Comments URL: https://news.ycombinator.com/item?id=47521964

Points: 4

# Comments: 0

http://ocapn.org/

Though I agree, that although not a technical justification, an explanation as to why there are no tests is because Apache HTTP is from the 90's. Not writing unit tests was par for course back then. Most FLOSS code bases in the 90s didn't have unit tests, let a alone a CI to run the test suite for each change. Adding tests later is hard. Though there are some tests under the test folder.

> Flagging old live (i.e., potentially immortal) allocations is easier (https://github.com/backtrace-labs/poireau) and more closely matches the goal of identifying why the heap grows instead of entering a steady state.

Ej. Why is Sidekiq memory consumption growing over time.

[0]: https://github.com/0intro/wmii [1]: https://github.com/0intro/wmii/blob/main/cmd/wmii.rc.rc [2]: https://github.com/sunaku/wmiirc

https://m.youtube.com/watch?v=1vui-LupKJI

Comments URL: https://news.ycombinator.com/item?id=41449625

Points: 165

# Comments: 71

Kara-cancel is a mechanic that lets you extend the range of your moves, in 3s is used for throws.

So the input for throws in 3s is lp+lk. Now, what happens if while one is trying to press the buttons at the same time they press one slighty before the other? A move will start to come out and then you can't throw because you are doing another move. To make it easier to input throws, devs made so that _any_ move can be canceled into a throw in the first 5 frames of the move. 5 frames is 5/60ths of a second.

Separate to that, some moves move the character forward. Ej. Chunli's HK. So people figured out that if you press a move that moved your character forward and then canceled that into a throw you can extend the effective range of your throw.

Mind you doing this means pressing a button 83 milliseconds before the other one. Which is of course not something you can do by thinking about it, instead you learn to position your hand in a way so that when you move it down together one finger lands before the other two. The name kara-cancels comes from the Japanese word for empty, because you are canceling a move that never came out.

Now I don't know the history, whether the mechanic was first found in 3s and then in SF2T or not, but it is an example of a mechanic intended to ease the input of something being used to expand the toolkit of a character.

V-cancel (not sure if that was the name ppl used for it, didn't play sfa2) refers to the fact that in sfa2 the number of frames to go from standing to downblock is more that the number of frames a character needs to go from standing to a low attack if the character is in v-ism. This means that if two characters are standing next to each other and one activates their v super, they have a guaranteed hit.

This was first used by Ricki Ortiz in a tournament setting in a finals and that is how it became wildly know. The story of it was documented in Sirlin's Play to Win book, which is how I learned about it. https://www.sirlin.net/ptw

Yes, that it something really important that the modern landscape of competitive gaming lost and the frequency of patching increasing. Not just in BW, older fighting games, especially those arcade based ones, people found ways (which sometimes blur the lines between exploit or technique) to keep pushing the bounders. Things like Kara-cancels in 3s, were one is using a mechanic meant to facilitate the throw input to extend the range of the throw. Or wave dashing. It is wild how even character tier lists changed in the original SmashBros 64.

And also how people adapt to this meta changing discoveries, like the bisu-build in BW or when Ricki Ortiz unleashed v-cancel in sfa2!

And is true that most package managers for popular language allow arbitrary code execution during the install process. That is how husky adds git hooks to the developers machines.

For example in Ruby I need to patch the Kafka gem, karafka because it downloads, builds and stores librdkafa.so in the gem's directory.

I understand that this as well as the husky example comes from a desire to make developer lifes easier but I'd rather we erred on the side of caution. Making sure that software builds without access to the network and without being able to modify your system (ej. Adding files to $HOME)

The sly debugger is better than most debuggers, but calling it the most powerful debugging experience is a stretch. Just try the debuger in a Smalltalk implementation (Squeak/Pharo/Cuis)

https://github.com/PuercoPop/.emacs.d/blob/9d0b99d332d619fe3...

> Other than using version control to determine what are the sources of the derivation

I don't think that defaulting to using git to list the sources when inside a git repo means that it is tightly coupled to git. If one doesn't want to commit/stage the flake files in the repo one can do `nix develop .` and it works . It will take a little longer because it will copy the entire directory to the store iiuc.

> Flakes are tightly coupled to Git.

How is it tightly coupled to Git if it supports Mercurial out of the box?

Other than using version control to determine what are the sources of the derivation, what does flakes lean on version control?

> Also, dcop is probably the best GUI automation tool I’ve used.

Everytime I have to use dbus from the CLI I think to myself, dcop died for this?!

To me the best GUI automation tool (in Linux at least) was Kommander, another KDE app that didn't survived the KDE3-4 transition. A pretty decent 'low-code' tool that unfortunately didn't make it to KDE4.

I used KDE4 up to 4.2, when it supposedly 'got good'. It wasn't bad by any means. But KDE3 was better.

Not 2002, 2008. Trinity was forked after KDE3 was EOL.

I'm curious if they are keeping DCOP alive or have they embraced dbus?