2,351,210 events, 1,170,429 push events, 1,875,846 commit messages, 135,748,657 characters
Add user socials remove alternative for user socials clear
- For some god damn reason I decided to specify you can use remove but I never added it, it's been clear this whole time
"9:15am. I am up. I've been getting used to getting up earlier than usual lately, but today it was not the case.
So let me chill for a little.
9:50am. Let me finally start. What is on the table today?
10am. pathsWithInfo <- traverse (\fp -> (fp,) <$> getInfo fp) inputPaths
Uf, this kind of heavily monadic code is hard even for me.
...Ok, I get it.
10:05am. No actually I do not.
(fp,)
What exactly is supposed to go on the right side here?
The -XTupleSections flag enables Python-style partially applied tuple constructors. For example, the following program
(, True)
is considered to be an alternative notation for the more unwieldy alternative
\x -> (x, True)
Ahh, I see. I was wrong about what this does. Clumsy me.
10:10am. I am looking at the documentation for </> and the ways directories are combined on Windows are quite new to me. It just goes to show there is something new to learn every day.
10:30am.
outputPrimDocs :: Make ()
outputPrimDocs = do
codegenTargets <- asks optionsCodegenTargets
when (S.member Docs codegenTargets) $ for_ Docs.Prim.primModules $ \[email protected]{..} ->
writeJSONFile (outputFilename modName "docs.json") docsMod
Just studying the code like this is interesting, but I'll want to do something real even if it turns out to be HackerRank exercises.
foreignInclude <- case mn `M.lookup` foreigns of
Just _
| not $ requiresForeign m -> do
return Nothing
| otherwise -> do
return $ Just $ Imp.App Nothing (Imp.Var Nothing "require") [Imp.StringLiteral Nothing "./foreign.js"]
Oh so it is possible to use the guards like this. This is new to me. Personally, I'd just use an if statement.
...
Ok...
I had enough.
codegen :: CF.Module CF.Ann -> Docs.Module -> ExternsFile -> SupplyT Make ()
codegen m docs exts = do
let mn = CF.moduleName m
lift $ writeJSONFile (outputFilename mn "externs.json") exts
codegenTargets <- lift $ asks optionsCodegenTargets
when (S.member CoreFn codegenTargets) $ do
let coreFnFile = targetFilename mn CoreFn
json = CFJ.moduleToJSON Paths.version m
lift $ writeJSONFile coreFnFile json
when (S.member JS codegenTargets) $ do
foreignInclude <- case mn `M.lookup` foreigns of
Just _
| not $ requiresForeign m -> do
It is not realistic for me to just study this code, I want to play around with Haskell now that it has come to this.
10:35am.
executables:
hello-exe:
main: Main.hs
source-dirs: app
ghc-options:
- -threaded
- -rtsopts
- -with-rtsopts=-N
dependencies:
- hello
Oh, package.yaml does have a ghc-options section.
10:40am.
2020-03-18 10:41:01.8779772 [ThreadId 1216] - Source error on cradle initialisation: module `Paths_hello' cannot be found locally
2020-03-18 10:41:01.8779772 [ThreadId 1216] - Cradle did not load succesfully
I keep getting this error even though the plugin works. Well, nevermind it.
10:45am. Now it is downloading GHC 8.8.3.
I do not understand why stack build failed for me in the hello folder. I had to recreate it.
Oh, goddamit. Stack and Hie are downloading 8.8.3 separately. What the hell?
...I'll cancel the stack download.
Let me find a HackerRank problem.
https://www.hackerrank.com/challenges/botclean?hr_b=1
Let me go with this.
10:50am.
module Main where
getList :: Int -> IO [String]
getList n = if n==0 then return [] else do i <- getLine; is <- getList(n-1); return (i:is)
next_move :: String -> [String] -> String
--logic here
main :: IO()
main = do
-- Take input
pos <- getLine
board <- getList 5
-- Do I/O stuff here.
putStrLn $ next_move pos board
Oh, it is really nice of them to give me the reader function.
This kind of problem would be absolutely trivial for me in F#, but in Haskell, it will be somewhat challenging.
10:55am. The easiest way to beat this would be to implement A*.
Ugh, Hie is complaining that it wants 8.6.5.
Now an impossible bug happened and it is asking me to report this.
panic! (the 'impossible' happened)\n (GHC version 8.6.5 for x86_64-unknown-mingw32):\n\tDynamic linker not initialised\n\nPlease report this as a GHC bug: http://www.haskell.org/ghc/reportabug\n"
11am. Goddamit. How do I set the ghc version in the yaml file?
The final field is resolver. This tells stack how to build your package: which GHC version to use, versions of package dependencies, and so on. Our value here says to use LTS Haskell version 3.2, which implies GHC 7.10.2 (which is why stack setup installs that version of GHC). There are a number of values you can use for resolver, which we'll cover later.
11:05am.
resolver: ghc-8.6.5
Setting it to this fixes everything.
It is really a good thing that I canceled the download of 8.8.3 and let Hie do it instead.
Now where was I?
11:10am. Right now I am thinking of how to do this problem.
...And am getting annoyed at how more difficult Haskell is making this.
Yeah, this brings back memories. Sigh.
It would be perfectly trivial to do this problem with some help from dictionaries. It would be trivial to implement a dynamic programming solution, but this is one thing Haskell makes annoyingly difficult.
11:15am. Forget memoization or anything else.
What I need here is a logger monad to keep track of the moves taken. Since we are talking only about a 5x5 grid here, I should just be able to do an exhaustive breadth first search. It should not be a problem. Is there a queue in the standard Haskell library somewhere?
11:20am. http://hackage.haskell.org/package/queue-0.1.2/docs/Data-Queue.html
But I can't just import outside packages in HR problems.
...Let me take a break here.
11:35am. Done with the small break. How about I take a big one?
I said yesterday that I would learn enough Haskell to 'do it professionally', but thinking through how I would do the HN problem is showing me just how much pain in the ass programming in Haskell would be.
Last time, back in 2016, I sort of did it all myself, but here if I am going to be doing HN problems seriously, one thing I would like to do is instead look at other people's solutions.
Even if I made up my mind to do this problem, right now I am at a loss on how to do breadth first search in Haskell. Yeah, it is possible, but everything that I can think of seems like a pain.
11:40am. You know what, forget this. Doing these problems is a different sort of challenge from understanding the compiler or the monad transformers.
No doubt HR tier problems would be a lot easier if I was intimately familiar with Haskell's standard library, but for now they are besides the point.
I am going to go back to what I was doing which is studying the compiler.
11:40am. Let me take a real break here. After I am done with breakfast, I'll get back to studying the compiler. Forget everything else. Seriously, what a waste of time Haskell is."
Merge pull request #126 from JonathanSpeek/master
Adding Tech Confessions to Examples
Thanks for this example! (I love this @JonathanSpeek! ) Apologies for the silence. Getting this site and contributions rolling more formally now.
Update Secure Cell API: Seal mode
Start preparing Secure Cell to passphrase API introduction by updating encryption/decryption API. It is not directly related to passphrases, but since we're going to reeducate users about Secure Cell, it's a good point to improve the API.
One annoyance with (all) Secure Cell APIs is how they are mapped into Swift. This is somewhat asymmetrical mapping: wrap and unwrapData, because of some Swift renaming magic. This naming scheme is also inconsistent with every other Themis wrapper on desktop systems which all use encrypt end decrypt.
Update the API to use "encrypt" and "decrypt" naming scheme. Also, provide more overloads for Objective-C to omit optional arguments. Note that we have to hide some API from Swift in order to get better error handling conversion. We need to expose only methods with explicit NSError out-parameters.
The old API is declated deprecated from now on. It is implemented via the new API as a compatibility shim.
While we're here, take this chance to improve inline API docs which make for a better IDE experience.
Improve pull request template proposal (#56756)
Looking through Kibana PRs many currently ignore the check boxes or just delete them. The goal here is to make the process easier, delete some unused options and align the checkboxes with what's most useful. We've had some discussions on the Uptime team about ways the PR template could be potentially improved. These changes are based on an extended discussion we had on the topic. We'd love to hear if other teams would be OK with these recommended changes.
The changes here are:
- Allow authors to just delete unnecessary items rather than strike through. Adding all the
~~in markdown is painful - Remove the unnecessary checkbox for release notes, the build bot catches this and blocks merges without that being set.
- Add a checkbox for testing in various browser sizes / mobile responsive devices
- Move IE checkbox to the bottom of the list since it's seldom checked and makes the checklist seem daunting
"Implement" internal modules as a single import. Also establish somewhat of a "uniform design" for all of them. Let's see how long it will take me to discover arcane magicks known as "interfaces".
Still think it's largely garbage but at least the final result and functionality check out with what I initially wanted.
Switch TS from 3.7 to 3.8 because I'm a retard and couldn't figure out a clean way of exposing internal modules to the main thing. Blessed be "export * as ", for saving idiots as myself.
Update late Work
Hello evryone , Lately this month I worked on the consumption of fuel but I didn't integrate the code in the UI for now because now I want to keep focus on the feature that are already integrated in the project.
The other idea that aren't for now implemented will be integrated later but for now it's time to work on the DOC and the main program.
I had also worked on a way to dynamize the background and I think that after all test that I run that put a Video on kivy with the kind of program that i've made isn't a great idea for many reason : 1- It's make 'laging' small computers as RPI, so it's makes ugly and unusable UI. 2- Python don't use all the cores availables in the UC as I want so It's why I don't want to continue to use video in this kind of case. Maybe in the case where an Kivy app doesn't need to reach info on the web all the time, it's can work but not in our case.
So to resolve this problem, I think I will only put a simple JPG to the background (or a GIF, if it's works).
So now, I'm back and I will work on the doc and new project will comes up soon on my Github.
Thank You, Franck. 🌃 🏩 .
requirements: Bump python-social-auth version.
We had a bunch of ugly hacks to monkey patch things due to upstream being temporarily unmaintained and not merging PRs. Now the project is active again and the fixes have been merged and included in the latest version - so we clean up all that code.
Add Phil Nash (#228)
Hi Phil, Welcome to the list. I just checked them all out and they work great.
Yes, you definitely qualify as awesome. The Criteria is more on 'developers who stream' than the 'awesome' part - I mean Corey is on there ;) (you know I love you really @clw895 )
Also Phil, you get a 🥇 1st place medal for editing your markdown and getting the PR accepted first go. Make sure you mention that to Corey a few times ;) Ooops, sorry @clw895
;)
Blueorbitlife
Experience God with worship songs filled with the Holy Spirit.
Reorder my life, please
Implemented operator precedence as ast tree reorderings Accidentally reinvented some tree structure, can't remember the name Holy shit, it works? Added a proper precedence doc Precedences hardcoded for now Fixed small issue with mangling and demangling of types Modules now own their own symbol table, which acts as the root Fixed SELECT ::[] operator acting on non-identifiers Program AST now printed conditionally based on args
Bosch Group Video- und Bilddatenbankexpertin Deep Learning | SmartRecruiters. Bosch Group Video- und Bilddatenbankexpertin Deep Learning | SmartRecruiters. Bosch Group Video- und Bilddatenbankexpert*in Deep Learning | SmartRecruiters. Deep learning rethink overcomes major obstacle in AI industry. Deep learning for mechanical property evaluation | MIT News. Deep learning for mechanical property evaluation | MIT News. Deep learning-based automated speech detection as a marker of social functioning in late-life depression - Northumbria Research Link. Inter IKEA Group Deep Learning Specialist – Lead Visual data science | SmartRecruiters. Postdoctoral Researcher in Deep Learning for Time Series. Deep learning rethink overcomes major obstacle in AI industry.