1,989,209 events, 1,149,222 push events, 1,640,704 commit messages, 93,011,376 characters
Created grid() and print_img() so that I can visualize the sudoku boards. Unfortunately getting the puzzle borders set up has been very troublesome and annoying. It's still ugly but you can see the solved puzzle.
holy shit that was annoying ok migrations work again
feature: add death animation
The animation will wait for the hurt animation to finish and the player to be on the floor before it starts.
Issue:
- So... yeah, physics am I right? When you die they kinda brake. At first I thought they were working as intended, but further testing proved otherwise :D But I'm tired, sorry my frend.
My kingdom for a simple working headless Xserver!
The key was:
-
Using Xquartz to get the headers
-
Running Xvfb from that package (we want to be headless)
-
An amazing hack from https://discussions.apple.com/thread/3989835, by "Gratino":
I had the same issue where X11/XQuartz would not start because /tmp/.{X11,ICE,font}-unix were not created properly as an unprivileged user (me). They were removed (along with everything else) from /tmp on startup.
I resolved the issue by putting this line at the bottom of my /etc/rc.local file (runs as root on bootup):
/opt/X11/lib/X11/xinit/privileged_startx.d/10-tmpdirs
thus the needed dirs are created at bootup by root and all is well.
It turns out (from much experimentation) that this is an issue when installing Xquartz in a Github Actions workflow, as we don't want to reboot the image yet the installation assumes that that's what we'll do. Not many people have hit this before, but that's because they're not trying to do headless X11-enabled testing with macOS. We're world leaders in awful kluges!
-
Miscellaneous small hacks, of course.
Update PNG.lua
fuck you builderboy learn how to capitalize properly sped
Update stickers.yml w. HunnieBumbleByStefSchultz (#346)
99352183d0b5f2c4d32e0d0eec2796d7: key: b095df72132d031a5340809a6470c918c662055dbb6f1e9493ab0d03ad1ba6f8 source: 'https://sorrylittlesharky.com/hunniebumblebysteflsharkyschultz.html' tags: - honeybee - bee - bumblebee - bumble bee - cute - girl - dance - bees - skydiving - rock climbing - cooking - burning - funny - sleeping - buddies - love nsfw: false original: true
Co-authored-by: Romain Ricard [email protected]
Create PokemonCards.c
Misty is fond of pokemons and likes to collect pokemon cards. She has P pokemon cards with her at present. She wants to have a particular number of cards on the Dth day. Her friend Ash is ready to help her to achieve the number. Ash will provide her N cards daily so that a day before the Dth day, she will have the required number of Pokemon cards.
Example:
Misty has cards with her, P = 5
D = 6
Ash will provide her with cards daily, N = 4
Ash provides pokemon cards:
Day 1 = 4
Day 2 = 4
Day 3 = 4
Day 4 = 4
Day 5 = 4
Total cards Ash provides = 4 + 4 + 4 + 4 + 4 = 20
Total number of pokemon cards Misty has on the Dth day = 5 + 20 = 25
Misty is busy with her tournament and wants to know the total number of pokemon cards she will have on Dth day. Can you tell her?
Input Format The first line of input consists of the number of test cases, T
The only line of each test case consists of three space-separated integers, P, N and D.
Constraints 1<= T <=100
1<= D <=100000
0 <= P, N <=100000
Output Format For each test case, print the required answer in a separate line.
Sample TestCase 1 Input 2 5 4 6 2 2 2 Output 25 4 Explanation
Test Case 1: As explained in the example.
Test Case 2: P = 2, N = 2, D = 2
Cards provided by Ash on Day 1 = 2
Total number of cards with Misty = 2 + 2 = 4
Time Limit(X): 0.50 sec(s) for each input. Memory Limit: 512 MB Source Limit: 100 KB Allowed Languages: C, C++, C++11, C++14, C#, Java, Java 8, Kotlin, PHP, PHP 7, Python, Python 3, Perl, Ruby, Node Js, Scala, Clojure, Haskell, Lua, Erlang, Swift, VBnet, Js, Objc, Pascal, Go, F#, D, Groovy, Tcl, Ocaml, Smalltalk, Cobol, Racket, Bash, GNU Octave, Rust, Common LISP, R, Julia, Fortran, Ada, Prolog, Icon, Elixir, CoffeeScript, Brainfuck, Pypy, Lolcode, Nim, Picolisp, Pike, pypy3
messed around with select fields; jquery and materialize are crap as hell, docu sucks, fuck off
Wanderer buffs
- Point Aversion bonus damage per point difference increased from 25 to 75.
- Increased boss tier (for boss regen and degen) from 1 to 3/4/5. This means all Wanderers will be harder to kill with Bleeding.
- Each Wanderer kill increases its regen tier. This means Wanderers that spawn after Wanderer 3 (wanderers that give only points) will be way harder to kill with Bleeding.
- Spawning Wanderer with a cheat will set the regen tier to 3.
- Added more comments to ai_wanderer.lua and did some cleanup.
- Wanderers 3 and those that spawn after now have Siltbreaker protection (immunity to most stuns and silences). I am still thinking about if Wanderer 2 should have this as well because normal tier 3 bosses and some tier 4 bosses have this Siltbreaker protection and they are not that hard to kill.
- Disable Sticky Blood proccing if Wanderer is not aggroed.
deletes new one (FUCK YOU GITHUB WEBSITE FOR ADDING THIS SHIT TO MY MUNI PR)
adds the old one back in (AGAIN,FUCK YOU GITHUB WEBSITE)
- fuck you credits;
- fuck you cutscene tweaks;
- fuck the foot particles because i still cant fix it;
Add "one's opinion of me" social statistics
r4558 2018-11-15
- Distinguish between "my opinion of someone" vs.
"one's opinion of me".
- Add a new table showing "one's opinion of me"
- Add getTheirOpinionsOfMe() in RelationshipManager.
- Add "my average opinion of them" score
- Add "their average opinion of me" score
- Rename myFriends() to getMyOpinionsOfThem() in RelationshipManager.
Remove arm64e slice; return dyn linking for CocoaPods (#18)
- Automatically enable Apple Silicon support
Since Xcode 12.2 (stable) is out, Apple Silicon support can be enabled automatically there. Check the version of the command-line tools and set APPLE_SILICON_SUPPORT value automatically if possible.
After a while, once Xcode 12.0.1 and earlier are not longer supported and widely used, this variable can be removed completely. Right now it won't be really necessary to set it explicitly, unless you want something strange.
- Revert "Use static frameworks for CocoaPods (#13)"
This reverts commit d21e3b7ee7ae5eaef8d3bb8355b06d644fdf7f83.
We can't just migrate CLOpenSSL to static frameworks as that requires all upstream dependencies to migrate to static frameworks too, and that breaks builds left, right, and center.
Instead, we have figured out what has been breaking dynamic linkage with CocoaPods, and now we're coming back to using dynamic linkage.
If static versions of CLOpenSSL are published, they will be published in a separate podspec.
- Drop arm64e to work around CocoaPods issues
arm64e builds are included to test the pointer authentication feature of iOS devices. We have to include it in binary CLOpenSSL builds so that upstream users of CLOpenSSL might enable it themselves for testing.
However, inclusion of this architecture slice causes issues with CocoaPods handling of vendored binary frameworks. arm64e builds include a certain linker command -- LC_DYLD_CHAINED_FIXUPS (0x80000034) -- which confuses CocoaPods' detector of dynamic binaries, making it believe that the vendored framework in a static one, not dynamic. This in turn causes issues when using CLOpenSSL as CocoaPods refuses to link "static" binary without "static_framework = true".
This is a know issue in CocoaPods, stemming from the missing features in Homebrew's Mach-O parser 1.
There is nothing we can do about it right now, other than disable "arm64e" builds for the time being. This does not affect deployment to App Store, but will break dependencies of CLOpenSSL which expect this architecture to be present. The dependencies will have to disable arm64e in their projects, if they have it explicitly enabled. (Considering that CocoaPods packaging of CLOpenSSL was never in good shape, the actual impact of this change should be minimal.)
- Set install name during linkage
Instead of using "install_name_tool" to fix the LC_ID_DYLIB value of the dylib, pass the "-install_name" parameter to the linker directly. This doesn't change anything in the resulting binary but looks a bit cleaner.
- Use ABI-correct install name on macOS
On macOS -- contrary to iOS, watchOS, and tvOS -- it is customary for frameworks to provide additional internal structure which helps with binary compatibility. For example, on iOS/watchOS/tvOS a framework typically has flat structure:
openssl.framework
├── Headers
├── Info.plist
└── openssl
while on macOS it's a bit more involved:
frameworks/MacOSX/openssl.framework
├── Headers -> Versions/Current/Headers
├── Resources -> Versions/Current/Resources
├── Versions
│ ├── A
│ └── Current -> A
└── openssl -> Versions/Current/openssl
Note that the top-level files are actually symlinks to the "Versions/Current" which in turn is a symlink to the "A" version, which actually contains the framework content:
openssl.framework/Versions/A
├── Headers
├── Resources
│ └── Info.plist
└── openssl
Currently, the 'install name' of all "openssl" binaries is set to "@rpath/openssl.framework/openssl", which makes the linked binaries remember and use this path when loading OpenSSL. On macOS this will involve two additional symlink resolutions. Another thing here is that if we will need to introduce a different version of OpenSSL framework on macOS, existing applications will still look up the current one, instead of using the A version that they should.
Update the install name computation to use
@rpath/openssl.framework/openssl
for iOS, watchOS, tvOS while using more explicit
@rpath/openssl.framework/Versions/A/openssl
for macOS.
This is how Apple's system frameworks do it. Though they have a pressing issue of supporting multiple possible ABIs and it's not that important in our case, you never know when this turns out to be necessary or whether some weird tool might choke on the symlinks.
Prove stable_sound in new generation
This is super annoying becuase it turns out we could have preserved the
entire set of stable transactions at no cost, and it even would have
given us stability of diskEnd_txn_id (which we currently prove as a
special case of being the last transaction, which is automatically
stable_sound).
stable_sound is also defined in a really Coq-unfriendly way. It should
at least be stated as a property of each individual transaction and then
lifted to the set, which would help us since we take a subset and keep
them stable. nextDiskEnd_inv similarly has no abstraction around the
read-only map, which also makes it painful to work with.