Achievement for ClimbHome

Add an additional compatibility layer for video

Fix typo

Move pixel format conversion from uVideoRec to AVWrapper This has several benefits, being in C-land allows us to better use libav API and avoid mixing memory allocated from Pascal. Also the C code for the conversion loop generated by GCC or Clang is probably more optimized than by Freepascal. Finally it will simplify code in the future if we are going to enable any other pixel format than yuv420p. Change the coefficients to improve color accuracy during conversion.

Fix bundle already present check

Add FindOggVorbis to make sure all paths are set

Update osx bundle creation with SDL2 library names

copypasta to make AI use vampirism too

logging via physfs

Merge default

Do not check games lasting for more than 3 hours

german translation: fix error due to confusion of plural "you" with singular "you"

avwrapper: Explicitly check for macro existence

videorec: Drop setting x264 preset It is the default one anyway.

avwrapper: Adjust prefixed symobols version guard

libavinteraction: Adjust version guard for av_codec_is_encoder()

videorec: Drop a commented out option

videorec: Use future API for raw video encoding

videorec: Appropriately rescale timestamps and duration Applies to both audio and video for them to be in sync.

videorec: Always initialize audio and video time bases In both context and stream.

videorec: Fully set frame properties

videorec: Fix flushing encoders When passing NULL frames, encoders empty their queues, but we have to stop polling from them when they don't any more data.

videorec: Make sure the audio packet is considered empty

[bug #1] - Fixed simultaneous pressing of two arrow buttons on mobile

Fix TechRacer not recoding finish if time was >= 100s by raising this limit ten times.

fix pas2c breakage caused by recent changes in uLandGraphics

Break up the hog/object collision. Currently is $7F, allowing 128 overlapping objects accurately. Breaking it up into 15 for hogs, 7 for other objects. I'm thinking the overall accuracy should be just fine as far as people noticing even with a ton of overlapping hogs, and this way we can tell the difference between a hog and "something else". For experiment and rope-breaking purposes, make rope pass through hogs.

fix and/or screwup that was causing girders in tunnels to erase the tunnel. also a little bit of indentation/case fixing

- Enable CI builds with travis for hw iOS

Don't see a good reason for it to be here

Don't dump ghost points if map isn't detected

- Enable scheme sharing for Xcode project, this required for travis builds

Syntax fix

Fix wp issue, hopefully

- Fix room config being sent unmodified - Fix server ghost color

Fix build

Don't compare time value with itself

Recognize ghost points in checker

Finish implementation of ghost points served from server, not tested

updated french lua translation

More work on best time ghost feature

cmake: Set OLD for policy 26 We need to get target location because of our bundled dependencies.

Recognize ghost points in db interaction tool (no storing yet)

- Campaign for iOS: Finally can launch campaign's mission game from list! TODO: Customize UI TODO: Track user progress TODO: Localization TODO: Refactoring TODO: Get some sleep ^_

- Small refactoring for seed command

- Campaign for iOS: Fill campaign page with parsed missions from campaign.ini

Dump best time ghost on console

- Campaign for iOS: Don't forget to add last parsed section in ini file

- Campaign for iOS: Reinventing wheel for campaign.ini parsing

Use secure HTTP for Sparkle appcast URL

Info.plist: Set less important URLs to https

Use secure HTTP for Sparkle appcast URL

Treat #99 point type as a ghost trace

- Campaign for iOS: Stub for CampaignViewController

remove unused path functions

- Campaign for iOS: Simply retrieve list of campaigns

- Campaign for iOS: Stub for list of campaigns

- Campaign for iOS: Added 'Campaign' button on MainMenu screen

Oops, fix whose nick to compare to

Filter out hog speech messages with non-local team index (not tested)