3,001,771 events, 1,513,216 push events, 2,412,868 commit messages, 181,774,126 characters
Add files via upload
- Now stepbrother and sister are checked by default
- Added 4 new shower join scenes. Now you can join every member of your family while they're showering. You need to have > 50 corruption points and more than 60 relationship points with parents, and more than 40 relationship points with siblings
- Added a new scene for sissy brother. See the "Help" section
- Added Cuckold Mother scene co-wrote by Rachael and Regent. You need to have > 70 corruption points, > 70 relationship points with mother AND father points. It can only be triggered on Fridays between 21:30-00:00. Also, you must be naked and have had sex with your father and mother more than 3 times
- Now, you can go to school through a link location! (Thanks to Sad+Lonely, our new developer and tester)
- Also, Sad+Lonely fixed a few bugs
- Finally added a pic for sister
- Added 3 new locations - Gym, Night club, Strip club. But only Gym is filled with some content. Content for other locations will be added later
- Increased the chance of granddad catching you at the farm
- Fixed the Problem Solver: Everyday check pregnancy error
- Updated SugarCube version from 2.34.1 to 2.35.0!
Fuck you Aaron; I added Consistent Code and changed locations of files; fuck
[libpsyseg] Add proof of concept library
Basically something I had in mind since forever, and never sat down to bash out. It's a very simple idea of prepending files with certain metadata so that you can have a portable way to sort them.
The files I'm interested in are usually media files such as pictures. Pictures have many formats, and over time have evolved to allow for the possibility of metadata.
However, different formats have their own ways of implementing metadata, and sometimes said formats will not support metadata at all, which make maintaining said collections much more annoying.
So I wondered how easy it would be, to write a container like format for image files, that could store the metadata in the header, and thus make it portable quite easilly instead of fiddling around with a database (which you'd have to remember carrying over to another computer for example).
The next question is how to view these files, as they'll most likely break the common viewers. For this I have two ideas:
- patch a custom
fehto simply skip over the metadata, and use its usual viewing routines - write a simple SDL viewer for media
Another aspect that I want to piggyback is appending gpg signatures to pictures. This is for another idea I had, where artists could claim certain works as their own (as there is quite a bit of art theft on the internet).
That's it for now I guess.
Fix hasMessageAvailable return true but can't read message (#10414)
I temporarily fixed this problem in PR apache/pulsar#10190. Now we have found a better way, this way can avoid the seek, then avoid trigger another reconnection. Thank you @codelipenghui to troubleshoot this issue with me all night.
We have added a lot of log and found that this issue is caused by some race condition problems. Here is the first reason: https://github.com/apache/pulsar/blob/f2d72c9fc13a33df584ec1bd96a4c147774b858d/pulsar-client/src/main/java/org/apache/pulsar/client/impl/ConsumerImpl.java#L1808-L1818 Now we have an acknowledgmentsGroupingTracker to filter duplicate messages, and this Tracker will be cleaned up after seek.
However, it is possible that the connection is ready and Broker has pushed message, but acknowledgmentsGroupingTracker.flushAndClean(); has not been executed yet.
Finally hasMessageAvailableAsync returns true, but the message cannot be read because it is filtered by the acknowledgmentsGroupingTracker
clean the tracker when connection was open
Preliminary Blood Magic and Occultism Progression
Reworked ID removals to use ID replacements where valid. This requires an update to KubeJS.
Eidolon will open up the Blood Altar, Alchemy Table, and Low Grade rituals in Occultism, opening up some extra crafting mechanics. Blood Altar Tier 2 and up will require parallel progress in Occultism.
Petal Apothecary can make Sprouted Fungus, making a simpler way to create potions. I'm thinking enchanting should be pretty late game, considering the power level. We have a lot of really fancy potions and a cool new system with Ars Nouveau for making some very powerful brews, but enchants almost entirely invalidate potions. By making them easier and enchanting later, I'd hope to get people to use them a bit more.
Alchemy Table recipe for Nocturnal Powder and Illumination powder added. Nice reagents for other crafting, and very handy when searching for mobs early on.
Cutting Fluid made a bit easier, but more varied. Now requires snake rattles (farmable), apatite dust (goes a long way), charcoal dust, and plant oil. This should hopefully give magic players a decently automatable ore doubling system.
Considering boosting output for ores, considering Create's Milling is also fairly early game.
Otherstone now crafted from Arcane Stone
Leather Stripe now a cutting board recipe
Spirit Attuned gem now made from Shadow Gem
New Recipe for Divination Rod
Impure White chalk moved to Alchemy Table
White Chalk moved to Blood Altar
Occultism Bowl recipes moved to Enchanting Apparatus
oh my fucking god i hate my parents
internet turns off at 9, and i figured out how to do what i wanted to do at 8:55, so uh yeah
also fuck not having a database and using actual files that i have gitignored
did it go through now? i swear to god git is as ass as snv, linus torvalds didn't fix SHIT
ZEN: Set MuQSS' yield_type to 0 (don't yield)
Turns out when I investigated performance issues with RPCS3 on Linux with MuQSS, my choice to set yield_type to 2 was flawed since I didn't benchmark or any other applications that cared about it.
Phoronix wrote an article measure performance of Liquorix against a stock 5.4 configuration here: https://www.phoronix.com/vr.php?view=28750
All the benchmarks measured framerate and Liquorix for the most part got up to 20% less FPS than stock CFS, depending on the game. Turns out some of it had to do with yield_type, and always yielding when requested dropped minimum frame times quite a bit. Disabling yield entirely raised the average frame rate a bit and the minimum frametimes on Deus Ex: Mankind Divided by nearly 10%.
Also, Linus Torvalds wrote in a forum about sched_yield. He indicated that yield used to make sense on uniprocessor configurations, but now with multi-core being the norm, yield almost always causes performance issues due to cache thrashing and thread/process migration on multicore systems: https://www.realworldtech.com/forum/?threadid=189711&curpostid=189752
And finally, even if we don't yield, MuQSS will reschedule the thread that's spinning anyway. All setting yield_type to 2 did was reschedule the thread sooner. Lets let MuQSS decide when a thread needs to be rescheduled, not the program.
Previous commit messages:
muqss, zen: Set yield_type to 2
In my previous attempt in making games and emulators behave better, I found that making sched_yield a no-op by setting yield_type to 0 caused emulators like RPCS3 to perform far better without frame drops. Unfortunately, on my next boot I noticed the whole system felt less responsive in general.
Fortunately, always expiring the timeslice of the yielding process and calculating the next deadline also fixes this weird performance anomaly. Setting yield_type to 2 with a reduced rr_interval (2ms), should help offset the disadvantages of rescheduling a process that has yielded.
muqss, zen: Disable sched_yield by default
While experimenting with audio stutter and frame drops through the RPCS3 emulator, I found that yet again, sched_yield is the source of unpredictable performance drops. Reading the original post[1] by Con where he outlined his last (final?) change to sched_yield, he changed the behavior of sched_yield in MuQSS to only yield to better priority or deadline tasks.
It seems then on my system, emulators and games are yielding to some number of higher priority / next deadline tasks and performing much worse. If you look at the comments on the blog post, however, many people found that setting yield_type to 0 improve performance on their games.
Lets just turn sched_yield into a noop for now and see how that fairs. Worst case we'll receive a report about an application behaving badly, and best case everyone will get a net benefit in gaming and emulators.
[1] http://ck-hack.blogspot.com/2016/12/linux-49-ck1-muqss-version-0150.html
Revert "muqss, zen: Disable sched_yield by default"
This reverts commit 8ec985bcd0ded1bcdd8bb999c90c094dc9536a0b.
This change, as expected, has strange side-effects with system responsiveness when many applications are launching at boot. Also at random times, the mouse becomes unresponsive while over saturating CPU bandwidth.
Leetcode 665 Non decreasing array
As someone that hates python i sure as fuck love solving questions in PyThoN
46.2. useTransition in a Neat, Reusable Module
The skipHiddenProperty Value Anyways, back at the browser, type to re-open the suggestions then click off of to close it. That happened instantly! Where was our transition? Inspect the element. Ah: see that little hidden attribute on the results div? That was added by the stimulus-autocomplete controller as soon as we clicked off. Thanks that, the element became hidden instantly instead of waiting for our transition. Normally, that's great! It's how stimulus-autocomplete hides the results. But now that we are controlling the hiding and showing with our transition behavior, we do not want this hidden attribute to be added. Fortunately, assuming my PR is merged, we can pass a value to disable that behavior. In the template, on the autocomplete controller, pass a new value called skipHiddenProperty set to true. That literally says: please do not set that hidden property: we are handling the hiding and showing ourselves. Let's try it out again. I'll type... we still get the nice fade in... and when I click off. Yes! It fades out! And.... we're done! I mean, the whole tutorial is done! I hope you found this journey through Stimulus as refreshing as I did. I love coding with Stimulus. In the next tutorial in the series - about Turbo - I hope to show prove that we can have an even more dynamic and speedy app while writing even less custom JavaScript. Let us know what cool stuff your building. And, as always, if you have any questions, we're here for you in the comments section. Okay, friends, seeya next time!
42.1. CSS Transition Classes
Right now, the search preview results are hiding and showing correctly... but there are no CSS transitions yet. Why not? Because... adding the transition is... actually up to us! By defining it in CSS.
Adding the Transition Classes Go back to assets/styles/app.css and head to the bottom. I'm going to paste in three new CSS rules. This 2000 milliseconds here is probably too slow... but it will make it easy to see how the feature works. Before we talk about what's going on... it should already work! Move over and refresh the page. Type and... beautiful! It faded in! When I click off, it fades out! Amazing!
The Class Lifecycle of Showing an Element But this... deserves some explanation. Back in the controller, we defined six classes that useTransition should use. The option keys come from useTransition, but I totally made up the class names on the right. Though... they make sense! Because we're going to create a "fade" transition, each class starts with fade and then matches the option name it relates to. Anyways, move back to the CSS file where we define the style for these classes. Here's the magic. When we call this.enter(), useTransition immediately adds the fade-enter-active class. That doesn't cause a transition, but it establishes that, if the opacity changes, we want it to transition over 2000 milliseconds. One frame later, it adds another class - fade-enter-from - and removes the d-none class. The result is that the element is now "shown"... but with an opacity set to 0. One frame after that, it removes fade-enter-from but adds fade-enter-to. Thanks to this, our browser starts transitioning the opacity from 0 to 1! Awesome! So... what happens next? useTransition is smart. It detects that a transition is currently happening and will take 2000 milliseconds. So... it waits. Yup! It literally waits for two seconds for the transition to finish. And then it removes both fade-enter-to and fade-enter-active because its work is done. The element faded in and is now fully visible. Isn't that amazing? stimulus-use didn't invent this idea: you'll see it in other libraries like Vue. But it is so handy.
The Class Lifecycle of Hiding an Element In our controller, when we call this.leave() to hide the element, a similar process happens. First, fade-leave-active is added to the element, which establishes that we want a 2000 millisecond transition on opacity. Next, it adds fade-leave-from, which makes sure that the opacity is definitely set to 1, which it already was. One frame later, it removes fade-leave-from and replaces it with fade-leave-to. The result is that the element starts a 2 second opacity transition from 1 to 0. Two seconds later, after the transition has finished, useTransition adds the d-none class and removes both fade-leave-to and fade-leave-active. The element is now fully hidden. How cool is that? Learning how this works is fun. But the result is even better. And in your day-to-day use, it's really simple. Now that we have these three CSS rules defined, we could reuse this exact useTransition in any other controller to add, fade in and fade out functionality to it. Heck, you could even create a re-usable JavaScript module that sets up the behavior and these options automatically for you! Next: there's one last thing I want to talk about. Stimulus is used by a lot of people, including the Ruby on Rails world. And so, it turns out that there are a bunch of pre-made Stimulus controllers that you can download and use directly in your app! Yay! Let's install one and learn how to register it with our Stimulus application.
41.1. CSS Transitions with useTransition -- yarn upgrade stimulus-use@beta
I want to add a CSS transition so that our search preview fades in and out, instead of showing up immediately. This may seem like a small detail, but this kind of stuff can be the difference between a mediocre interface and a truly beautiful experience.
Why CSS Transitions can be Tricky So how can we do this? We learned some CSS transition basics earlier. In assets/styles/app.css, if you look for cart-item... we set transition: opacity 500ms. When that element also has a removing class, we set opacity to 0. The result is that, as soon as we added this removing class to an element that already has the cart-item class, that element's opacity changed from 1 to 0 over 500 milliseconds. So CSS transitions are simple, right? Well... not always. If we repeated that trick here, we could transition the opacity of this element to 0. But then... the element would still technically be there... it would just be invisible. So if the user clicked right here, to them, it would look like they were clicking this photo. But in reality, they would be clicking the invisible dropdown above, which would take them to the disappearing ink pens page. Ok then! So we just need to fade this element to opacity 0, wait 500 milliseconds for the transition to finish... and then fully hide the element. Right? Yea! That's exactly what we need to do! But sheesh, this is getting complicated! And if, later, we decided to change the transition in CSS from 500 milliseconds to 1 second, we would need to remember to go into our Stimulus controller and also change the delay there to 1 second... so that we don't hide the element before it's finished fading out. So that is why CSS transitions are trickier than they seem at first. Fortunately, the stimulus-use library we installed earlier - to get behaviors like clickOutside and dispatch - has a solution for us! Head to their docs and find a behavior called useTransition. This is a brand new feature that allows us to add CSS transitions when we need to hide or show an element. At the time of recording, this is currently a beta feature... which means it could change a bit - but I'll update the video if that happens.
Upgrading stimulus-use The feature was introduced in version 0.24.0... and it's so new that I need to upgrade my stimulus-use to get it!
Find your terminal and run:
yarn upgrade stimulus-use@beta I'm using @beta because the feature isn't included yet in a stable release at the time of this recording. The 0.24.0-1 version is a beta release. Once they release 0.24 stable or higher, you won't need that. Open up the current search-preview_controller. As a reminder, as we type, this makes an Ajax call to an endpoint that returns the HTML of the search preview, which we then put into resultTarget. That's this element right here.
44.4. Using the autocomplete-controller
Making third-Party Controllers Lazy I have one more wish... or question: could we make this autocomplete controller load lazily? Earlier we made the whole chart-js controller "lazy" in controllers.json by setting fetch: 'lazy'. We also made our own submit-confirm_controller lazy by adding this special comment above the class. But what about a third-party controller? We don't register this in controllers.json and... we can't exactly open up the file and add a comment in it. So, can we make it lazy? We can! Though, due to some rigidness in how Webpack works, the syntax... isn't amazing. In bootstrap.js, we need to change the import to pass through the special @symfony/stimulus-bridge/lazy-controller-loader. You can do that by literally saying import { Autocomplete } from... - the name of that loader - an exclamation point, and then the name of the module that you want to import. So, this module will now be passed through the loader. That... by itself, won't change anything. But now, before the exclamation point, add ?lazy=true. The lazy-controller-loader is what looks for the stimulusFetch comment above the controller. Since that won't be there, this ?lazy=true is our way to force laziness. And, normally, this would be all we need! It's a bit ugly, but not too bad! And we get the laziness we want. However, notice that the stimulus-autocomplete module uses a named export instead of a default export. What I mean is, when we use this, we have to say import { Autocomplete } - that's a named export - instead of just being able to say import Autocomplete. If we tried this, it would not work. Anyways, since the library uses a named export, we need to notify the loader about this so it knows where to find the code. We do that by adding one more loader option: &export=Autocomplete: the name of the import.
"11:45am. Ok...I want a break. Right now, for the first time in my life, it feels like I have a complete foundation. Literally everything is there to make the poker agent. With today's test, I've cleared the hardest hurdle when it comes to using neural CFR.
11:55am. I have an answer for quite a lot of things now. For the optimization parts of the monthly review, I reasoned them out from first principles so I am confident about them. I was wrong about noise keying, but I was right in general about ensemble averaging. So I can still consider my intuition being sharp here.
12pm. The duality gap method should be enough stabilize GANs so for the first time ML practictioners a viable path towards unsupervised representation learning in real life circumstances.
Uncertainty estimation and deep exploration can be done via ensembles. Via dropout and keeping the key steady to a lesser extent as well.
What is missing for AGI currently? How to do credit assignment over extremely long distances. Better NN models will be needed for dealing with long term memory. And also reasoning, but GNNs are being researched. This is not something I can jump into just yet.
Also what we are missing is hardware. GPUs are ultimately not scalable and I need better hardware to progress.
What I have now is just enough for poker and not much else.
12:05pm. GANs are just the first step, we also need to direct the unsupervised learning models towards optimizing representations that would be good for dealing with subgoals.
Overall, things could be summed up that we need better hierachical learning.
After six years of effort, I've just broken into the very first realm of skill. But this first level is finally the place where I can do actual magic. Previously I could only dream about doing it.
12:10pm. Now I need to put myself back on track. I need to compile the Leduc game to a representation that NNs can deal with.
12:15pm. Let me slack here a bit and do my reading. I've been under extreme pressure to figure things out for the past two weeks and it feels like my mind has finally calmed down. I've always been determined and persistent, but now I am starting to feel hope and confidence as well.
Let me gather my will for the next part. Slowly and steadily, I should just go forward without stopping until the agent is complete. There is nothing that can stop me from winning."
bug fix - Fucking .equals istedet for god damn == fuck cs og fuck java
i fuck up this life, that tottally shit component and to refactor this shit and alother shit, main functionality was maded
Perry and Perry's Annoying and Incoherent Student
Perry is a saint. He is perfect and amazing, unlike his idiot of a student, who can't even formulate a sentence on her own. She didn't type this, and if she did, Perry was probably spell and punctuation checking every millisecond. She is that much of an idiot.
Merge pull request #3 from 6ACBTA35/juliocesar
Subiendo rola de Queen Love of My Life. Todo bien y listo.
WTF IS GOING ON? IT WORKS!
I DON'T FUCKING UNDERSTAND, WTF IS GOING ON, WHY DOESN'T IT WORK ANOTHER WAY, BUT THIS WAY THIS FUCKING STUPID INEFFICIENT SHIT WORKS!
There are some special moments in your life when you realize that it's better changing life and learn new stuff. To all the person who helped me through this wonderful experience, I'd like to say thanks. Thank you all guyz ly
Merge pull request #3 from rev1853/revan-1
fuck you
mostly [modular] renames brig off to correction officer, on request of gandalf (#5453)
-
[semi-modular] Renames Brig Officer to Correction Officer on request of Gandalf
-
Update jobs.dm
-
my name i
-
s Patrick, I'm 31, and I'm not here to cause harm. Please do not contact the police , please don't alert the authorities, please don't freak out. I'm not going to kill myself or hurt anyone or anything like that, but please don't freak out. My name is Patr
-
go on
-
???
-
Update security.dm
-
more, more
-
Dear top I'm writing you this tomato I'm writing this email to make sure you know that I tried to make a fund video unfortunately, unfortunately that didn't turn out so well my hypothesis in my experience Haitians were not happening as well as I plant and I'm really sorry you'll never happen again, please don't fire me I hope we can still be best friends your if if fewer from your best friend sure not now if if if if if, Thank you again,. Sharma 908 auburn 500 Joe favor employee
Co-authored-by: louiseedwardstuart
Ancient of the Burning Wastes, Mummy Improvements, and Insight Monster improvements
Misc: Add comments and fix a macro mistake in other Ancient breaths.
Ancient of the Burning Wastes -mhurtle(): Add a flag that will cause huge monsters to be thrown. -Explosion effects are now compatible with Elemental damage types, monster exploding code simplified. -Casts spacewarp: blades and spacewarp: throw -Inhales water (or blood) and exhales heat and ruin (effects non heat resistant or nonliving targets) --Exhalation relocates the ancient. -Melee attack inflicts mummy rot (see below)
Mummy Improvements -mummy_curses_x --Called from the xp functions (xkilled and grow_up). --Called from AT_NONE AD_MROT. Can also be used as a melee attack damage type AD_ROT. --Bad luck: lower luck to minimum --Curse equipment: curses ~1/4 of open inventory --Pain: Halves current HP (doesn't kill) --Insect bite: inflicts Sickness --Mummy rot is a persistant stat damaging effect that can be cured in much the same way that WA contamination and morgul shards can. --Heart attack: Instant death if < 100 HP, 10d8 otherwise. --Remove protection: Halves intrinsic protection. --Reduce maxHP: Halve max HP bonus or reduce by 25. -Add soldier and priest mummies -The Hmnyw pharaoh, regular pharaohs, soldier, and priest mummies have curses, generic mummies do not.
-grow_up() audit: grow_up() sometimes kills things, but was typically used in ways where that will cause problems. -Spacewarp: throw: monster spell, flings target in a random direction
Insight Monsters -The Hmnyw-pharaoh and the good neighbor still buff their followers when they're visible. -Slightly tune down the rate at which Hwnyw revives Nitocris, so that the player can take steps to destroy her body more easily.