Opengl improvements backport

See merge request !304

Update readme

See merge request !313

Update copyright info

See merge request !309

Minor remote viewpoint bugfixes.

Some code cleanup.

See merge request !298

Winsock error handling fixes

See merge request !303

File loading hotfix

See merge request !310

Make sure handles of files opened by W_LoadWadFile are closed if we abort loading the files for whatever reason.

ESPECIALLY if the file is already loaded in SRB2, that's just silly.

Patch offsets fix

See merge request !305

MI minor refactoring

See merge request !294

Multi-patch texture support for transparency AND translucency

See merge request !293

now that V_FLIP's effects are always supported in V_DrawFixedPatch, there's no need for the "flip" variable anymore

HWR_DrawFixedPatch: Add V_OFFSET support for V_FLIP in hardware code as well, also add missing SHORTs for gpatch fields

Make V_OFFSET no thing without V_NOSCALESTART, to remove any potential glitches with using it on its own

Removed all remaining traces of VID_X11 code in hw_drv.h and hw_data.h, the macro is no longer used by Linux etc versions of SRB2.

Unlike the rest of the commits in this branch (as of writing), I didn't make this commit between 1 and 2 years ago, I made it right now ;)

Remove pointless drawtextured variable and redundant fake planes checks (R_FakeFlat would already have made gr_frontsector/backsector something else if they were)

Added missing checks in HWR_AddLine from the software version, move v** vars to the top since R_PointToAngle calls use the same values anyway

this fixes a crash in (old) GFZ2 at the ramp as a result of creating pv1/pv2. This probably means before pv1/pv2 there could have been some silly typecasting from vertex_t to polyvertex_t to get fixed vertex coords and such...

Yes I know they're void * in r_defs.h's seg_t definition, it's quicker than trying to figure out if including hardware/hw_glob.h is a good idea or not

I also optimised those two functions while I was there (why keep a "floating" variable when setting it to false guarantees the functions return false?)

gr_correcttricks fix: don't check if top/bottom textures are missing for sloped sectors, just ignore and cancel the hack

This fixes GFZ2's mysterious flying flats at the ramp to the big room

Add MD2_INDENT and MD2_VERSION so we can cleanly check that it's a valid MD2 from magic number/version

Use temporary buffer with a max size of 255 bytes instead of having Microsoft's FormatMessageA alloc one for us. Also, provide a fallback message in case no message was available for some reason

Save the result of errno (aka WSAGetLastError() for WinSock) as soon as possible, to prevent anything in SOCK_GetNodeAddress resetting the value to 0 while trying to print the message for the error itself!

Creating a quick get_WSAErrorStr function to act as a wrapper for FormatMessageA so we can string-ify Winsock errors properly

Untested!

mapheader lua unarchive hotfix

See merge request !302