a/src/scripts/auto.lua      Mon Jun 11 23:04:09 2018 +0200
+++ b/src/scripts/auto.lua      Fri Apr 05 12:40:21 2019 +0200
@@ -84,6 +84,8 @@
                        counter = -1
                        --append the output to what we had, plus a newline character (0x0a is newline)
                        out_data = ("%s%s0x0a,"):format(out_data, line:gsub("\r", ""):gsub(".", tohex))
+               else
+                       out_data = out_data .. "\n\t0x0a,"
                end
        end
        --close our input

(While on it, 'function auto(...)' could be local, as could the variable 'out_file')

[slime]

Could we have the embedded boot.lua include the empty lines in the original source?

I've applied your change, thanks!
https://bitbucket.org/rude/love/commits ... ad4b938620

[rjoukecu]

Hey, I think I've found strange bug or I'm missing something.
Anyway the simple line below does not work

Code: Select all

if love.filesystem.getInfo("Religion", "directory") == nil then love.filesystem.createDirectory("Religion") end

It simply refuses to create a folder.
The solutions I've found is to removing "if" part or if I change Religion into religion (from upper to lower r)

But then this piece of code works perfectly:

Code: Select all

if love.filesystem.getInfo("Chars", "directory") == nil then love.filesystem.createDirectory("Chars") end

Tried on 64bit Linux only.

[bobbyjones]

Hey, I think I've found strange bug or I'm missing something.
Anyway the simple line below does not work

Code: Select all

if love.filesystem.getInfo("Religion", "directory") == nil then love.filesystem.createDirectory("Religion") end

It simply refuses to create a folder.
The solutions I've found is to removing "if" part or if I change Religion into religion (from upper to lower r)

But then this piece of code works perfectly:

Code: Select all

if love.filesystem.getInfo("Chars", "directory") == nil then love.filesystem.createDirectory("Chars") end

Tried on 64bit Linux only.

do you have a Religion folder in the game source? love.filesystem.getInfo would check both game source directory and save directory I believe.

[yintercept]

Are you running the game with "--fused" as one of the arguments or no?

I had a folder in the source directory called 'data'. Kept running into trouble.

Turns out the call to open the file was checking the %appdata%/LOVE directory, NOT the data folder in the source, until I ran it with '--fused' (no quotes)