Adarq.org

Quote from: LBSS on May 06, 2014, 04:03:52 pm

Quote from: seifullaah73 on May 06, 2014, 03:03:02 pm

It was a nice day
slight wind from north direction, which be
hitting me from the side and it was a good day as
I found
Something, which had caused to start feeling fast
during the runs compared to the last 2 weeks.
I found
out that I hadn't fully mastered the arm
Swing yet, not in the run but from start to driving out.
I found
that before I was swinging fast when I come out,
but this time I focussed on exploding out first then swinging arms
fast and I felt
fast again,
which was good.

found poetry.

Hall of fame post , easy!

i didn't get it... but now i do...

hah 

Bout to sleep. Spent some time tonight with the streamjs-go port. Pretty much a direct port from javascript to golang, which is sick imo.. garbage collection + scope/lambda's make for a smooth dev environment. ;f

https://github.com/adarqui/streamjs-go

Would like to finish that up tomorrow.. if so, the time spent on that port would be amazingly low (~1 full day) -> which is one of my goals. I want to be able to port software over extremely quick/efficiently.

gn!

Ohhh. More logging. Feels good.

Posted a few projects to git:

A somewhat decent file event cli tool in go. Performs crazy good so far. Had to make a few modifications to the fsnotify library so that I could hook CLOSE_WRITE inotify events. These are important because one use of this tool is to monitor 'file uploads' to a directory(s) and trigger resque events when the uploads are complete. This happens via 'intelligent guess work' instead of the uploader/uploading client triggering some kind of api.. So a CLOSE_WRITE is so far decent when trying to predict that a file has been opened for writing, written to, and then closed (upload behavior).

https://github.com/adarqui/watchque-go


This library is pretty nutty and unfinished.. Trying to port stream.js over to C -> to learn more about implementing lazy evaluation. Lazy eval can be incredibly powerful as I've learned during my research into haskell. So, the library is just a basic starting point which achieves some lazyness, but will eventually need to be re-written/re-designed. I figure, if I could implement a decent lazy list/data structure library for C, then I will have a pretty excellent grasp on lazy eval by that point.... I'd need to lazy eval various data structures, IO streams, etc.

https://github.com/adarqui/streamjs-c



These libraries are part of my 'port of haskell patterns to other languages'.. These are mostly experimental, no idea how they will turn out. I love haskell's libraries; they are on another level (of professionalism). For example, Shellac is a great example. All languages have some kind of third party/open source 'shell framework' for developing cli programs. None of them come close to shellac. Shellac is some beautiful & impressive code. So, one of my goals is to be able to port something like shellac over to C, go, node etc..... Once I'm able to accomplish those tasks, i will also be significantly better at coding haskell - which is another major goal. Want to become extremely skilled in that language. So here are some libraries I recently git'd:

https://github.com/adarqui/Prelude-js
https://github.com/adarqui/Data.Maybe-js
https://github.com/adarqui/Data.List-js

Here's the go Data.List lib I just started tonight:

https://github.com/adarqui/Data.List-go

Tomorrow I will also start the C variation of the lib, which I'm pretty excited about tbh... I'm hoping to implement most all of the C code as MACROS That'll be really dope and somewhat haskell-like in terms of baby-level generics & optimized code (without excessive function calls). So that should be very interesting to play with.


Some small debian build/lxc container, extremely stripped down: ~54M

https://github.com/adarqui/quark


This code is ~3-4 months old. I need to tinker with it near-daily. It's not a project that I would ever 'finish soon'. It's basically an attempt at turning 'ghci' (haskell's interpreter) into a unix-shell-like environment. Sooooo much you could do with functional shells. GHCI is perfect for this project. Some cool stuff in README.txt. For people familiar with linux/unix etc, check out the README.txt for some haskell examples that look exactly like shell commands. You'd think those were shell pipelines etc, but they are haskell functions & haskell function composition pipelines etc.. so sick.

https://github.com/adarqui/HBox




So my biggest focus right now is the 'haskell port' project.. I have lots of libraries to port over to node, go, C etc (these seem to be the languages i've settled in for now). Libraries such as Data.List, Data.Char, Data.Bits, Data.Tree, Data.Set, Shellac, etc etc.. Lots of work to do here. The idea is to provide myself (and others) with a 'unified data pattern' environment for these languages.. ie, calling map's/fold's/bitwise funcs the same (as much as i can) across the languages, so that something I do in haskell will look very similar when I port it to C.

For example, one little project is to write some key/value store similar to redis which exposes my various haskell ports via a cli/api.. That way you could map/fold/use any of the haskell libraries I port over via a cli/web api etc.. Which might be pretty fun/cool. So, I'd like for the code to look very similar between C, go, haskell and node.



Some other projects I have are porting other libraries (non haskell) too.. One such library would be async.js. I think it would be pretty cool to port that library over to C/go, especially go.. ie, using waterfall/series/parallel etc in a very similar manner, but in go.. So ya, i'm trying to abstract alot of patterns out and port them to other langs so that I can hit that next next next next level in eng/arch/programming.. So much work to do

#jumping @ efnet.

peace



edit: hope you feel better LBSS.

damn injuries, it's the only thing I don't miss about training.

hah cool.. ;f

i'm going to start prototyping tonight, most likely.. going to create an 'adarq.org' github project on my git.. then the following directories:

adarq.org/prototyping/{forumjs,forumfs}

I find nodejs the quickest language to prototype in.. I often find myself creating a prototype in node and then porting it to go/c/haskell. Three main requirements of a custom forum to me are: 1) the forum itself is live/event driven 2) completely api/cli based 3) it must also exist as a file system.

The forum file system will be cool... imagine logging in to adarq.org via ssh:

ssh user@adarq.org

Then cd'n to the forum/vim'n your thread:

cd /forum/progress-journals/
less "ADARQ's Journal"/last <-- to view the last post
vim "ADARQ's Journal"/new <-- to create a new post
vim "ADARQ's Journal"/100 <-- to edit post 100

etc

A dope way to search:

grep -ir "half squat" /forum/progress-journals

:F

As long as I prototype first, instead of trying to save the world initially.. I should be able to get the project started.. One of my main focuses with development now is:

simplicity
testing

So the prototype needs to be simple, not necessarily accompanied by unit tests.. the end product however needs to be fully test driven.

;f

i don't usually show love to michigan state, but...

<a href="http://www.youtube.com/watch?v=wGw-v1AFYas" target="_blank">http://www.youtube.com/watch?v=wGw-v1AFYas</a>

 

holy shit hah.

Quote from: adarqui on March 31, 2014, 02:37:33 am

bradley vs pacquiao

martinez vs cotto

mayweather js vs maidana


!!!!!!

i got everyone who i wrote first.. i think bradley wins legit this time vs pac.

cotto could def beat martinez the way he's been fighting lately... if martinez is healthy i got him.. if not, going to be a rough fight for him.. cotto is banging lately.

mayweather jr obviously.. but maidana is no joke.. jr is just unbeatable right now so.. can't wait to see how he dissects maidana.

;f

I see pacquiao outboxing bradley like he did against rios, but bradley doesn't have the knockout power to pose any threat to pacquiao and so i see pacquiao hurting bradley a lot but won't be able to knock him out maybe wobble him like ruslan did, i don't see much from bradley, as he knows how to box but extraordinary skills he doesn't have. so i see it to pacquiao by unanimous decision.

Mayweather obviously would defeat maidana, maidana is slow, flat footed and comes against him straight away, like ortiz mayweather didn't find any difficulty in those type of fighters and wears them down and the last few rounds he finishes them off.

Martinez vs cotto, its close to call, because of martinez's age i feel he may lose as well as coach in the picture, but i don't think they will be able to get past his speed as roach claims and reflex, cotto won't be able to hit him with a good power shot to win.

but they are good fights for sure, also froch vs groves, close last time, they are also hoping to get chavez jr vs golovkin on july, would be a real opponent for golovkin.

oooh chavez jr vs golovkin ??? ? ?

nice..

I guess i'd have to go with chavez jr for that fight.




man bradley was bringing it early on.. after the end of the ~5th round, he was sort of limping.. then the fight changed. His body just can't handle all of those intense slips/movements he does in the ring, sucks..

pacquiao's composure/conditioning was very impressive.. never tired at any moment.

slightly disappointed because i wanted bradley to win.


can't wait for pbf/maidana and cotto/martinez.

Well, it works fine. It has some periods when it's really slow... but fortunately they aren't too often...

It would be interesting to know why those periods exist, though.

daym.. maybe it's the database backups i do? i mysqldump --single-transaction so it shouldn't impact the forum from a lock tables perspective.. but the server isn't super powerful so mabbbe that's what's causing it.