Add CMake option to enable searching for DynLibs in an extra folder#373
Add CMake option to enable searching for DynLibs in an extra folder#373hhyyrylainen wants to merge 2618 commits into
Conversation
…ch the convention
…disc at full length
…ter initialization
…cross scene loads
- Escaped relevant characters from XML comments
…ompact stream of bits
- Added nullptr init and comparison, creation methods, fixed casting
…s to a fixed-size GUI panel
… GUI immediately, instead of having a one-frame delay
…ding export now supported
- Added bs_convert_timet_to_string function for time to string conversion. - Move all time-to-string functions in bs::Time to bs_convert_timet_to_string. - Fix the time not being correctly written to the log output. - Added the function to save the log as a text file. - Modify the Debug::saveLog function to allow for setting the type of the log. - Added a class to manage and get the string names of the categories. - Register all found categories during the framework initialization. - Log verbosity now has a toString function. - Fix the documentation to match the logging system revamp. - Fixed some other minor things.
- Move it to BsString and fixed code that used it.
- Port the code that used the old class to use a instance of the Log class in Debug. - Fix the documentation to reflect the changes.
…ing to release null pointers
…ferencing finishes loading
…d block size, as this is okay in some cases (e.g. mip-maps smaller than block size)
…ly handles rotations over 180 degrees - It no longer forces them to use the shorest path, causing them to wrap around
- It was flipped on the Y axis
Also wraps the makro in do { ... } while(0) to make the compiler
complain about a missing semicolon after a BS_LOG invokation.
I was unable to fully test this as it seems that even though I have my libs in a sub folder but it seems that just loading them by name works. Maybe my RPATH setting comes into play here. At least on Linux.
|
Hi. Can you provide a use-case for this functionality? |
|
Not having to clutter the same folder as the executable with a bunch of libraries is nice. I coded this before I found out that at least on my computer the ld open lib function seems to take RPATH into account. |
|
Makes sense. Although I'd prefer this path is something we provide to |
|
I suppose it makes more sense to have this in the START_UP_DESC. I was in the mindset of adding cmake flags, as I'm going to next do a PR for adding flags for enabling RTTI and exceptions, and leaving debug info alone. So I just made this a flag as well. |
This adds an CMake option that when specifies enables the DynLib
loader to look for the library in an additional folder.
I was unable to fully test this as it seems that even though I have my
libs in a sub folder, it seems that just loading them by name
works. Maybe my RPATH setting comes into play here? This is at least
so on Linux.