org-mode - markdown backend -- Emacs 24.3.1

A little reminder of something I failed to do before. Org-mode has optional export backends, some are loaded by default:

(message "%S" org-export-backends)
;; "(ascii html icalendar latex)"

(add-to-list 'org-export-backends 'md)

(message "%S" org-export-backends)
;; "(md ascii html icalendar latex)"

Now in an org-mode buffer, `C-c C-e` (org-export-dispatch) will have a markdown entry accessible through `m`.

ps: done on GNU Emacs 24.3.1 (i686-pc-linux-gnu, GTK+ Version 3.8.2) of 2013-08-07 on -mnt-storage-buildroots-staging-i686-eric; archlinux

lambda linguistic and logical interface

For a long time I've been amazed how much the `lambda` expression construct gives you. The introduction of name into expressions is seen through `let` which is syntactic sugar for lambda underneath. I believe Scala `for` nested iterations are also syntactic sugar toward flatmap + lambda. It's also a clean expression of data dependency. The sub-expression has meaning only if the lambda abstraction can be reduced by application to values. Otherwise well it's another smaller lambda expression. Wonderful.

reified side effects

Just wondering if there was some research into reifying the effected external entities back into the 'denotation'.

Instead of

f :: a -> b -> IO ()

you'd get

f :: e -> a -> b -> e

now that e is inside our algebra, maybe we can see programs as pairs (local,global)
bound by a protocol, or a fixed model of state transitions.

that's all.

Form vs editable trees (DOM). Shifting toward simplification.

html has this split between content (p,li,..) and communication (form,input) that will seems less and less justified. With html5 you can modify any node. Then transposed as json which can be sent to another peer as a structured message. The old way is a nice in-standard microDSL implemented by the browser, which guarantees freedom from javascript. This would be a large move off the 'web' as it's perceived but ...

Pedestal.io is already far away from the current paradigm, I believe they do send subtrees diffs between clients and servers. This is fascinating, clojure guys are jumping boldly on the fringe (edn, datalog, datomic, pedestal).

ps: to be edited.

pyalpm : archlinux pacman bindings in python

Wanted to toy with python and pacman. Starting with subprocess calls to `pacman` and ugly string processing. But the google fairy told me I was heading in the wrong direction.

sudo pacman -S pyalpm

5 minutes later, I choke. I can't find a way to instantiate anything. ipython helps navigating modules, 5 more minutes I find a "hack" to get a Handle and then the local package database.

import pyalpm

PATH="/var/lib/pacman"

def init():
    "() -> db"
    h = pyalpm.Handle(PATH,PATH)
    return h.get_localdb()

DB = init()

def installed_pkgs():
    return DB.pkgcache()

In [115]: pkgs = installed_pkgs()

In [116]: pkgs[:10]
Out[116]: 
[<alpm.Package("ghc-7.6.3-1-i686") at 0xb5a1b130>,
 <alpm.Package("qemu-1.5.2-1-i686") at 0xb5a1ce50>,
 <alpm.Package("chromium-29.0.1547.65-1-i686") at 0xb5a18ab0>,
 <alpm.Package("emacs-24.3-4-i686") at 0xb5a18ae0>,
 <alpm.Package("faenza-icon-theme-1.3.1-2-any") at 0xb5a18d40>,
 <alpm.Package("python-3.3.2-2-i686") at 0xb5a1ccb0>,
 <alpm.Package("jre7-openjdk-headless-7.u40_2.4.1-3-i686") at 0xb5a1b8f0>,
 <alpm.Package("gcc-4.8.1-3-i686") at 0xb5a1b080>,
 <alpm.Package("python2-2.7.5-2-i686") at 0xb5a1cd50>,

 <alpm.Package("linux-3.10.10-1-i686") at 0xb5a1c420>]

<yay/>

archlinux :: mint 15 month long swap

update: 1 pacman -Su and 2GB later, boots ok, runs fine. One little detail, kernel warns about root being mounted rw when it shouldn't .. I couldn't read, boots too fast.

Fed up by, well, my mediocrity, I boot my archlinux machine again. Firefox shows a cached version of Google News July 30th 2013. It's been a month already. I just needed a debian based distro to try docker.io, arch is supported but it needs an aufs kernel module, which requires quite a lot of disk space to compile, my partition setup forbis such an amount. Anyway, Mint 15 was a agreable experience, after a while I forgot about arch, the overall better desktop configuration (better colors schemes, fonts, themes) made it feel stabler. I greatly suffered the windows manager intercepting all my emacs keybindings. also Xmonad screens suits me better than floating windows ala Mate.

The Mint 15 machine is less old, core duo => core 2 duo, with twice the ram and a non sluggish hard drive. A lot of things were faster. But a few days ago, I messed up the user/groups in Mint, I can't sudo anymore, booting singler user is useless since I believe root doesn't have a password, which is required by ubuntu/mint recovery mode. I don't even know what group I belonged to in the first place. Now I boot arch, the disk screams but 10 seconds later I'm under Xmonad. The terminal fires up fast, emacs fires up fast. Everything seems a little more reactive. This core duo can't decode 720p without choking but for most things everything is a little better. It's not even an engineered and polished system, I blended configuration examples, tweaks from wikis until it was good enough (c).

I have to admit, I forget a lot of things. Passwords took a little time to come back. I don't remember what I told Xmonad for launching the browser or the editor, fortunately I left the default dmenu one. Day saved.

Mint is not rolling, I missed the constant contact with arch repos. I'm in for a treat now. Pacman says I have 163 presents to open before midnight. Let's see if arch breaks as easily as I read so often on r/archlinux.

$ sudo pacman -Su
:: Starting full system upgrade...
:: Replace python-distribute with extra/python-setuptools? [Y/n]
resolving dependencies...
looking for inter-conflicts...

Packages (167): ack-2.08-1  alsa-plugins-1.0.27-2  archlinux-keyring-20130818-1
                bash-4.2.045-5  binutils-2.23.2-3  bison-3.0-1
                ca-certificates-java-20130815-1  cairo-1.12.16-1  cantarell-fonts-0.0.14-1
                chromaprint-0.7-6 chromium-29.0.1547.62-1  cloc-1.60-1
                cmus-2.5.0-3  cpupower-3.10-2  cryptsetup-1.6.2-1
                curl-7.32.0-1  device-mapper-2.02.100-1  dhcpcd-6.0.5-1
                dvdauthor-0.7.1-6  emacs-24.3-4  ethtool-1:3.10-1
                fbida-2.09-3 ffmpeg-1:2.0.1-1  file-roller-3.8.4-1
                fluidsynth-1.1.6-2  fontconfig-2.10.95-1  fuse-2.9.3-1
                gcc-4.8.1-3  gcc-libs-4.8.1-3  gdb-7.6.1-1
                gettext-0.18.3.1-1  ghostscript-9.10-1  giflib-5.0.4-2
                git-1.8.4-1  glew-1.10.0-2 glib2-2.36.4-1
                glibc-2.18-3  gnokii-0.6.31-6  gnome-icon-theme-3.8.3-1
                gnome-settings-daemon-3.8.4-2  gnupg-2.0.21-1  gnuplot-4.6.3-1
                gnutls-3.2.4-1  go-2:1.1.2-2  gpgme-1.4.3-1
                graphviz-2.32.0-1  groff-1.22.2-5 grub-2.00.5086-1
                gst-libav-1.0.10-1  gst-plugins-bad-1.0.10-1  gst-plugins-base-1.0.10-1
                gst-plugins-base-libs-1.0.10-1  gst-plugins-good-1.0.10-1  gst-plugins-ugly-1.0.10-1
                gstreamer-1.0.10-1  gtk3-3.8.4-1 guvcview-1.7.1-2
                hwdetect-2013.08-2  iftop-1.0pre2-1  imagemagick-6.8.6.9-1
                imlib2-1.4.5-5  intel-dri-9.2.0-1  inxi-1.9.14-1
                iproute2-3.10.0-1  iputils-20121221-3  ipython-1.0.0-3
                isl-0.12.1-1  iso-codes-3.44-1 jdk7-openjdk-7.u40_2.4.1-3
                jre7-openjdk-7.u40_2.4.1-3  jre7-openjdk-headless-7.u40_2.4.1-3  jshon-20130815-1
                kbd-2.0.0-1  keychain-2.7.1-3  kmod-15-1
                lftp-4.4.9-1  libarchive-3.1.2-2  libbsd-0.6.0-1
                libgexiv2-0.6.1-2 libmariadbclient-5.5.32-2  libmng-2.0.2-2
                librsvg-1:2.37.0-1  libsamplerate-0.1.8-2  libtiff-4.0.3-3
                libtracker-sparql-0.16.2-1  libusbx-1.0.16-2  libvdpau-0.7-1
                libwbclient-4.0.9-1  libwebp-0.3.1-3  linux-3.10.10-1
                linux-api-headers-3.10.6-1  lirc-utils-1:0.9.0-58  llvm-libs-3.3-1
                logrotate-3.8.6-1  lshw-B.02.17-1  lvm2-2.02.100-1
                man-pages-3.53-1  mariadb-5.5.32-2  mariadb-clients-5.5.32-2
                mercurial-2.7-1  mesa-9.2.0-1 mesa-libgl-9.2.0-1
                mkinitcpio-0.15.0-1  mkinitcpio-busybox-1.21.1-2  moreutils-0.49-1
                mpd-0.17.5-1  mplayer-36285-4  mupdf-1.3-4
                nasm-2.10.09-1  netctl-1.3-1  nodejs-0.10.17-1
                nspr-4.10-2  nss-3.15.1-1 ntp-4.2.6.p5-17
                openresolv-3.5.6-1  pacman-mirrorlist-20130830-1  parallel-20130722-1
                perl-5.18.1-1  perl-error-0.17021-1  perl-file-next-1.12-1
                perl-time-duration-1:1.1-3  php-5.5.3-1  php-apache-5.5.3-1
                php-pear-5.5.3-1  pixman-0.30.2-1  powertop-2.4-1
                python-distribute-0.6.45-1 [removal]  python-numpy-1.7.1-2
                python-pip-1.4.1-2  python-pygments-1.6-2  python-setuptools-1.0-1
                python-tornado-3.1.1-1 python-virtualenv-1.10.1-1  python2-mako-0.8.1-2
                python2-markupsafe-0.18-2  python2-paramiko-1.11.0-1  racket-5.3.6-1
                redis-2.6.16-1  redshift-1.7-7  reiserfsprogs-3.6.24-1
                run-parts-4.4-1  sbcl-1.1.10-1  serf-1.3.0-1
                sip-4.15.1-1  smartmontools-6.2-1  smbclient-4.0.9-1
                sqlite-3.8.0.1-1  subversion-1.8.1-2  syslog-ng-3.4.3-1
                tcl-8.6.0-5  testdisk-6.14-1  tk-8.6.0-2
                ttf-dejavu-2.34-1  unrtf-0.21.5-1  util-linux-2.23.2-1
                valgrind-3.8.1-3  vim-7.4.0-2  vim-runtime-7.4.0-2
                wayland-1.2.1-1  x264-20130702-2  xdebug-2.2.3-3
                xf86-video-intel-2.21.15-1  xorg-xprop-1.2.2-1  xorg-xset-1.2.3-1
                xorg-xwd-1.0.6-1  youtube-dl-2013.08.30-1 zsh-5.0.2-3

Total Download Size:    484.72 MiB
Total Installed Size:   2021.43 MiB
Net Upgrade Size:       27.16 MiB

:: Proceed with installation? [Y/n]

recomings.

The way on is always longer than the way back. It's insufferable. Once the circle done you know, not all, maybe a little, but enough.

The trip from high to low-level is strange, often difficult. Those who dive deep enough to see the atoms, and get back up will see the combinations.

Creating requires this loop, from ideal to real. Providing the loop is what matters. Iterating will distort it as needed.

As intended.