Mostly (lists of (lists of ...(useful links))), mostly programming-related
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

lrn :lrn:


learn by doing


IP intellectual property

from Icon mag: in europe, of 10k food products launched every year, 50% fail within 3 months, 95% fail within 2 years



  • make sure get VIN number before buying a car

lrn-clothing :lrn_clothing:



apparently loading data from a binary instead of a plaintext is 5-20x faster?

The Big List of Naughty Strings (strings likely to cause issues as input) | Hacker News

diagramming software: lucid charts chrome app




  • chromebrew relies on insecure hosts for dl'd packages




lrn-containerization lrn-docker

lrn-crypto lrn-encryption




  • (Note to anyone reading: I will breathe vim until my dying breath, but after thinking long and hard about it, for a language like Python that has such a powerful, FOSS IDE like PyCharm, the time, complexity, stability, and cross-platform cost to get a similar environment working is just too high.
    • also, "org-mode" alone justifies at least some usage of Emacs/Spacemacs ;)
  • :( in thinking about vim autocomplete working with libraries etc., didn't even think about if they have anaconda/environment support...which this says it's more trouble than it's worth
  • more likewise vim failure comments
  • These and similar posts on r python make it VERY clear that the plugins powering python in vim are not as feature complete, integrated, or powerful as those in a real ide like pycharm
  • steven yegge ( says, in your editor, if where you need to go is > 5 lines away, you should be using incremental search, not moving there. just as well applies to vim
  • vim notes
    • just discovered that in vim, can search for newlines with 'n' a la s/n/???/g, but only way to insert a newline is with 'r', a la s/n n/r/g
    • you can get the values of the current underlying character via :as
    • how to delete all but X...Y in vim: g!/.*cosampl0pt7.*15001.*/d
    • sliming in vim via tmux
    • libs (plugins, really)
      • autocompletion
      • file finder
        • Unite.vim (probably best)
        • CommandT (fast, but need to painfully compile Ruby, build with vim ruby support, etc.)
        • CtrlP (oldest? slowest? in VimScript)
      • snippets
        • firstly, see the popular, cross-compatible snippet store at
        • UltiSnips (by default incompatible with YouCompleteMe, but can work well via SuperTab? or .vimrc changes?)
        • Neosnippet (same author as Neocomplete, probably works well with Neocomplete?)
    • incremental search / replace: :let i=1 | 38,46g/foo/s//="bar_".i/ | let i=i+1
    • insert date: :r !date
    • quick paragraph prettify gqap
    • folding, 'zfap' folds the paragraph, 'zo' opens, 'zc' closes
    • :abbr pn penguin
    • in insert, ctrl-t adds an indent to the line?
    • read json: :r! curl -s "some_urltojson" then :%! python -m json.tool
  • emacs notes
    • general emacs notes
      • "redo" is weird, some kind of circular buffer
      • to switch the "direction" of un/redoing, hit C-r it seems
      • how to browse websites?
      • mark (for region select) is set by C-space (since I had to rebind tmux :()
      • due to super different emacs indention, in Insert mode, can only force a "new" indent via M-i
    • tutorial notes (a la emacs mode)
      • C-v and M-v move forward and back a screenful (save for 2 lines so you can read continuously)
      • h and l in vim correspond to C-b and C-f
      • j and k in vim correspond to C-n and C-p
      • b and e (words) in vim " to M-b and M-f
      • "meta is usu acting on objects of the language, like sentences, whereas ctrl is usually for raw text"
      • 0 and $(9, for me) in vim " C-a and C-e
      • ( and ) (sentences) in vim " M-a and M-e (Emacs' sentence detection is far superior)
      • the cursor location is the "point"
      • M-< and M-> move to the beginning/end of buffer
      • instead of just prefixing command with a number (since it would type the number), either "M- " or prefixing with "C-u " works; the latter is recommended
        • the number in this case is called a "prefix argument"
          • for some commands, the mere presence of a number changes the fundamental behavior of the commands, meaning it acts as a "flag" instead - this is not true of anything so far
        • thus, C-u 8 C-f moves forward 8 chars.
        • this can also be used for individual characters, like C-u 8 a
        • there are some exceptions, like how "C-u C-v" SCROLLS by lines
      • C-g is sort of the universal "interrupt and exit command entering"
      • some commands are disabled for new users, where typing them opens a window explaining them
      • Emacs, according to its descriptions, appears to have vim-like newlines, where they're simply treated as their own character
      • windows
        • C-x 1 closes all windows except one
      • killing/yanking = cutting/pasting
      • deletion
        • Emacs calls backspace , NOT the same as the usual delete key
          • despite this nomenclature, it is not considered a control character
        • Emacs can use C-d as the "delete next character" key, or colloquial delete
        • M- deletes the previous WORD, and M-d deletes thee next word
        • when any of these are given an argument, they kill instead of deleting
      • killing
        • "killing" means you can reinsert the text, while you can only undo deletion
        • C-k kills from cursor to end of line, and M-k kills to the end of the sentence
          • this means, to kill a line AND the newline at the end, you need to hit C-k twice; however, C-u 2 C-k will instead kill 2 whole lines AND their newlines
        • C-w kills a region, selected by making a mark via hitting C-space, and then moving the cursor
      • yanking
        • "yanking" is the REINSERTION of killed text
        • C-y is the basic yank command
        • once text is yanked in, can hit M-y repeatedly to exchange the text with earlier kills
      • undo
        • C-/ is the basic one
          • C-_ and C-x u are both identical to this
          • all equivalences accept a numeric prefix argument as the number of undos
        • just like vim, undoing has no effect on the kill ring
        • just like vim, commands that don't change the text aren't undone
      • files
        • opening files are called "finding" or "visiting" them
        • C-x C-f opens a "minibuffer" for looking for a file
          • has rudimentary text completion, including "ls" of the dir in a new window upon TABBING
          • this is also how you create new files
        • C-x C-s saves changes to a buffer to a file
      • buffers
        • C-x C-b lists all the buffers
        • ALL things with text in Emacs exist in buffers
        • you can reopen a buffer you have been editing with C-x C-f, same as finding the file, or C-x b
        • not all buffers are files, like Buffer List, and Messages, and scratch
        • you are free to navigate to other buffers without saving your changes in some
        • C-x s conveniently asks you if you want to save all changes made to all modiifed buffers
      • Extended commands
        • C-x is "Character extend", and is followed by a character
        • M-x is "Named command extend", and followed by a long name
    • whoa, apparently a vertical split with the large file causes interaction to slow down significantly? but not when it's in a single window in emacs, or even horiz split...
    • basically don't use vertical split when working with
    • libraries
    • python (not sure if properly workign? haven't tested in a while as of 2015-03-10)
  • emacs on ipad
  • emacs ESS doesn't work with the wickhamverse/tidyverse? then it's not usable by modern standards (use Rstudio instead apparently)
  • resources
  • org-mode notes
    • simple commands
      • tab cycles
      • C-c t cycles through TODO-DONE-blank for a correctly formatted star/headline item
      • C-c , <either case priority letter> gives an priority to that item, though it needs to be a TODO to show up on agenda
      • whoa, apparently the below two work not just on org headers, but also markdown-style unordered lists! however, they convert any ordered lists into unordered lists in this case
      • M-return basically gives you new headings at the 'current' level. It either
        • if it's not a header already, adds beginning stars to current line, turning it into heading of the 'most recent' order (top-to-bottom), or
        • if the line is already at that level heading AND the cursor is NOT in the first column, starts a new heading on the line below it that's ready for text input, moving all text to the right of the cursor (including under it) down to that new header. (This means a clean heading is created below if you're in Insert mode at the end of the above heading.)
        • if the line is already at that level heading AND the cursor IS in the first column, starts a new heading on the line ABOVE it that's ready for text input (so can then M-down to move it)
      • M-left/right left = moves header to a higher level (less stars), right = lowers heading level. (This works too en masse for Visual general selection, even taking care of non-header material correctly, but not Visual Block / vertical selection.)
      • M-down/up down = moves current header down the buffer amongst headers of the same level, keeping the cursor on it, and likewise for up; stops when reaching a header of a higher level.
    • install
      • there is a bug using ELPA packages etc. where you must not call ANY org functions before installing the FULL org package correctly - this means you need to comment out org settings set in ~/.emacs. It's dumb.
    • you have to install capture mode - see the manual - where you use C-c c
    • MobileOrg
      • Note that org/ is the "index file" for MobileOrg, so it knows the notes I care about, I guess I have to individually link to them?
      • You add JUST the paths to all the files you want on MobileOrg, not direct files links. Note that this is documented nowhere.
    • org-babel / literate-programming
    • html export images can have attributes modofied, but must be members of unordered lists
  • lrn-idea-vim-lrn-pycharm
  • play with vim




lrn-games (including finished ones I already own)







lrn-irc (lrn-irssi)



do NOT use hyphens in filenames, module names, etc.
vim and python: conclusion: just use Pycharm
  • omfg, started trying to get vim plugins all working, started working really well like with YouCompleteMe, but it turns out that getting the general packages ale's pylint version, youcompleteme install package versions, vim's internal python version (which affects what is used by certain, but not all, plugins), and ALL of their respective python venvs is...a hassle. For plugins that support it, if can have venv control then can get it to work, but not all plugins have sane venv control. Conclusion for the 431095015143th time: meh just use Pycharm ffs
  • some plugins support virtualenv, some don't and rely on the system python version vim is attached to, and almost none support conda!
  • Spacevim may be worth trying, but could be unstable like spacemacs

tldr; If you want "ggplot-style" plotting in Python, use plotnine. If you want simple plots that just work, use Seaborn, Altair, Bokeh, plot.y, D3...the list goes on. If you want to dive deep and invest time learning the "engine" powering most Python plotting tools (thus allowing you to plot anything) then learn Matplotlib.

I think I'm going to at least start out by trying to use "plotnine" for "ggplot2-style plotting". Supposedly it's the most similar to ggplot2 in R and the most comprehensive in terms of capability compared to other Python ggplot implementation attempts - here's a good blog post about the different contenders, showing them in action . There's also a good discussion here about learning/teaching ggplot-style for researchers vs a different plotting tool.

If you don't care about doing things the ggplot way, then there's also Seaborn and Altair (and others like Bokeh, that are built on top of Matplotlib . Seaborn and Altair are meant to be user-friendly tools for easy plotting of common stats/etc. data. Matplotlib is definitely the most widely supported and powerful plotting package in all of Python, BUT it can be a real pain in the butt to figure out how to do anything but the most simple plotting. I've heard Matplotlib described as, it works better as a plotting API on top of which to build simpler plotting tools (like Seaborn/Altair do) than a plotting library for general use. IMHO to do anything advanced in Matplotlib, you have to come to understand the underlying object structures for each part of the figure (figure vs axes, etc.), and the whole API is modeled after Matlab's, hence the name, so it's not exactly intuitive. This is the opposite of what most scientists want, so Matplotlib has an excellent Gallery showing the code for how to do different plots. You can totally get by in science just by building off of those examples and "cargo-cult" programming from that way. I will say, though, that if you do learn the Matplotlib API, then you will be able to make pretty much ANY plot that's possible, and that's something other libraries don't have. I made an attempt to learn Matplotlib for real, but abandoned it since the documentation was garbage for that and I couldn't find resources explaining how you're "supposed" to do Matplotlib; things are much better on that front today, though, ask me if you want links to guides on how to learn Matplotlib the RIGHT way

Regarding ggplot2-style libraries, there's 5 I'm aware of:

  1. "plotnine", which is under very active development (which is very good) and IMO the best candidate for, eventually, a full ggplot2 port in Python. It seems to be the most complete implementation in Python right now that doesn't resort to actually calling R's ggplot2 itself.
  2. "ggplot" as it's called on the Pypi package repository (aka the "pip" packages). This may work currently, but it is completely abandoned and hasn't had a single commit in 2 years. There's been a LOT of juicy drama with this repo. This was renamed to "ggpy" after Hadley Wickham, the creator of ggplot2 (and R programming god, creator of the "tidyverse"), complained about the name . However, while the repo has been renamed to ggpy, the Readme hasn't been updated with this new name, nor the website for this project ( , nor the Pypi package...basically they never got around to actually changing the name where it mattered. That alone is a big red flag that the creators no longer care about it (and thus it'll eventually stop working as Python advances). As the blog post at the beginning of this email explains, there's a lot of things in this package that are NOT implemented, even though it presents itself as a decently-full ggplot port. In other words, don't be surprised when only half of what you need works, from the blog post it sounds like pretty general stuff isn't implemented at all, and the syntax is different than plotnine/true ggplot2. Also note that because there's already a Pypi package (see below) named ggpy, they chose the new name without making sure they could distribute on the python package they'd need to change the name again if they cared enough to actually distribute it. This thing like super dead in the water to me.
  3. "ggpy", a different package from the one above that has the name reserved on Pypi BUT the Github repo it links to doesn't exist! Therefore the code doesn't exist.
  4. "pygg", which apparently wraps the ACTUAL ggplot2 library in R. Unfortunately, this means you have to transfer your data from Python to a similar, but not exactly the same, data type in R, every time you want to plot. Python plays well with the shell/system, though, so I would argue that if you were going to use this, it would be better to just generate/save your data in Python, then call an R script and do your plotting that way, independently. Jupyter notebooks are especially good for mixing languages like this, since the "r" in Jupyter stands for its support of the R language.
  5. "pyggplot" which is another wrapper around R's ggplot2. This has the same problems as pygg, but is also abandoned (2-3 years ago).
  • What seems to have happened: they changed the name and glamp pushed a big rewrite roughly a year ago, then basically all dev was completely halted, the pypi wasn't updated, and the readme wasn't either. And dev was apparently slow even before this


  • lxqt not ready for primetime as of 20180330
  • default ctags out of the box on Debian stable is NOT exuberant ctags and can NOT recurse by default, ugh
  • commands/programs for the terminal
  • conky system monitor
  • lrn-android
  • lrn-arch documentation (for macbook air at least) (much of this is out of date...and unnecessary if using a distro with a full DE)
    • backlight
      • echo 9 | sudo tee /sys/class/backlight/acpi_video0/brightness changes screen brightness!
        • max screen brightness for acpi_video0 is 15 - use this one as main
      • echo 5 | sudo tee /sys/class/backlight/intel_backlight/brightness is another screen brightness one, don't use
        • max screen brightness for intel_backlight is 1808, REALLY - don't bother using this one
    • copypasta / system clipboard
      • main thing: CLIPBOARD buffer is the one I use
      • emacs: omg, the headaches on arch, a surefire way / this worked once:
        • do NOT install emacs-nox (i.e. the one without X11 workings), instead install the normal emacs
        • not only that, but do NOT run emacs strictly inside the terminal, i.e. starting it with "emacs –no-window-system", as this prevents seemingly any X Primary/Clipboard clipboard buffer interaction
        • Install program xclip, as you do NOT need programs xorg-xclipboard or xsel for copypasta to work
          • you can test the Primary/Clipboard clipboard buffers via xclip -out -selection primary etc.
        • Enjoy - AFAIK, by Emacs >23, you don't even need to (setq x-select-enable-clipboard t) to make it copy to X11's Clipboard clipboard buffer. It should work at this point, but I haven't done it from scratch.
      • get arch copypasta working in vim and terminal emacs
        • notes 2014-08-21
          • so, with a new and improved omgclipboard, including now knowing the existence of the secondary and buffer-cut buffers
          • note that this is after I got the emacs clipboard working, which involved JUST running it in standalone-X-window style outside of the terminal
          • copying into buffers
            • emacs (nonterminal, that pops out) copies into the CLIPBOARD buffer
            • emacs terminal does not copy into any buffers! (it does use a program-private buffer though)
            • vim copies into the PRIMARY buffer
              • this includes, that if the vim session is exited, the PRIMARY buffer is emptied
              • this is with either the set clipboard=unnamed setting on, or NO explicit clipboard setting on in its .vimrc
              • hooray! with set clipboard=unnamedplus (see the plus), it ONLY uses the CLIPBOARD, not the PRIMARY as before from
            • firefox is complicated
              • firefox JUST selection of texts actually copies into the PRIMARY (though this usually happens when copying text)
              • ctrl+c or a "real copy" copies its contents into the CLIPBOARD
              • so, what ends up happening most of the time is it copies into both PRIMARY and CLIPBOARD
            • skype and vinagre right-click copies its contents into BOTH PRIMARY and CLIPBOARD
          • pasting from buffers
            • vim pastes from the PRIMARY buffer (e.g. from firefox' copying)
            • firefox pastes from the CLIPBOARD buffer
            • emacs, (nonterminal) under evil, pastes weirdness and apparently has an internal buffer, vim-style. Natively, though, it pastes from the CLIPBOARD buffer, as you would expect, since it only writes to the CLIPBOARD one.
            • emacs terminal also doesn't paste from any of the other buffers
            • middle mouse in tmux in urxvt pastes something??? just a hyphen? it doesn't appear to be any of the 4 buffers from xclip
              • more testing reveals it pastes a PORTION of the text in the PRIMARY
            • as a test of general desktop stuff, skype and vinagre pastes from the CLIPBOARD
          • after looking online, apparently many people complain about this with firefox, and it seems difficult to change
        • therefore, based on firefox's (and other desktop apps) pasting preference for CLIPBOARD, and the customizability of vim and emacs, the best thing to do is apparently center everything around the CLIPBOARD buffer, e.g.
          1. make vim copy/paste from it, and
          2. make terminal-emacs copy/paste from it
            • so, nonterminal emacs is fine, it doesn't even need (xclip-mode 1) in emacs settings to copy/paste to CLIPBOARD, it works out of the box
            • terminal emacs (i.e. with '-nw' passed) is more complicated, has two requirements:
              1. emacs needs the TERM environment variable to be set to 'xterm' or 'xterm-256color'
                • this can NOT be done automatically in emacs settings, apparently, even with the setenv function, since it still doesn't seem to work. Solution now is to add a 'TERM=xterm-256color' definition prior to all alias'd emacs bash aliases, so it's pretty much fixed
                • variants of screen and urxvt will not work! AFAIK this includes 'screen-256color', 'screen', 'rxvt-unicode-256color', 'rxvt-unicode'
                • note that this turns the emacs colors awful, the 'color-theme' package can rectify this
              2. emacs also needs the 'xclip' emacs package (you also need it installed on the system), and you have to turn it on with just (xclip-mode 1)
      • got copypasta working in urxvt tmux terminal!
    • file manager
      • can change nautilus settings via running dconf-editor and going to 'org->nautilus->' to hardcode the objects, or the thing in the upper left duh, or can use gnome-tweak-tool
      • can use locate via the 'mlocate' package, must first update with sudo updatedb - it's pretty fast
    • file system
      • filelight for visualizing hdd space
    • firefox kde? theme color change bug and
    • kernel modules
      • there are some details in the wireless note
      • the proper place to blacklist things, as per , is in a /etc/modprobe.d/<modprobe|blacklist>.conf file where you just put an entry for blacklist <module name>
      • lspci -vnn | grep -7 14e4 tells us what actual modules are being used by the hardware, in this case the wireless card in a macbook
      • you can see the status of some? modules via lsmod, and you can see where all the modules are loaded from via mkinitcpio -v | less
      • you can potentially individually load modules by making files in /etc/modules-load.d/<module>.conf where the contents is just the module name
    • keepass / passwords
      • had to install custom keepass portable to home dir for FF extension "KeeFox" – for deets, see its main wurbsite
      • update: i have no freaking clue where the above wurbsite actually is, it might actual be INSIDE the plugin (via html file)
      • Note: must install xsel package to get copying directly from KeePass, though still don't have paste TO KeePass
    • keyboard / keymap
      • ctrl-caps swapping? only diff is, as of 2014-12-10, I now use '/usr/share/kbd/keymaps/i386/qwerty/' where the Caps_Lock keycode is turned into Control vis-a-vis
      • note that this is an update that affects the terminal but NOT X11, which is still dominated by 10-evdev.conf, which now has "ctrl:nocaps" instead of constantly switching "ctrl:swapcaps"
      • now, if NEED capslock, just loadkeys .../ (or maybe can just do loadkeys us-capsctrlswap)
      • can set default keymap persistently via localectl set-keymap –no-convert us-nocapslock
        • might be able to do this as well with this, but untested
          export KEYMAP=us-nocapslock ```
      • passing alt+L/R arrow through virtual console: further edits to '.../' from here fix it (though NOTE that there is a typo in the answer, the second string is F106 not F105, simple and obvious typo)
    • installing arch
    • laptop power
    • lrn-dotfiles
    • maintenance
      • obviously, pay attention to your programs when pacman -Suy
      • pipelight is a special case, as it needs to be activated (see its entry under netflix) every time a new version is installed
      • haskell: run cabal update and cabal install world to update installed libs for haskell use
      • what is what is file "000033.log" in dropbox? how to search for that, find? is it emacs?
    • microphone
      • starting pulseaudio server and switching input in pavucontrol to analog seems to make mic work
    • mime / xdg-open types
      • installed 'mimeopen' package to be able to 'mimeopen -d file.pdf', then can set /usr/bin/okular to be the program, it works! with calibre and xdg-open and everything
      • use `mimeopen -d . to CLI-set which specific program you want to use for that filetype
      • xdg-mime query default application/pdf to query program
      • xdg-mime query filetype file.pdf to query filetype
      • some applications appear to force their Mimetypes in a way that conflicts with even when you set it
        • check out `grep 'application/pdf' /usr/share/applications/*' to see which ones reset
    • neovim
      • got everything working (minus powerline, which appears to be a fundamental issue with vim's bindeval) by installing from yaourt neovim-git and python2-neovim-git
    • netflix and az prime
      • how I got netflix working on FIREFOX:
      • just to be careful, make sure you close all browsers before updating:
      • write down that needed to activate multilib section of /etc/pacman.conf in order to install pipelight for netflix ANY way, since it was necessary for ANYTHING for wine, including most/all of it's lib32-x dependencies
        • this is done by uncommenting the [multilib] lines in /etc/pacman.conf
      • to make it easier, add pipelight to /etc/pacman.conf
      • just to be safe do "sudo pipelight-plugin –disable-all"
      • "sudo pipelight-plugin –enable silverlight"", then accept the licenses (SUDO is important here)
        • note that enabling the plugin and accepting the licenses are currently needed every update
      • also did "sudo pipelight-plugin –create-mozilla-plugins" but this may not be necessary
      • then get a user agent extension to firefox like the "UAControl" extension, giving it the user agent for "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:15.0) Gecko/20120427 Firefox/15.0a1"
        • "chromium –user-agent="derp" " doesn't seem to work anymore?
      • AZ prime
    • photo editing
    • peripherals
    • pictures program
      • use 'geeqie', dev'd fork of dead gqview?
    • python
    • screenshot
      • using imagemagick, which is probably already installed
        • run import -window root screenshot.jpg to save a screenshot to /home/USER
      • sometimes, opening the screenshots in certain image viewers, or the actual screenshot itself, will be screwed up, probably something to do with its headers – in that case, basically just run mogrify -density 90 screenshot.jpg from
    • skype
      • okay, so you're going to need to switch your audio to PulseAudio (which, AFAIK, is headphones/speaker-sensitive and just as good/capable as native sound stuff.)
      • packages needed to be installed (from pacman):
        • multilib/skype
        • extra/libcanberra-pulse (needed for pavucontrol)
        • extra/libpulse (necessary, this'll probably already be installed due to chromium etc.)
        • multilib/lib32-libpulse (specifically needed for skype?)
        • extra/pavucontrol (maybe unnecessary, but acts as a good alsamixer for PulseAudio INCLUDING the fact that it shows the amplitude of the ACTUAL SIGNAL either being output or input, so you can see if you're getting any mic signal)
        • extra/pulseaudio (needed for input processing and pavucontrol working correctly, aka it acts as a PulseAudio server?)
        • extra/pulseaudio-alsa (optional, equalizes alsamixer behavior with the PulseAudio channels)
      • another thing: as per Solution 1 at , add PULSE_LATENCYMSEC=60 to the beginning of the exec... line in /usr/bin/skype. This is needed for the horrible crackling quality on both input and output.
      • lastly, you MAY need to go into either pavucontrol? or at least alsamixer and increase the input capture and Digital channels to nonzero
      • after this, you may not need to restart X or the computer, but do it anyways just to be sure. It should work at this point.
      • note: after restarting the computer, you will need to reenable PulseAudio using the simple pulseaudio –start, after which you can use pavucontrol to control it
    • spotify
    • sound
      • make sure to switch to PulseAudio using pulseaudio –start on startup (see the skype section for install)
      • see skype section for how to setup pulseaudio
      • audacity doesn't like pulseaudio, use pulseaudio –kill
    • steam
    • system, incl X11 settings, package configs (including journal for manual changes to system)
      • getting started
        • (unnecessary to actually have an xorg.conf in arch, but it's way better to have one)
        • make sure you install the xf86-video-intel or whatever it is, that's not explicitly mentioned in the arch Beg Guide
        • generate a skeleton via running Xorg :0 -configure
        • then sudo copy the new file /root/ to /etc/X11/xorg.conf
          1. not necessarily necessary, but do it
        • start it with startx, which uses ~/.xinitrc
      • fonts
        • (as in, in xterm in X11)
        • add the fonts to /usr/share/fonts/<folder for whichever filetype they are, like TTF>
          1. after this, they should be see/grep-able using fc-list, just to make sure
        • in .Xdefaults or whichever your WM is using, add
          XTerm*faceSize: 10 ```
      • important files for X:
        • ~/.xinitrc (where you choose your wm)
        • ~/.Xdefaults
        • /etc/X11/xorg.conf
        • also see files listed in $lib/bash/rsync_utils/maint/
      • whoa xdg-open can be used on terminal or even dmenu with arguments to open in native app
        • you use xdg-mime to edit default applications
        • to open pdfs in firefox, for example, (see xdg-mime –manual), use xdg-mime default firefox.desktop application/pdf
        • can maybe edit ~/.local/share/applications/mimeapps.list ? not sure
        • can use mimeopen -d <file> via xdg-open for interative mime default app selection on CLI! see
      • dmesg gets like the kernel cycle buffer or something? what even is that?
      • changing keyboard settings using xkboptions a la X11 (archaic)
        • can change any of these to default via setxkbmap -option by itself
        • swap capslock and control, like for laptop keyboard, with just setxkbmap -option "ctrl:swapcaps"
        • swap alt and windows keys, like for home keyboard with just setxkbmap -option "altwin:swap_altwin"
        • so, you may have to start by additionally installing the evdev general input (incl. keyboard and mouse) driver, idk
        • once done, enabling caps/ctrl swapping on startup is as easy as expressing an XkbOption in /etc/X11/xorg.conf.d/10-evdev.conf like
          catchall" MatchIsKeyboard "on"
          Option "XkbOptions" "ctrl:swapcaps" Driver "evdev"
          EndSection ```
        • as of 2014-12-10, have just started using "ctrl:nocaps" instead of messing with swapcaps for X11, and made new keymaps to fix both caps/ctrl switch and alt+arrow (alt+L/R arrows) console switching in non-X11 console via:
      • swapfile see
      • 2014-12-03: so apparently running "systemctl mask systemd-udev-settle" will disable that thing which was slowing down boot
        • this worked! briefly reading forums say that this settling thing isn't really used or at least is obsolete
      • visual framework settings (gtk, etc.)
        • qt, affecting konquerer, dolphin, vlc: run qtconfig-qt4
        • some others? gtk-chtheme
      • apparently config files for lots of installed pacman packages are in "/usr/share//stuff.conf"
      • CLI X11 display scripting DISPLAY=:0.0 .screenlayout/
      • xprop run through terminal allows you to click on a program running in X11 and see its properties?
      • use xev to find keycodes
    • lrn-thinkpad-specific
    • time
      • change timezone by timedatectl set-timezone america/new_york or america/chicago etc.
    • trackpad
      • dl "mtrack" driver
      • add either the section or just the preferences here to /etc/x11/xorg.conf: ### aes: some of these settings are new/nondefault, and this entire section might be as well section "inputclass" matchistouchpad "on" identifier "touchpads" option "thumbsize" "50" option "scrolldistance" "80" option "sensitivity" "0.5" option "fingerhigh" "9" driver "mtrack" endsection
      • as of 2014_0730, new update to xorg-server 1.16, having that stuff in the xorg.conf was sufficient to completely disable the trackpad!
        • however, deleting then reinstalling the mtrack driver through yaourt and adding in 10-mtrack.conf to xorg.conf.d didn't have any effects
      • so adding these settings into 10-mtrack.conf makes it seem like they're working!
      • ugh so on 2014-08-21, seemingly upgraded xorg, andt disabled touchpad, though it was an easy fix since /etc/x11/xorg.conf.d/50-synaptics.conf suddenly was demoted to just a backup with .pacsave. by copying the pacsave to the file's name before, it now works! ugh, i don't want to have to do this every time!
    • urxvt
      • transparency in xmonad through compositing
        • enable composite in /etc/x11/xorg.conf through section "extensions" option "composite" "enable" endsection
        • dl prog xcompmgr, and in your ~/.xinitrc, put (maybe needs to be before exec monad) xcompmgr -c &
        • if want background picture, program feh seems to work better than program xloadimage
    • vnc
      • apparently i'm using vncviewer which is really aur/tigervnc-viewer but may come with default vnc package?
    • webcam
      • package 'isight-firmware-tools' or something like it from aur, and then reboot of computer, seems to fix nonfunctioning webcam on air
    • window managers
    • wireless (broadcom 43224)
      • NetworkManager
        • to save wifi passwords, go to 'nm-connection-editor', right click on people next to password and "store password for all users"
      • bu / general wpa wireless support
        • use wpa_supplicant.conf example in z wpa_supplicant entry in the db
          (802.1x)" priority=15 key_mgmt=wpa-eap
          eap=peap mschapv2 identity="user" password="derp"
          phase2="auth=mschapv2" } ```
      • autoswitch
      • troubleshooting wireless network
        • first set up the interface with sudo ip link set wlp2s0 up
        • then do the handshake via sudo wpa_supplicant -b -dwext -iwlp2s0 -c/etc/wpa_supplicant/wpa_supplicant.conf
        • then get a dhcp lease via sudo dhcpcd wlp2s0
        • finally, ping to test ping -c 2
      • oh my god, don't even get me started.
      • as of 2014_0601, now i need to load module brcmsmac with modprobe brcmsmac after booting in order to load the wireless interface, after which everything is peachy
      • as of 2014_0619, everything is back in order, the normal interface wlp2s0 loads on boot, and it works.
        • the issue was, while something did go wrong a while back, i def made things worse by blacklisting pretty much all available broadcom wireless drivers in /usr/lib/modprobe/broadcom-wl.conf
        • the proper place to blacklist things, as per , is in a /etc/modprobe.d/<modprobe|blacklist>.conf file where you just put an entry for blacklist <module name>
        • as per looking at the actual module being used by the hardware, lspci -vnn | grep -7 14e4 tells us that now, the bcma module is the one being used
          1. this is good, since this appears to be the default one, as it's one of the few that are already loaded in the kernel.
        • so in other words, if it breaks again, do another lspci... to see if bcma is not being used, etc. note that brcmsmac relies on bcma, which apparently can still be loaded if blacklisted if you do sudo modprobe brcmsmac
        • you can see the status of some? modules via lsmod, and you can see where all the modules are loaded from via mkinitcpio -v | less
      • as of 2014_0729, there was a kernel upgrade and now broadcom-wl module doesn't appear to work, so had to modprobe brcmsmac to get working again. since wl automatically blacklists brcmsmac, i uninstalled wl from yaourt, and upon reboot brcmsmac starts automatically and works. so will delay getting broadcom-wl for a while. wow i actually knew what to do at this point.
      • seems you need to manually reenable/etc netctl profiles via sudo netctl enable/start/reenable wlp2s0-derp before you can enable them in netctl-auto
  • lrn-debian documentation
    • battery threshold control (thinkpad)
    • bluetooth woes
      • may have to install 'pulseaudio-module-bluetooth'
      • if no audio coming through at all, completely forget device from computer, then pair as if for the first time
      • then, should have crappy audio coming throuhg - go to 'System Settings > Audio and Video > Audio Hardware Setup' and change the device's/"card"'s profile to A2DP instead of the crappy headset one
      • to get volume control, go to pavucontrol > Playback, and for the main source switch the hardware on the right
      • repeat ad infinitum when it inevitably stops working
      • a different attempt, without forgetting
        • restarted bluetooth service on laptop
        • waited a minute
        • turned on headphones (after messing around a bunch before)
    • mime / xdg-open types
      • use `mimeopen -d . to cli-set which specific program you want to use for that filetype
    • networkmanager wifi
      • nmcli - one cli client built in
        • commands
          1. to show list of wifi networks do nmcli dev wifi list
          2. to add a network, do nmcli dev wifi connect <ssid>
          3. to connect if already have a connection, can do nmcli con up id <connection id
        • notes
          1. i guess it treats wifi as a generic device?
    • r
    • troubleshooting X11 screen EXTREME flickering (in both gnome and kde): according to the combination of the options from but in the file placement from , need to put the below in either the file '/usr/share/X11/xorg.conf.d/20-intel.conf' or '/etc/X11/xorg.conf.d/20-intel.conf' Section "Device" Identifier "Intel Graphics" Driver "intel" Option "AccelMethod" "uxa" EndSection
    • troubleshooting minor window jitter when moving windows or changing sizes
      • issue seems to be caused only when starting laptop connected to external monitor AND laptop screen shut closed.
        • Re-opening the laptop screen makes the problem go away partially, but
        • rebooting laptop with clamshell screen open AND connected to external monitor seems to fix problem
  • lrn-kernel
  • lrn-make and other tools like autotools
    • what does make -j 4 do really?
    • Introduction to the Autotools (autoconf, automake, and libtool)
    • gnome-screensaver is abandoned
  • lrn-mpd
    • on debian stable, regular install installs to systemd service, but that uses the GLOBAL config in /etc
      • so, need to disable the systemd service and run user autostart for MPD to read user config
      • this works for both ncmp and cantata
      • can then set global media controls in KDE to control cantata, but only if it's open
  • lrn-wayland
  • lrn-wifi
    • rfkill - once installed, can use sudo rfkill list all to see if wifi blocked
    • can then use sudo rfkill unblock all to unblock it
  • lrn-windows (ugh)
  • magic sysrq key
  • awk sed
  • filesystem
  • security
  • tutorials
  • terminal
  • chroot



  • archaic nginx notes
    • old Kewl filez on it: /var/log/nginx/error.log /var/log/nginx/access.log /etc/nginx/sites-enabled/www /etc/init.d/nginx
    • HTML locations:
      • On Mac:
        • /opt/local/share/nginx/html
      • On Linux: (probably)
        • /usr/share/nginx/html
    • config locations:
      • for nginx.conf, mime.types:
        • On Mac:
          • /opt/local/etc/nginx
        • On Linux: (probably)
          • *etc/nginx*
    • Instructions from
    and at the end of the install, you'll be shown a command that causes
    it to run at system start: sudo port load nginx.
    Configure nginx. Start by setting the default config files:
    cd /opt/local/etc/nginx sudo cp nginx.conf.default nginx.conf sudo
    cp mime.types.default mime.types sudo nginx -s reload Load
    <http://localhost> in your browser to see that it's working. You're
    now done installing nginx. ```

lrn-open-science lrn-reproducibility

lrn-organizational software




lrn-regex (regular expressions)



  • add simulation filesystem org to coding practices (use as an exercise to revisit it)
    • see what solutions are already published, poll people, etc.
  • read misha's good practices wiki for tips






  • learn xpath, apparently a good way to go through xml
  • emacs also has nxml-mode, which is supposed to be good
  • stackoverflow says, if you really need industrial usage of XML like in a corporate environment, open tools just don't cut it

lrn-cycling :lrn_cycling:

road bikes to check out

  • peugeout, fuji, panasonic (sp) old?, raleigh is okay but not for speed bikes, bianci, trek, vilano shimano?
  • vilano shimano looks good
  • Hub Bicycle shop woman named emily says $6-700 is a good price for a good road bike


core principles for design could just be sufficiently complicated that it's difficult to think of good counterexamples

see kim's list of good websites

go through hack design website

crap way to improve usability

proposal for an interactive introduction to graphics programming

minimalist websites





lrn-economics-lrn-finance :lrn_finance:




  • lrn-bb-brian's recommendations
    • basically, vanguard roth IRA into the VFIFX fund – actually maybe VFFVX to target 2055 retirement
    • brian-recommended financial metrics
    • bstout brian recommends 'futureadvisor'
    • bb finance says $10,000 into roth ira through vanguard
      • says start in vfifx (3k min)
      • once 12k, split into indiv mutual funds
      • now is good time to buy 20150911 esp intl
      • definitely don't put in more than max, or withdraw (obvi)
      • if more, "invest in a vanguard taxable accoutn", "something like vsivx, since doesn't pay much in dividends, and is therefore tax efficient" - "just builds equity which you don't get taxed on until later, and laxed at lt cap gains, which is much more favourable"
    • more
      • firstly, find low fee mutual fund company in planning on 2040/50 retirement via roth ira and roth 401k
      • max out roth ira (or roth 401k if so chosen) ($5500/yr) first, super important in 20s
      • then do vanguard "small cap value fund" and "total stock market index fund" ticket visvx, vtsax
      • then open tax deferred life insurance policy
      • then don't take out until not in high income bracket
      • after that, it's up to you, including "vanguard/peer-to-peer lending/equity"
    • 20160125 convo Re advisors: vanguard and fidelity is fine, but only if you have a balance large enough to get free advice, no commission (salaried), and recommend only vanguard funds which are cheap Hmm,, but I use roboinvesting software to optimize portfolio weights Future Adviser Analysizes finances in all current online accounts, balances based on goals, and provides top 3 inexpensive and close beta index funds that meet target Pretty easy, and totally free (if you act on advice yourself rather than them buying/selling)
  • No 401(k) at work, do I have investment options besides Roth IRA?
  • software



ask salim about that as an intro book, or maybe ap econ?

revenue maybe prop to "buying power", which inv prop to rate (a la actuary guy), for which of course premium is calc from

tightwads and purchasing pain as prop to income and expenses. how to get data on what these are for most people? could match with good accuracy to address alone?

read law stuff on the shell-eni nigerian. malabu deal, in the british high court and another in arbitration?

global witness ngo, tom wayne, maybe discusses docs

keep read mankiw (while questioning assumptions, or at least tracking them)

best books to learn econ

side money


(1 hr) read signals and systems section 8.7 for fm and instant freq for fun

arduino/raspberry pi

lrn-signal-processing (lrn-sig-proc)


lrn-healthy-eating (eating comes before exercise in importance!)

  • omg make frozen burritos for lunches? aka other frozen food on sunday :@home:
  • whole 30
  • notes from gohar, having done both paleo for years and Whole 30
    • Gohar says whole 30 , plan all meals, must go to market or whatever on Sunday - non negotiable. Also she had 2 cheat meals per week
    • Nom nom Paleo site
    • Put stuff like meat in individual packages in freezer
    • Roast big sheets of vegetables
    • Gohar says work lunch, do frozen containers of chili and soup etc ready to go in freezer
    • Pre prep on Sunday even if not cooking on that day
  • gohar cabbage salad
    • dressing option 1. honey, apple cider vinegar, olive oil.
    • dressing option 2. rice vinegar, mirin, and tiny bits of olive oil and sesame oil
    • main thing: "mandolin" a head of cabbage, roast and toss in some seeds, add big handful of chopped mint, add a wee bit of mandolined shallots or red onions
    • can add roasted chicken

class notes

  • 20140224-adult education cajun cooking class
    • cajun roux
      • darker so more flavor and color, less when formal
      • used as a french thickening agent
      • low heat!
    • jambalaya - rice dish with stuff, while gumbo is a stew w/ rice in it
      • "red" jambalaya - tomato
      • uncooked rice (he did 3 cups to 5c stock and large tomato sauce)
      • you can bake jambalaya
    • shrimp remoulade - "remolaude" means condiment
    • cajun "trinity" - celery, green chilis, pepper
      • "holy trinity" - trinity + garlic
    • okra - word originally "okigumbo" but now gumbo
      • use it in gumbo!
    • his greatgrandma used tomatoes in gumbo
    • if you "move around" the chx, it'll break apart, so make the call
    • Andouille sausage
      • can you get good stuff even at supermarket now, since it's a thing
    • diagonal cuts of jalapeno are quicker
    • poaching liquid - lots of lemon, celery, white wine
    • 15 mins on softening his vegs - until the color starts to dull
    • add lemon to remoulade stuff (and many other things)
    • "crystal" hot sauce, less vinegary
    • poach just a few shrimp at a time, since some will chamge temp too much to be effectual
      • don't want rubber overcooked, get "just at or undercooked"
      • look at the head of the spine to check if center underdone
      • shock the shrimp
    • cajuns use animal fat in roux, Creole used butter and were richer, it's more traditionally French, simpler (less ingredients), more finely differentiated
    • hoppier beer: drink it sooner
    • incr alc = it ages better / you can wait on it
    • Ommegang, Allagash, and Cambr Brew Comp have "wild yeast" beers
    • CBC is isolating Cambridge "urban" natural yeast
    • "GumboTales" book

Get good at drinking!

  • try montenegro digestif
  • william recommends Amaro Dell'erborista, said it was amazing
  • buy ricardo, amaro montenegro @liquor store :buy:
  • beer: kevin wang wants Trillium and Tree House beers (made in boston), says they're the best on
  • rsh more on very old fashioneds
  • Zephyr cocktail at Hawthorne, made by Jason, very good
  • mojitos
  • good mezcal: jolgorio tobala
  • 1919
  • Bourbon & ginger
    • either Talisker 18 and ginger ale, or
      • ginger syrup via
        • 1/2 cup sugar (fine turbinado)
        • 1/2 cup water
        • 1/4 lb. fresh ginger root (peeled and finely grated)
        • In a small saucepot bring the sugar, water, and ginger to a boil. Turn off the heat, and allow to steep and cool for one hour. Strain off the ginger pieces, and reserve the syrup.
      • mixed via
        • Large ice cubes
        • 3 oz. Maker's Mark (or your favorite bourbon)
        • 1 tbsp. ginger syrup
        • 2 slices fresh ripe peach (Try pear if peaches are out of season.)
        • 2 oz. soda water
      • Fill a 10 oz. highball glass with ice. Add bourbon and ginger syrup, and stir. Add peach slices (or pear), and pour soda water to the top of the glass. Stir gently.
  • Prof moriarty fernet drink at Peter kern lib
  • beers to try
    • headytopper
    • Great lakes brewing co, commodore perry IPA
  • wines to try
    • Sokol Blosser - Evolution
  • ward eight recipes
    • old recipe:
      • 0.5 oz lemon juice
      • 0.5 oz orange juice
      • 2 oz rye whiskey
      • 4 dashes grenadine
      • shake, serve with cracked ice
    • from a latin quarter ? nightclub in 1950s:
      • 1.5 oz rye
      • juice from half a lemon
      • 1 oz orange juice
      • 1 tsp grenadine
      • shake and strain, garnish with maraschino
    • 1934 G Selmer Fougner for the New York Sun
      • juice from one lemon
      • 1 "barspoon" of powdered sugar
      • put above in large whiskey glass 3/4 full of bourbon
      • add ice
      • 3/4 dashes orange bitters
      • one-half "jigger" of grenadine
      • either bruised mint or creme de menthe
      • top off with water or seltzer
      • add two half slices of orange, a piece of pineapple and a cherry
    • 1936 Mr Boston advertisement
      • 1.5 oz Old Mr Boston Rye ""Liqueur"
      • 1 oz lemon juice
      • 0.5 oz orange juice
      • 0.5 oz grenadine
    • modern Ward Eight from Drink
      • 3 oz Old Overholt Rye
      • 1 oz lemon juice
      • 0.5 oz simple syrup
      • 4 dash Ango orange bitters
      • 0.5 oz grenadine
      • top with soda water
      • muddle mint into a glass and then remove the mint
  • south boston cocktail, from Mr Boston official bartender's guide - 1 part Wire Works American Gin - 1 part apple brandy - 0.5 part lemon juice - 0.25 grenadine or pomegranate syrup
  • the lone tree from Hotel Touraine, Boston - 1 part american gin - 1 part sweet vermouth - 1 dash orange bitters - strip of orange peel
  • Fanciulli is a Manhattan but Fernet Branca instead of Angostura
  • Dillionaire
  • Fernet vallet Mexican fernet?

get good at cooking!

coffee and tea

  • chris johnson says peruvian from Harvest near babe's is good

already made cooking recipes that are good

lrn-grooming :lrn_grooming:

lrn-hair :lrn_hair:

  • if need pomade, Murray's (sp) or Imperial (water-based) are supposedly good


ergonomic chair stuff


sony WH-1000xM2 noise-canceling headphones




  • Riley Bogema's list of books for chinese history
    • The open empire – Valerie hanson
    • The Chinese frontier – 221bc – 1857?
    • The perilous frontier – Thomas barfield
    • Ancient china and its enemies – Nicola icosmo
    • The rise of the east empire
    • Mao’s biography
    • 1845-1864 – the largest uprising in human history
    • God’s Chinese son – jonathan Spence
    • The unknown story of mao – jung chang, john halliday
    • Chinese civilization – a source book — patricia ebrey
    • Cheng-du – “the best city in china” – travel hub


digitizing paper records from home


dod's learning resources

Katrina says Pimsleur language learning courses are great


  • french duolingo
  • rosetta stone



How to study mathematics

visualization of math for pedagogy

Ten lessons I wish I had learned before teaching differential equations" [pdf]

peak finding






  • read ch4 of dayan and abbott for info theory
  • read the original shannon paper




  • best way to learn data analysis/stats probably do it, including NOT just reading books, but REAL problems or active HOMEWORK "You'll learn faster by doing problems than reading material"
  • help
  • Recommended Self-Study Path for Statistics | Hacker News
  • Probability & Statistics |
  • print / write down p value yes/no definitions on wall :@office:
  • go through
  • dl book: "statistics for spatiotemporal data" - rec'd by lepage?
  • always? list assumptions for stat rsh analysis
  • program mutual information, be able to explain it
  • Stats book: ch5 initially, "random sample X1,...,Xn is a SINGLE kind of RV and n observations" "a sample size is the number of observations"
  • how not to do it
  • books
    • probability: reddit and HN seem to agree that Feller's Introduction to Probability Theory is a very good classic
    • Casella and Berger's Statistical Inference is then a possible next step?
    • HN's HilbertSpace comment
    • Early in my career, I asked that question. Eventually I got some answers. Broadly usually the 'big bucks' are in things that are new at least in some important sense. To find such things, one recommended approach is to stand "on the shoulders of giants". Then you may still do some original work, but you have one heck of a base. My view is that Moore's law, the Internet, etc. are not nearly fully exploited, that enormous value remains, and the best way to get some of this value is to find a big unsolved problem where a good solution will be very valuable, stand on the shoulders of some giants, do some original work to get some powerful 'secret sauce' for solving the problem (which appears not to be easy to solve otherwise), start a business, please customers, get revenue, grow the business and sell it. Here, however, just what problem, giants, original work, secret sauce are NOT easy to see. Broadly we are trying to see, no, MAKE, the future, and usually that is not easy. While we can see what applications have been made of statistics in the past, what statistics will be applied where in the future is not easy to see. Yes, broadly statistics is highly promising. Or, in 'information technology' we want valuable, new information, are awash in cycles, bytes, bandwidth, infrastructure software, and data, and want to use these to create the information. Here, on appropriate problems, statistics can commonly, easily, totally 'blow away' anything else. The marriage between (A) 'information technology' in business and (B) statistics has yet to be consummated, even to reach to 'going steady' stage, assuming the traditional order of events. For 100 years or so, people have come to statistics from various areas of work. Usually they had some data and some questions. Some of the areas have been educational and psychological testing, experiments and testing in agriculture and medicine, industrial quality control, model building in economics, experimental work in physics and chemistry, investing, attempts to create mathematically based sciences in the social sciences, especially economics, psychology, sociology, and political science. In sociology, old examples would be James Coleman, Pete Rossi (professors of my wife when she got her Ph.D.), Leo Goodman. Likely the best fast, practical path into statistics is via books, courses, etc. intended for students in the social sciences. These students commonly do not have good backgrounds in mathematics. For the mathematical prerequisites, generally can get by, as a start, with just high school first year algebra. With this path can learn about probability distributions, the central limit theorem, the law of large numbers, statistical estimation and confidence intervals, hypothesis testing, cross tabulation, analysis of variance, regression analysis, principle components analysis, and more. For more, statistics is a rock solid part of mathematics, as solid as any part of pure mathematics, e.g., topology, geometry, analysis, algebra, etc. That is, statistics is based solidly on theorems and proofs, sometimes relatively deep ones. Statistics as theorems and proofs is called 'mathematical statistics'. Long standard has been (with TeX markup): Alexander M. Mood, Franklin A. Graybill, and Duane C. Boas, {Introduction to the Theory of Statistics, Third Edition,} McGraw-Hill, New York. The main prerequisite for this book is just a not very good course in calculus, and the book actually makes not much use of calculus. Mostly all a student will need from calculus is that it can find the area under a curve. Since the book has long been standard, can't really ignore it, but it's ugly. And often, with just calculus, the book doesn't really give solid proofs of the results. E.g., their treatment of sufficient statistics has some nice intuition, but their proof is junk. The subject cries out for a good book, but I'm not trying to write one or waiting for someone else to. Can get some of the flavor of mathematical statistics done with high quality, as mathematics, in, say (with TeX markup): Jean-Ren'e Barra, {Mathematical Basis of Statistics,} ISBN 0-12-079240-0, Academic Press, New York. Robert J. Serfling, {Approximation Theorems of Mathematical Statistics,} ISBN 0-471-02403-1, John Wiley and Sons, New York. P. Billingsley, {Convergence of Probability Measures, 2raise0.5exhbox{ed},} ISBN: 0-471-19745-9, John Wiley, New York. R. S. Lipster and A. N. Shiryayev, {Statistics of Random Processes I, II,} ISBN 0-387-90226-0, Springer-Verlag, New York. However, pursued mathematically, statistics has some relatively advanced prerequisites some of which curiously are not popular in US university mathematics departments. For the prerequisites, High School. Should have had high school first and second year algebra (reasonable facility with algebraic manipulations, the binomial theorem, complex numbers, both of which will see again in important ways), plane geometry (where nearly all the work was proofs – first place to learn about proofs), trigonometry (usually assumed in calculus and important in, say, analysis of organ tone harmonics and, thus, the most important example of an infinite dimensional Hilbert space), analytic geometry (especially the conic sections, especially ellipses which definitely will see again), and, if can, solid geometry (for more intuition in three dimensions). College. Need a standard calculus course, not necessarily a very comprehensive or difficult one because will do the subject all over again, maybe two or three times, and more later, WITH the proofs! Then need linear algebra, that is, how to work with data of several dimensions, which is just crucial. The big result is the polar decomposition, and there get to think about ellipses and get to use complex numbers. Also the course is an introduction to functional analysis and Hilbert space. Use any popular book to get started but in the end cover the classic, Halmos, "Finite Dimensional Vector Spaces'. Halmos wrote this when he was an assistant to von Neumann and intended it to be a finite dimensional introduction to Hilbert space (which once von Neumann had to explain to Hilbert) which it is. It also has some multi-linear algebra, of interest to exterior algebra now popular in relativity, but likely for nearly any business applications of statistics for the next several decades can skip that chapter. Then need some advanced calculus. That is a poorly organized, huge, catch-all subject beyond any one course. The usual start is 'baby' Rudin, 'Principles of Mathematical Analysis'. So, that's calculus with the proofs. Warning: The book is severe, succinct, with zero pictures. Have to draw your own pictures in your head. The book is packed solidly with powerful material just awash in important applications from statistics, economics, and engineering to physics, but there is hardly a hint of the applications in the book. I enjoyed the book, but few people will enjoy it or even get through it. Hint: Get a really good teacher! Then for more, popular is Spivak, 'Calculus on Manifolds', mostly because it is short. Actually, it's too short. I prefer Fleming, 'Functions of Several Variables' until get to the exterior algebra chapter at which time, if care, can now get the thin Henri Cartan, {Differential Forms,} ISBN 0-486-45010-4, Dover, Mineola, NY, 2006. in English. Between Halmos, baby Rudin, and Spivak, you will have covered Harvard's Math 55 with a colorful description at Harvard tries to cover these three for freshman, but in most math departments the material will take you through all or nearly all of a focused undergraduate pure math major. If somewhere take a course in abstract algebra, e.g., with a little group theory, then that might help! Graduate School. Might learn a little more about topology, say, from Simmons, 'Introduction to Topology and Modern Analysis'. So, get good with metric spaces and get started on duality. The next big step is a course in measure theory and functional analysis. The Simmons work will help. Baby Rudin will be crucial; Halmos is recommended. So, with measure theory, do calculus over again and in a very different and much more powerful way and a way just crucial, even central, for mathematical approaches to statistics. The functional analysis will concentrate on representation theorems, the Radon-Nikodym theorem, and Hilbert and Banach spaces. Long popular, from Stanford and sometimes aimed at statistics students, is Royden, 'Real Analysis'. It's gorgeous. Should also read the real half of Rudin, 'Real and Complex Analysis'; it's a few steps up in difficulty from baby Rudin. Again, hint: Get a good course from a good teacher who can get you over the material without getting stuck. Then go back and study the material 2-3 more times, apply it, do some original research in it, and finally begin to understand it. We're talking high, top, center crown jewels of civilization here; the stuff is of just awesome power; my view is that it is, for the rest of this century, one of the main pillars of increases in economic productivity via the exploitation of Moore's law; on 'what to program', computer science is stuck and this material is the most promising way forward; as a famous restaurant owner once said about some Morey St. Denis, "you won't find better". Now are ready for probability. I recommend: Leo Breiman, {Probability,} ISBN 0-89871-296-3, SIAM, Philadelphia. M. Lo`eve, {Probability Theory, I and II, 4th Edition,} Springer-Verlag, New York. Kai Lai Chung, {A Course in Probability Theory, Second Edition,} ISBN 0-12-174650-X, Academic Press, New York. Jacques Neveu, {Mathematical Foundations of the Calculus of Probability,} Holden-Day, San Francisco. Neveu is succinct, gorgeous, but not easy. This material is NOT popular in US departments of mathematics. At Princeton, see Cinlar. Then should make some progress with stochastic processes: The big book is Gihman and and Skorohod, right, in three volumes, but mostly people settle for shorter treatments. Whatever, should learn about Poisson processes, Markov processes (discrete time, finite state space is enough to get started), Brownian motion, and martingales. Might also learn about second order stationary processes. A good course in stochastic processes is NOT easy to find, especially in mathematics departments. Now are ready to attack statistics mathematically! I don't know of a good, single 'mathematical statistics' book at this level. Instead, there are many books – I gave some above – and then the journals. Thankfully, the field is relatively close to applications; so can take a practical problem and concentrate on what is relevant to it. One of my papers was some new work, at this level, in mathematical statistics for a problem in practical computing and computer science. The computer science community struggled terribly with the mathematics. So, it was some progress in computer science that community will have to struggle to understand. One approach to work in computing is just to try things, that is, just to throw things against the wall and see if they appear to stick. Or, maybe the truth of the situation really is a simple statistical model. Likely that model will fit the data well. So, try many simple statistical models. We use these models mostly ignoring the mathematical assumptions; mostly we are proceeding 'heuristically', that is, with guessing. If any of the models fit well, then they can be considered candidates for the truth. So, are throwing things against the wall to see if they fit. This approach also called 'data mining'. Problems: (1) Will be quite limited in what statistical models can use. That is, will be drawing from a cookbook instead of being a real chef who can create good, new dishes appropriate for the available ingredients and customers! (2) Don't have much, e.g., have not proceeded mathematically where from the deductive logic of assumptions and proofs actually know in advance some good things about the results. Something like breaking into a pharmacy, mixing up a lot of pills, taking them, and seeing if feel better! Uh, I'll pass and let you do that without me! (3) May have gone through a lot of computer time in an 'exponential, combinatorial explosion' of efforts throwing against the wall. (4) Have ignored a LOT in statistics that can add to what know about the results. (5) Will be tempted to conclude have have found 'causality' but will likely not have. (6) Will be tempted to conclude that have a model that predicts, but that is on shaky ground and risky and needs more work. Applied to important problems, this approach can be dangerous. There are not many healthy statistics departments. Much of the career interest is in biostatistics, especially related to FDA rules. It appears that among the top statistics departments are Berkeley and UNC. Since Breiman and Brillinger are at Berkeley and since Stanford, long good in statistics, is not far away, if I were looking for a Ph.D. in statistics then I'd pick Berkeley. There is a general problem getting a 'job' in a technical field and likely also with statistics. The assumption in US business is still as in factories 150 years ago: The supervisor knows more than the subordinate; the subordinate is supposed just to add common labor to what the supervisor says. In particular job descriptions are written by the supervisors, not the subordinates! Well, there are nearly no supervisors in US business who have even as much as a weak little hollow hint of a tiny clue about the material described here. So, won't need that material to qualify for the job descriptions. Moreover, if actually know such material and let that fact leak out, then will likely not make it past the first level HR English major phone screen person who will tremble and conclude that you are not like the employees they have! If you do get hired and someone in your management chain discovers that you have used some mathematics they don't understand, you might be on the way out the door, especially is your work was valuable for the company! Of course, the solution is to find a valuable application and start your own business. While maybe biomedical venture capital can understand crucial, core technical content, in information technology venture capital, likely you will be trying to explain this stuff to, say, history majors who worked in fund raising, marketing, general management, or financial analysis or have a background in just relatively elementary parts of computing. Just will NOT find more than six people, maybe not more than zero people, in US venture capital who can work the exercises in Royden or explain the strong law of large numbers. Sorry 'bout that! So, if you explain that the value of your venture is the powerful material in your 'secret sauce', then you will be regarded as a kook, far outside the mainstream of venture funded entrepreneurs, discarded, maybe even laughed at. As it is, some of the venture people are making money now, and the rest just want to be more like the ones who are making money. Looking for anything really new, powerful, and valuable is just NOT in the picture. So, once you have some results in users, customers, revenue, etc., then maybe you can get some venture funding; just why at that point, owning 100% of the business, you would take venture funding, a Board that can fire you, etc. is less clear! Or venture funding is not for everyone! Or venture firms prefer to give money to people who don't need it! For the real power of the 'secret sauce', you have just to keep that a secret! Once mathematicians have yachts, at the venture firms math will be to info tech like biochem is to biotech. In the meanwhile note that a valuable application of statistics can put you on the Forbes 400 where there are not many people! Generally if you are making a valuable application of advanced or new statistics, then you will not know many people who understand what you are doing. Or, if lots of people understood it, then it wouldn't be valuable!
    • one of, if not the best recommended books on Bayesian? "jaynes' Prob theory: the logic of sci"
  • p-values
    • friends don't let friends calculate p-values
    • table of alpha and beta errors in experiments, from table 6.2 of matlab for neuroscientists
      • ** effect exists (h0 false) effect doesn't exist (h0 true)**
      • we conclude it exists discovery of effect alpha (false rejection of h0)
      • we conclude it doesn't beta (false retention of f0) failure to reject the null hypothesis
    • if think understand them, go through
  • arima
  • pca
  • anova
  • misleading statistics
  • lrn-bayes
  • lrn-machine-learning-artificial-intelligence
  • good books for data science etc
  • probaility theory for scientists and engineers


Nanodoc software


lrn-NS lrn-neuroscience

what is known about coding in neurons? prob have to get into specific region

maybe classify different ways of transfering information between them?

sawtooth beta?????? wtf is up with these waveforms

if learn stuff about thal, can maybe help with parkinson's?

rlnshp b/w D1 DA rec b/w epi & park's

inactivation variables

  • h is not so much an inactivation variable (i.e. level of inactivation) as much as a DE-inactivation variable (where completely inactivated is 0 NOT 1!!!!!)
    • "Recall that h is the probability that a Na channel is NOT inactivated", p. 49, Ionic Channels of Excitable Membranes
    • another way of thinking: h is prob that the "inactivation gate" is OPEN, allowing current to flow – so if it's 0, then the gate is closed
    • NOT inactivated = the inactivation gate is OPEN, omg
    • see and "Ionic Channels of Excitable Membranes" for why the original HH paper theory that it's an inactivating PARTICLE has been deprecated for theory of an inactivating GATE


  • fmri methods wiki
  • cwatson's notes on fMRI
    • look at Friston & Dolan, NeuroImage, 2010 (Volume 52, pp 752-765), and the references therein
    • also look at Friston's Dynamic Causal Modeling
    • first paper for that was Neuroimage, 2003
      • although that isn't quite like the modeling we did in, e.g. cn510
    • also I think there's a paper from 2003, with Friston and Breakspear as two of the authors, I can't remember the journal
    • a very recent one (that I haven't read) is Hellyer et al., J Neuroscience, 2014 (Vol. 34, Issue 2)
    • and you might want to check out Deco et al., J Neurosci, 2013 (Vol. 33, Issue 27)



lrn-neuro-anatomy (lrn-anatomy) lrn-neuroanatomy


  • compile list of NS methods, bycategory
    • a very old ephys review calls LFP, or "micro EEG", as most similar to electro-"histology"
  • technologies
    • electrode types (tons)
    • cochlear implant types
    • neuroprosthetic types
    • are there any major catalogues of these?
    • emery says fronto has no anatomical connectivity to parietal, only goes through thal
    • emery agrees human thal has 1 IN for every ~5 TCs, inside nuclei (i.e. not including RE)








go back over krane's intro (nick james has krane)

study general relativity, like einstein's book?

Could warp across space-time by generating points of huge gravity, from either huge masses or masses from energy?

books (look at this e-library);brand=ucpress

why momentum works


find news source for local political news, pay attn to local politics

go over thinktanks < united_states>


also check out google's publicdata thing?

read less wrong, maybe take notes on essays? do the starter sequence

  • study biases!

look anti nuclear stuff to see what they say point by point

add to um to listen? podcasts

open-politics, a la open political data


wtf! mdwiki allows wikis on github pages sites???


  • notes from procrastination talk
    • we put things off thinking that we will have fun in the meantime, but usually it's guilt instead, a paralyzing emotion
    • we put things off by cooking, cleaning, other work things, things that have a "moral" component of "well look i'm doing this thing"
    • what we THINK we understand about procrastination is not actually true, but we still think these things in everyday exp:
      • we think it's about having fun (it's not)
      • we think we're not affecting our future self (we are)
      • it's all about giving in to feel good (it's not)
    • "tomorrow is a mystical land where 98% of human productivity is stored"
    • one new defn of procr is "the gap b/w intention and action"
    • procr is a subset of delay, "all procr is delay, but not all delay is procr"
      • example: if you plan to do hw for friday on thurs, that's not procr. If you get to thur and decide you don't "feel" like doing it, that's procr
      • voluntary, irrational delay
    • xxx [therefore, an important part of deciding BETWEEN delay and procr is planning!]
    • grad students esp feel this badly because there is much they need to and SHOULD delay, but they equate that with procr in their heads!!!
    • procr doesn't just affect performance, but also well-being, health, relationships, etc.
      • procrastinating students are actually sick more
    • think about how procr directly affects spending time with others
      • and how the stress you get from it spreads to other people like a disease
    • regrets of people bereaving those close to them who'd died is basically procr
    • fixing it
      • thinking about your agency
      • the most precious thing you have is time
    • procr is "NOT a time-management issue", a planner does NOT help
      • procrs tell time just like normies (study done by one of his students)
      • another of his students studied if procrs suffer the "planning fallacy" where they plan they're going to do a TON and then don't live up to it
        • procrs did NOT suffer planning fallacy (difference between expected and real understanding of how much studying they'd done)
        • procrs did study less and study later, but completely understood that they had done this
    • it is NOT an issue with planning
      • lots of people do their planning, but don't plan what they're going to do right now
    • he gave an example of a meeting with a student after which he knew the student didn't keep working for the rest of the time that day
    • xxx need to keep working AFTER successful meeting, use those good feelings for productivity
    • he says he never procrs anymore since he UNDERSTANDS the self deception
    • part of his understanding is of how actually long (and not that long) it takes himself to do something
    • just writing out a plan relieves stress, but does NOT solve the issue!!
    • procr is "weakness of will" and exerting self-control
      • the more you say "well I'll just do this thing that only takes 2 minutes", you're still facing the same hard decision and willpower threshold 2 minutes later
      • in that moment, you're either going to work or not work
      • "at its heart, it's about weakness of will, and about self-regulation"
      • it's a self-regulation failure, like buying things you don't need or eating things you don't need to eat
      • xxx it's about willpower
      • "I want to feel good now"
      • esp with a thesis, when you get it back and it's not perfect/has lots of edits to do, you feel bad! [and that's okay] you get negative reinforcement (emotionally)
      • "negative reinforcement" is the removal of a negative stimulus, like using an umbrella
        • "procr is neg reinforcement", since the task makes you uncomfortable, therefore you're avoiding the task (to avoid the neg stim)
        • xxx therefore, by USING procr, you're REINFORCING that it's effective
        • it's a reinforced habit, and is difficult to change
      • [i originally was planning to talk to therapist about lack of "energy" of me doing things, but what if the real issue isn't energy but willpower? what if it's nto that i'm tired (although that does happen, but less frequently and a nap cures it), but it's self control and not enough willpower to work?]
        • the real issue may not be that i'm tired, but that I'm UNWILLING
        • listening to this discussion esp about personality and impulsivity, sounds like to work on solving this procr, i need to better understand the psych report and my personality
    • there are multiple parts of this "self-regulation failure" - how do you intervene in these?
      • personality
        • how to intervene
          1. dutiful, self-disciplined people don't procr much
          2. low "conscientiousness" (wishing to do a task correctly and thoroughly) means procr
          3. high "impulsivity" means procr
            1. "they can't shield one intention from another intention"
          4. perfectionism
            1. difference b/w self-perfectionists vs socially-prescribed perfectionism (driven by other people)
              1. the socially-prescribed perfectionists live up to OTHER people's standards, have lots of negative internal talk, and score high with procr
            2. "perfectionism isn't bad, it depends on the flavor"
          5. emotional intelligence
            1. higher emo intel, lower procr
            2. it's a skill that you CAN learn
            3. "I said i was gonna do this, it doesn't matter how i feel"
            4. task: just track how you feel about doing different things and write it down
            5. neuroticism is emotional instability
              1. those who're neurotic have emo rxns to work that are WAY out of proportion, low lows and high highs, even when the work is NOT A BIG DEAL
            6. procr will stay up all night, do task, turn it in, feel great, but then not use that good feeling to do more work
              1. real learning should be constant (at lower magnitude), not stop and go with cramming and then unworking
            7. you have to learn to USE your emotions but also not react to neg emotions
            8. "You can have fear but you do not need to BE your fear" (fear is the mind killer...incl how it applies to learning!)
          6. procr is also a developmental thing
            1. PFC is last brain part in dev
            2. if you've "achieved" your identity, less likely to procr
            3. "you can't expect an 18 yo to act like a 25yo"
        • lots of personality traits are inherited???
      • the nature of the goals
        • how to intervene
        • task aversiveness
          1. what makes a task aversive?
            1. splitting a task into inception, planning, and action phases,
            2. boredom/frustration at every stage
            3. frustration at inception and planning phases = due to lack of meaning/enjoyment/(importance?)
            4. frustration at action phase = due to lack of structure/control/uncertainy
          2. temporal construal theory = are you thinking abstractly or concretely, and how that affects our temporal interpretation
            1. "when you think about things concretely, you think of them in terms of today, but abstractly, tomorrow" (recent psych result)
            2. so it's not jsut breaking things down into baby steps, but the fact you're thinking about real world actions, which makes your brain give the things more urgency
        • how we think about the task
          1. "I work better without pressure" fallacy aka the "arousal procr"
          2. quote from thesis "The mental discipline necessary to work towards a deadline is something that you must develop. It can become a habit just as lettings things slide until the last minute can become a habit. That pattern leads to staying up all night and writing in a blind panic. Besides ruining your health, you never can write your best. If anyone tells you "I have to wait until the pressure is on before I can start to cook", don't believe it. Occasionally, you may be able to work under pressure of a deadline, but stop kidding yourself, it won't be your best."
          3. uh make deadlines
          4. several quotes from music industry people about how you THINK you'd work better under pressure, but that's not true and not how it works
          5. procrs think "it could have been worse", but that's just making an excuse (and you're always going to be able to say that to yourself, but it's never helpful)
      • self-control and willpower
        • how to intervene
        • willpower is like a muscle
        • self-regulatory depletion
        • implications are that you can BUILD it up like a muscle
        • can build it using "value affirmation"
        • even when you think you have nothing less, there are ways to motivate yourself to increase your willpower reserve
        • meditation helps
          1. [is that because it helps you recognize what you're doing from an external POV? what is that called?]
          2. meditation helps attn, and attn helps willpower
        • "implementation intentions"
          1. sort of like Ivy Lee method where you just say, "When X happens [e.g. done with task, end of talk], I'm going to do Y (read a paper), then Z"
            1. prob esp helpful at beginning of day
          2. "people who make implementation INTENTIONS tend to follow through"
          3. (sort of like a super todo list, but tied together like triggers)
          4. THIS is writing down the list of what to do today in nb
          5. saying "i'm going to spend all day doing a task like writing" when you've NEVER spent the entire day doing that is not helpful – what is helpful is finer granularity, like "11am-noon I'm going to write, then eat lunch, then 12.30-1.30 i'm going to write more"
            1. that's also asking too much of a single interval
            2. once you get into the habit of doing something like that more frequently, then you'll have more of it done and won't need to do it all at once
      • cognition and beliefs
        • how to intervene
    • procr will stay up all night, do task, turn it in, feel great, but then not use that good feeling to do more work
      • real learning should be constant (at lower magnitude), not stop and go with cramming and then unworking
    • fear is the mind killer and the procrastinator


20130326 professional devlpmt talk

  • lots of assumptions about attn during talks (esp w/ science)
  • long pauses
  • engagement w/ audience
  • human contact and stories
  • diff b/w msg, speaker, and audience
  • aud doesn't on verbal or visual; integrates both all the tiem
  • "provide a mental struct that uses prior knowledge"
  • don't turn off lights for a post-lunch talk
  • narrow down message to land only 1 thing
  • "you have to want the audience to learn the thing"
  • he discourages using notes
  • now he says 3 points
  • hard to distinguish b/w his real Q's and rhetorical Q's
  • "hyperlinks" and linking to web during talk, omg
  • wtf multitasking? "being intentional"
    • use consistent themes
  • go grab related pieces of news
  • "stories are the most engaging things"
  • could always put it more casually (like at the end of a slide)
  • have a midway synopsis slide?
  • he keeps pushing relevant engagement as if people don't know your stuff at all
  • stupid handout
  • what to do if eqns put people off?
    • he doesn't seem confident in his response, more about "what's most relevant to most people"
  • think of the ones in the audience who know the LEAST on your stuff
  • give enough for the people who follow to graspp the details, then shift and bring in the poeple whoo don't need it
  • this guy has a 90s-level understanding of powerpoint
  • he uses hyperlinks as if "techniques"
  • powerpoint is low-info-resolution; when need higher, go to print
  • "the Google search engine"
  • "be metaphoric in your designs"
  • use metaphorical images
  • carve out a specific niche and try to attack the likely questions people will have
  • make a media library for yourself, esp if it's diff take, esp w/ human element
  • how do you control your speech?
    • arrive at a place of ease, "who gives a s—"
    • "less is more" for speaking
    • once again, simplify your speech
    • not as important as content
    • "be present" [he means mindfulness]
    • look at people and smile at them
    • make sure people come AND leave with questions, not answers
  • review articles for quotes


  • Nancy at SZ collab meeting, 2014_0402: "let's try to write down questions for which this group are the right people to do it"

lrn-networking lrn-professional-networking / long-term-networking

  • frank: biotech/industry/startups, ESPECIALLY in boston expect YOU to reach out to them, there exist certain pipelines from certain institutions

lrn-presn, lrn-presentations


  • Scott and Jon answering questions about working in industry at GPN retreat 20170608
    • most results in industry are negative, and expect a lot of them (unlike academia)
    • what they strongly frown upon is sloppy sci tech, which are more important than what you know




  • Exact Audio Copy and ffmpeg for transcoding

lrn-sailing :lrn_sailing:

sailing notes, day 1 2014-09-20

  • agenda: the boat, wind, turning, knots
  • can take (discounted) $39 multiple choice test after finishing class for most basic registration
  • chris allen teaching
  • "Courageous" an NPO, and "Community Sailing on the Charles" also recommended
  • get 35 days of member sailing after class, that may roll over
  • parts of boat
    • omega symbol on sails means it's a student, everyone on water will see and understand that
    • hullspeed eqn determines literally the max speed of the boat, f(length on water line) = 7 knots for solings
      • need 12-15 knots windspeed for max boat speed, but >15 knots doesn't make it any faster!
    • need "one hand" of wtr b/w ground
    • can't use tiller to brake
      • xxx "can't brake, can only take foot off the gas"
    • heeling = when boat turns lat on its side
    • "lines" = boat term for lines
      • can use "camp or cleat" to catch/sit lines near winches
      • you "trim" vs "ease" the line
    • "blocks" = boat term for pulleys
    • standing rigging = literally the rigging on it 24/7, even when at mooring
    • running rigging = stuff like sails that you put on to go
    • can BEND the sails like wings, which planes can't do
    • can trim the aft stay to bend the mast (don't worry about plastic deformation)
    • the boom is free in its range
    • watch spongebob episode about parts of the boat/learning to drive
    • luff = Leading edge of sail (see "luffing")
    • clew = no clew what the name is, it's not the tack or head
    • 4 lines
      • main halyard, hoists main sail up
      • jib halyard
      • jib sheet
      • main sheet
    • in rigging, you "tack on the tag" [connect the tack of both main and jib to their things]
    • sails shaking = luffing [boat is telling you it wants directions]
    • stays = permanent metal wire lines going up the mast connected to the boat. Frontmost stay is what the jib tack goes on
    • jib sheet loosens/tightens jib clew
    • "roaches and leeches" on main sail (leech is the perfect triangle, roach is the extra fabric on sail)
  • port/starboard are fixed, but there are also wind-relative side terms "winward aka weatherside" and "leeward" (pron loohr-word)
  • right of way: if starb = win, you have RoW
  • "the boat leans to lee[side]" [but it wants to turn into winward/wants to approach the wind]
  • helm = person holding the tiller, and sometimes the main sheet
  • the wind
    • spd and dir
    • there exist nautical vs statue miles
    • 17 mph = 15 knots / good enough
    • solings want 5-10 for learning, 10-15 is fun, >12 is intense
    • there are smaller storm sails for high wind
    • gusts are often 20 knots
      • gust mgmt is adv
      • can see them coming on the water sometimes
    • 0 is the worst wind speed
    • Boston harbor wind mostly S-SW (meaning coming from)
    • winds on the sail
      • when going upwind, front of sail is a high-pressure sys, back of sail is a low-pressure sys, which makes a lift vector called "attached airflow" or pilots call "laminar flow"
      • air on both sides reaches end of sail at same time
      • the lift vector is usually very perpendicular to the wind dir, but when combined with vertical keel hydrodynamics, makes an upwind vector
      • this all is called lift
    • jib has red and green "telltales", and you want them BOTH to be strong
      • if inner red weak, trim it, and vice versa
    • "when in doubt, let it out"
    • leech has tails that come out in rear
    • "points of sail" 360 degree wheel, from
      • into wind / "in irons",
      • "close reach" maximizing lift,
      • "beam reach" named for flat beam behind helm
      • "? reach" more downward, starting to use "drag" instead of lift
    • wind coming on left side = "port tack", wind coming on starboard = "starboard tack"
    • "head up" = turn nose to point up to source of wind
      • done by pulling tiller into wind
      • as opposed to "fall off"
    • solings are good at not staying in irons
  • turning the boat, aka "tacking"
    • to turn, push tiller into sail
    • pushes rear of boat around wrt the bow
    • when tacking, don't need to adjust main, since is "self-tacking"
    • NEVER let go of the tiller, also bad if not looking at front
    • newbs tend to turn around when tacking - don't
    • tacking the jib
      • when turning L, loosen the R jib line and pull the L line
  • in irons is sometimes called the safety position
  • xxx in "jibbing aka gybbing" (turning downwind), can get "uncontrolled jibe" where jib starts luffing and then turns around to be opposite the direction of the main; this is VERY bad, as the boom will then come snap around VERY quickly
  • helmsman is "in the box"
  • "listen to the jib"
  • man overboard rescue is usually either a figure 8 or a "quick turn"
  • knots: cleat hitch, bowlein, stopper knots (8, pigtail)

sailing notes, day 2 2014-09-21

  • if going downwind in a far reach (can't remember exact phrasing), can be "sailing by the lee" where boom is still on the winward side
  • downwind : drag :: upwind : lift
  • starb tack has RoW over port tack
  • if get a "puff", can
    • let off main sail (best, simplest)
    • head into wind
    • sit differently
      • "the boat hates people"
      • "everyone should sit on the winward side => the boat will point to leeward"
  • most difficult thing is keeping a course
  • jib vs jibe
  • best time to do jib during tack
    • to release, just when jib? start luffing
    • to tighten, wait until luff of jib crosses the mast
  • he trusts candied ginger for seasickness/nausea
  • "crash jibe" = immediately going to start a jibe bc of emergency, watch your head
  • mooring
    • exiting mooring
      • let the jib fully luff when unmooring
      • can artificially (aka with hands) tighten jib to get a kickstart
        • aka "backing the jib"
        • this is used to change dir, NOT to go fwd
    • entering mooring
      • open the jib completely
        • this decr wind power
        • this change the handling too, so be careful
      • once "on the moor," should luff and/or drop main
      • can easily drop the sails by pulling on luff
      • xxx "luffing a sail" = loosening
      • can always "pass" on a particular mooring = good practice too
      • "shopping window" of potential moors is usu b/w 270 degrees and 320 degrees, if on a reach/going upwind
      • lots of mooring maneuvers exist
      • "sneak in from behind, let sails out early"
  • man overboard
    • important things
      • cry out (say it)
      • throw flotation, even if they're wearing a lifevest (so you can see the flotation)
      • get a spotter who constantly maintains eyes and arm pointing on the person (4.) do NOT jump out, since more people in water = less control of boat; only jump in for elderly, invalid, or infants
    • "quick turn" easiest, best
      • immediately turn to beam reach, go 5/6 boat lengths,
      • then jib,
      • then turn or tack to return to where the person is, loosening sails to slow down when close
  • Balance
    • CLR/"the pivot" = center of lat resistance, can rotate around this point/vertical line
    • CE/"the push" = center of effort, midpoint of mass b/w the sails (and maybe people as well?), so it provides torque around the pivot, and change location depending on absolute and relative location of the sails
    • knowing this really helps for better sailing
    • "weatherhelm" is idea that boat tends to turn into the wind
      • this is by design, so it keeps you awake etc.
      • dealing with weatherhelm is newb stuff, you should be able to compensate
      • this includes that you need to "fight" the boat doing this to go in a straight line
    • if strong wind, then main will get more, which shifts the push more to main
    • if main gets too much push, can get enough weatherhelm so that the push is so rear that tilling isn't effective
      • letting main out is best solution to get rid of this
      • "the less balanced the boat, the less effective the rudder"
        • this includes heeling
        • too much heeling is inefficient
    • bending the mast also moves the push to the rear
  • leeway
    • if trying to reach a point, should aim a little towards the source of wind to actually get there (have to go sideways a little)
    • this can happen especially for tidal reasons (think of the hydrodyns of the keel)
    • xxx read more about this
  • rules of the road
    • if vessel collision, EVERYONE is at fault
    • diff boat classes has more or less RoW
    • stand-on (so) or give-way (gw)
    • gw means the boat with more control (often facing the wind more) has an easier time getting out of the way
    • so means you should maintain your course and speed
    • starb tack: if port tack and starb tack both going upwind and headed on collision course, port tack is gw and should move
    • leeward boat: if in irons/close reach, and intercept w/ a beam reach boat, beam reach is gw and should move
    • overtaken: ??
  • big boats: 5 loud blasts = danger, idk what you're doing and I can't tell
  • spinikerup = restricted maneurver (has RoW)
  • Apparent Wind (AW)
    • relative wind feeling
    • often when you feel wind loss, it's this
    • changing dir of AW
      • as boat goes forward, gradually wind vector will approach course
      • "your boat spd has bent the dir of the wind"
    • if want more wind, head up, since AW incr, instead of falling down (which would give you a better angle)

sailing notes, day 3 2014-09-27

  • backing the jib = using your hands
    • like when leaving mooring and want to turn
    • if main is tight, then can make this much harder
  • if someone falls off, can do
    • "quick stop" (where come by on downwind),
    • "figure 8" which is commonly taught since involves a tack instead of a jibe, and
    • "quick turn" which is what they taught me
  • pick people up on WINWARD side of boat, not leeward – wind can knock boat into person, better visibility
  • weather-helm: when rear of boat becomes boss
    • reduce it via
      • let main out
      • head up into wind
      • tighten jib
  • "fall-off" - away from wind
  • jibe has less turning power then tack
  • right of way: if both starb, tiebreaker is who gets wind first

sailing notes, day 4 2014-09-28

  • bending mast = brings CE back and tightens
  • outhaul = incr curvature of foot
    • thing with the weird tie at bottom of boom
  • cunningham helps upwind, most vertical portion of main
  • fuller sails = "keep the wind"
  • boom vang is how you take twist out
  • Navigation 2
    • "obtaining a fix..."
    • D = RxT
      • Distance comes from chart
      • Rate is difficult, use an app/GPS
        • recommends "iSailor" app
    • inside ring is mag compass, outer time [? can't read word]
    • dead reckoning - figuring outhere you are between two places you absolutely knoow where you are, like a buoy
    • cross lines that you get
    • range = where two tings line up
    • bearing = looking at single thing
    • all boats use mag north
    • tide/current is like +- 4 degrees
    • 90 is the amount of leeway you must allow to the ???
  • tide/current
    • add high and low knoots are added to soundings
    • be careful of supertide (goes below)
    • hours of tide (roman numerals usu used):
      • 1/12 of tide comes in
      • 3/12
      • 6/12 (half the tide comes in the middle)
      • 9/12
      • 11/12 (slows)
      • 12/12
    • going out: ebbing
    • if current against, go shallow
    • if with you, go deep
      • deeper water = stronger current
    • wind and tide and current: don't have 2 of them against you
    • "you plan tides and wind"
    • we know about storms 4-5 days in advance
  • weather
    • "anvil clouds" out of NW
      • if seem them, have 30 mins to get to shelter
      • turns cold at 10 minutes
      • squalls 15-30 min long, 25 knots
      • if get caught, drop sails, maybe store them
    • everyone runs aground
      • think where you are in the tide cycle, wear lifejacker (if happens)
      • keel is the thing usu caught
    • "topping lift" = same place a standing halyard, used to keep up the boom
    • float plan = just tell somebody when/where going
    • coast guard can stop you anytime, "courtesy inspection"
    • !!! need as many lifejackets as people, plus throwable
    • if motor, needs a fire extinguisher
    • must have a lookout at all times (can be the skipper)
    • CAN drink on boats, but just can't get hammered
      • no formal BAC amount, but can get in trouble with law if drunk
    • big boats
      • 1 toot means "leaving you on port", 1 response means "yes", 5 means "no"
  • sail shapes
    • wind at top of sail is much faster
    • when wind faster, usu dir is diff
    • so far, we've just been using the halyards just to raise sail, but can use it to change shape of jib or main
    • loose sails can "hold" wind for longer
    • jib down haul - pulls down jib tack to tighten
    • pulling Cunningham allows more curvature in main
    • can also loosen sail via loosening sail clews
    • draft of sail = how curved it is?
    • jubs and mains have backstays that can tighten
    • jib fairlead (that's the things that you pull the jib lines through)
      • pulling back lets off gas
    • twist (in main)
      • raise boom = allows twist
      • "boom vang" = triple pulley thing, cxn boom to mast
      • !!! "golden rule of sailing" = want top batten and boom aligned
  • buoyage
    • magenta on maps = has a light
    • box containing
      • orange/red circle = advisory
      • orange/red square = warning
      • orange/red diamond = don't go, really bad


researchgate and


criticism of peer review

reproducibility crisis




lrn-effective-altruism EA






lrn-studying lrn-reading

PQRST system - Preview, Question, Read, Summary, Test

Studying for Science, in calibre

how to read a book

  • follow your own goals
  • first read overview/discovery, second for detail and understanding, third for notes which he means exec summary in your own words about all major points - discovery should be very small, just major things
  • use multiple "modes" for review, like talking, visualizing
  • author context
  • general arguments are far more important than details
  • know how long you're going to read
  • read as if you imagine you're going to have formal argu with author
  • one hour (and an initial fifteen for easing in) is a good time for concentration. 3 1hr sess are better than 1 3 hr sess
  • actually estimate times for reading!
  • prefer margin notes that summ over highlights. And if highlighting, only words not sentences-want to minimize reread
  • [standalone mental rehearsal?]
  • these techniques will get better over a few months!



learn symbolic regression

jason's idea to optimize models based on fundamental laws of physical systems derived from symbolic logic,

  • read schmidt, m., & lipson, h. (2009). distilling free-form natural laws from experimental data. science, _324(5923), 81–85. doi:10.1126/science.1165893
  • his email:by the way, when i've seen it used, the term "model optimization" usually refers to "parameter optimization".i'm using "model optimization" to refer to the optimization of the mathematical form of the model (i.e., the same way statisticians discuss finding the best statistical distribution to model data).
  • continuation:ok, one more addition - in practice, when statisticians talk about choosing the best distribution to model data, they are choosing a distribution from a discrete set where the choice is based on known properties of the different distributions.when i say "model optimization", i'm thinking more of an optimization by the incremental variation of a mathematical form (than property-based selection from a discrete set).however, if i were debating myself, i would point out that most distributions used in applied statistics are actually derivable from the gamma distribution (a well-known fact considered in the theoretical domain).thus, the continuous variation of gamma distribution parameters will sweep a range of mathematical forms (= that are commonly applied distributions) that could be optimized in the same sense i was describing.haha, ok, enough of that.


avery's first-year grad school teacher starts class with an open ended question, and connects people's answers in a web through discussion, and then saves and posts to the web


eventually should learn air deals via < r_francereservationagentthat/>

can track trends via


brian nosek says In writing, lead with the evidence, follow with the explanation.

writing essay to eric denovellis after briefly reading chapters of his thesis:

(tl;dr: you often write sentences with many clauses, just like I do. For several reasons listed, it's probably a better idea to break most or ALL of those sentences with multiple clauses down into individual sentences.) essay: Also, do as I preach, not as I do. All this writing below is cursed by what exactly I'm saying that neither of us should do, but that we tend to do. simplify, simplify, simplify – you write a ton like how I natively write, specifically subjects with lots of adjectives/etc., like 'context-relevant sensorimotor information'. Recently, Nancy did more than tear apart my attempt at an SFN abstract, to the point where she effectively rewrote the whole thing - and I couldn't deny that what she wrote was way more understandable. The key difference was shorter sentences. Much of the time I would be trying to say very, very related things together within the same sentence, since the science itself flows between them, but every single clause you add to a sentence adds SO much overhead if the person only 90% truly grasps what you said in the first clause. Honestly, multiple-clause stuff is probably fine in a thesis (where the main readers ARE going to be able to directly follow you easily since they have maybe even already read the papers you're discussing) . But if there's one thing that stuck out to me when I was doing a direct comparison between my writing (very similar to this) and Nancy's for the same piece of writing, was (being brutally honest here), how STUPID the latter was. I mean that the latter was better written, but honestly the grammar was so simple that a child could have read it, it was SO plain. There were many more short sentences, but just like a function, each one served one purpose and one purpose only. I tried thinking "how do i change my perspective to write like this" and honestly the first thought in my head was "write it like you would for someone super super dumb"...and that's both horrifying that that was my thought, but also that was what made it great! As you read the simplistic version, your head responded like "Okay. Okay. Yeah. Well, yeah that follows. Yeah. uh-huh. Yep." By looking at the multiple clause sentences like right here at the end of page 91, i see the same thing I do natively, which I have to edit it out. Splitting up all these super-related clauses will, yes, make it sound less "professional" or academic...but I think that's because it exactly makes it SIMPLER, and I'm of the opinion now that that helps the reader understand the stuff better than ANYTHING else. The very pseudoscientific way I think of it is, imagine your brain, the compartmentalization/compression of each of these major points is probably FAR better in your brain than in mine – you've read these papers multiple times or at least have kicked around the key points and obvious things from the writing in your head for years, almost like you think of each of the ideas as if they are individual words. You encapsulate them in your mind very well, because you've spent years in grad school obsessing over these and related ideas. To someone else, even someone who is familiar (but not as much) with the work, each of these major points are going to be more like sentences. When I read "task demands' expected values increase" I have to process it, like "okay, the subject is task demands [which part is that in? let me look back a few sentences], what's it doing?". To you (just like propofol phase-amplitude coupling to me), that entire clause is a single "idea", fully encapsulated, but to someone else, it's an unknown connection between the subject, what's it's doing, and how that's different than a very similar effect on the same subject or vice versa, nevermind anatomical localizations. By the time I get to the next clause, I may (and was) still trying to process the first "connection" or result, as that's literally the first time I've been hit with it, and I need to learn it in order to move on. If two major points/results (from papers) are different clauses in the same sentence, then in all likelihood i'm still going to be trying to figure out what you exactly mean by the first as I'm moving onto the second, and then I'm trying to build two different connection/result ideas in my head, etc. Furthermore, I'm not going to have any idea if the second clause REQUIRES the first clause, or if they're completely independent (as they most of the time are). I think that, to someone who has extensively embedded such ideas in their head, each of these results/clauses are going to seem like single entities to them. For the same reason, several of these are going to be related, and it's going to be obvious that "result 1 is important, and result 2 is tied to result 1 so result 2 is this". But for the lay person, it's nontrivial just to figure out what result 1 is and all it entails, and putting more results in the same sentence is going to force them to almost try to parallelize figuring out results 1-3. I'm not saying this is a run-on since it's not, it's just a long sentence, but I think my pseudoscientific bullshit explanation above is a big part of why academics (people who understand a particular idea really, really, obsessively well) write such long sentences. YOU understand that "Ax=B" is Thereom 1, and so you say "Thm1, and Thm2, and Thm3", but for lay person seeing "Ax=B" the first time, it's harder on them than it is on you for understanding that the first time they see it. The "academic" way of writing would be this really long sentence (except in mathematics...but we'll get to that), but I think the "simpler" way of splitting it up into, say, individual theorems that the reader attacks one at a time, goes a LONGGG way towards simplicity. That may actually be part of why Nancy's version of my writing was so much clearer: the separation between what constituted individual "ideas" is so much cleaner, almost like how math manuscripts don't go "Thm 1 => Thm 2 => Thm 3", but instead treat each of them one at a time. Separating by sentences also tells the reader "you don't need Thm1 to understand Thm2, as they're independent, and you can take all the time you want just to focus on the full sentence that is Thm1" Then, later, when you actually DO need Thm1 AND Thm2 to imply Thm3, you can refer to them using MUCH simpler terminology. I'm sending you both my and Nancy's abstracts so you can see the differences/superiority I'm talking about; honestly, it was a super good exercise for my understanding of writing differences to compare/contrast them. Doing so convinced me that 1. her way is better, and 2. it's just plain simpler. Just making the sentences shorter overall makes each of them THAT much easier to understand. It can almost make you sound like a simplistic robot, but the separation of ideas reduces cognitive load SO MUCH that I'm currently of the opinion that that is the MOST IMPORTANT thing to change in my own writing. And because you seem to write VERY similar to how I do, I think that same change seems relevant.

A manual for writers Kate l turabian

lrn-writing-systems/software! currently using pandoc

  • The Plain Person’s Guide to Plain Text Social Science


  • notes on "The Science of Scientific Writing", American Scientist, 1990
    • expectation: minimize distance b/w subject and verb
      1. make each sentence ("unit of discourse", e.g. everything before a semicolon) on one thing and one thing only, or serve a single function, or make a single point
      1. stress position/emphasis: put most important stuff at end of sentence
    • use the beginning of sentences to serve as "topic position" either for linkage to previously established material or predicting future points
      • BUT don't start sentences constantly with brand-new ideas that haven't been introduced yet
      • don't use the begininng of sentences to give things their first appearance
    • "a sentence is too long when it has more candidates for stress positions than there are stress positions available"
    • number 1 problem in professional writing: misplacement of old and new (to the reader) information
    • [like i suspected] put OLD information at beginning topic position, put NEW information at ending stress position
    • always make sure to include connections to older information (linkages)
    • they consistently list out the main verbs of each sentence to examine the structure of the abstract
    • "articulate the action of every sentence in its verb"
    • provide context before giving new information
  • Purdon, Patrick on writing sci papers: "methods first, then results then...whatever intro/disc"