* Build system and packaging work

- [ ] =chicken= is using a positional argument for the file name... (note
  the confusing error message when you try specifying an option first)
- [ ] Missing LICENSE in egg tarballs
- [X] Implement an option for =chicken-install= to not fetch/install any
  dependencies
- [ ] Implement an option for =chicken-install= to retrieve eggs to
  install from a file (like with =@requirements.txt= or =<
  requirements.txt= or =-f requirements.txt=), with each line looking
  like =name[:version]=
- [ ] Implement an option for =chicken-status= to list versions in
  that format
- [ ] Report a bug for eggs installed via C5 not having a version
- [ ] http://code.call-cc.org/egg-tarballs/ should be official (see
  http://www.more-magic.net/posts/vcs-independent-distribution.html
  for details)
- [ ] Same goes for
  http://code.call-cc.org/cgi-bin/gitweb.cgi?p=chicken-core.git;a=tree;hb=HEAD
- [ ] Source links in the wiki (svnwiki hacking?)
- [ ] Do something against the hardcoded locations (having a globally
  installed CHICKEN and locally installed libraries doesn't mix as
  they're looked up in the global location unless you use environment
  variables, perhaps an approach à la =pip install --user= or =gem=
  and its user option could work)
- [ ] Eggs using =setup-helper= (like, =posix-utils=) are considered
  broken (as it's going for the =(include ...)= approach instead of
  the recommended =(use ...)=) and would need to be fixed before
  someone attempts packaging it as make dependency!
- [ ] Use =<chicken/chicken.h>= in generated sources

#+BEGIN_SRC
<Earnestly> wasamasa: All chicken needs is the ability for -p to be an actual prefix and an install location
<Earnestly> wasamasa: So `chicken-install -p /usr "$pkgdir"` for example
#+END_SRC

* =chicken-src=

** Notes

The equivalent to =chicken-doc= for (definition) source lookup.

It should use chicken-install to retrieve eggs to a source location
and be able to find definitions in both scheme and C sources of these
retrieved eggs.  Additionally to that it would be nice to have an
updating command and integration into the REPL.

See also henrietta-cache: https://wiki.call-cc.org/running-an-egg-mirror

Perhaps it's sufficient to add an option to that utility for fetching
the last n revisions (and cleaning older releases!), then one could
just run it periodically and maybe offer a tarball of it on a mirror...

http://wiki.call-cc.org/chicken-projects/egg-index-4.html
http://code.call-cc.org/svn/chicken-eggs/release/4/ (anonymous svn)
http://code.call-cc.org/svn/chicken-eggs/release/4/egg-locations
http://code.call-cc.org/release-info?egg=loop
http://www.more-magic.net/posts/vcs-independent-distribution.html
http://wiki.call-cc.org/chicken-projects/egg-rss-feed-5.xml

#+BEGIN_SRC
<C-Keen> wasamasa:  (toplevel-command 'doc (lambda () (repl-doc-dwim (read)))
<C-Keen>                     ",doc PATHSPEC     Describe identifier or path with chicken-doc"))
#+END_SRC

** Implementation

- =henrietta-cache -c ~/.usr/share/chicken/cache -e https://anonymous@code.call-cc.org/svn/chicken-eggs/release/4/egg-locations=

- Check whether last items are newest (otherwise implement version sort)
- Implement download-last-release-files (default to 1)
- Implement clean-cache (with option how many releases to keep,
  default to 1)
- Expose these through command-line arguments

- Write tool going through that cache and guessing what defines an
  exported identifier (typically it's a top-level (... IDENTIFIER ...) or
  (... (IDENTIFIER ...) ...) form, top-level usually being inside a module)

- How will this work on chicken-core itself?
- These appear to be part of library.scm...

* =moped=

A better MPD library!

* =doodle2=

[[http://wiki.call-cc.org/eggref/4/doodle][doodle]], but with [[http://libsdl.org/][SDL2]] (as it has numerous benefits, including the
windows behaving more like other toolkit windows with resizing and
whatnot).

https://wiki.libsdl.org/MigrationGuide

* Lispy grep tool

Basically a less snazzy [[http://mlsec.org/joern/][joern]] which parses lispy languages (with
support for archives by using the great [[http://libarchive.org/][libarchive]]), uses a query
language (maybe graph-style or something simpler?) and doesn't require
a database server.

- https://github.com/fullstackio/cq (very simple query language for js)
- https://github.com/hchasestevens/astpath (xpath for python)

An alternative to graph databases is using a triple store. Allegro is
a proprietary CL implementation of these, Redland is a free one with a
C API.

* Fuzzy matcher

That would be nice to have as basis for fuzzy completion inside Emacs,
though [[https://github.com/Fuco1/yafmp][yafmp]] might be good enough (as it is very new though it's still
got the classic plethora of C bugs...).

* Self-hosting [[http://wiki.call-cc.org/eggref/4/spock][spock]] and interpreter

#+BEGIN_SRC
<wasamasa> I've been looking at http://ceaude.twoticketsplease.de/js-lisps.html thanks to #scheme
<wasamasa> and was wondering whether spock got an interpreter by now
<DerGuteMoritz> wasamasa: doesn't look like it
<wasamasa> DerGuteMoritz: would it be a worthwhile goal to pursue?
<DerGuteMoritz> I think that would allow it to become self-hosted
<wasamasa> indeed
<DerGuteMoritz> well, not necessarily the only way
<wasamasa> something clojurescript hasn't managed yet :P
<DerGuteMoritz> but a worthwhile one
<DerGuteMoritz> heh
<wasamasa> the reason why I'm asking is because ever since getting started with emacs I really like starting out in an interpreter first
<wasamasa> which works well enough in `csi` (save some... interesting bugs)
<DerGuteMoritz> an on-the-fly compiler is also an option
<DerGuteMoritz> might be possible to conjure something like that for spock already
<wasamasa> so, if one would keep feeding lines of input into that compiler form, that might suffice?
<DerGuteMoritz> sure
<wasamasa> "eval is not supported."
<DerGuteMoritz> pseudo-js: eval(popen("chicken-spock", "(some-scheme-expression)").read())
<DerGuteMoritz> (also, pseudo chickne-spock interface, no idea whether it can be called like that)
<DerGuteMoritz> *chicken-spock
#+END_SRC

* Linter

- =csc -ASM stoll.scm=
- =csc -debug analyze-only stoll.scm=
- Either combine both kinds of output or figure out how to get line
  numbers into the former

http://wiki.call-cc.org/chicken-internal-structure#the-analysis-database

* Write a dedicated Emacs mode

See http://code.call-cc.org/cgi-bin/gitweb.cgi?p=chicken-core.git;a=blob;f=misc/chicken.el;hb=HEAD

* Pastebin

- paste.brause.cc
- put a file in a specific location (watched by inotify) and get it
  copied to an uniquely named file
- copy UI from ix.io/sprunge.us (minus the part of line numbers being copyable)
- allow highlighting with language via pygments
- allow highlighting line/range

* DONE inotify egg

- see https://github.com/xlevus/chicken-stuff/blob/master/inotify.scm
- learn how callbacks would work

* DONE GIF

- write giflib C code
- study other wrappers
- write giflib wrapper
- write imlib interface
- redo cssfuscator option parsing:
  http://api.call-cc.org/doc/getopt-long
  https://bitbucket.org/ckeen/pee/src/54dea32f4471a500d288579e9ac4596f51d92255/pee.scm?at=master&fileviewer=file-view-default

* DONE GUI

- https://github.com/vurtun/zahnrad
- this could be cool if it were usable from CHICKEN...
- felix would be interested in any findings and comparisons to, say,
  tcl/tk: http://www.call-with-current-continuation.org/ma

* ZNC

- I'd like having a queriable (SQLite) database
- One approach would be watching files grow and adding their new
  content to it
- The other one to just write a ZNC module logging new stuff into the
  database
- Either way I'd need to deal with old data first
- Once that's done, an UI for querying would be useful, maybe even
  something on the web which supports predefined queries (like,
  highlights, project requests, queries, work stuff, etc.)

* nginx traffic logs

- something is eating up a few gigabytes of transmitted traffic each month
- analysis of logs should allow reconstruction by figuring out file
  sizes and amounts and graphing these

* DONE SRFI-105 (infix syntax)

https://code.dieggsy.com/srfi-105/

* Interesting gazette episodes

** #9 - chicken-doc

http://gazette.call-cc.org/issues/9.html

** #11 - autocompile

http://gazette.call-cc.org/issues/11.html

** #16 - modules

http://gazette.call-cc.org/issues/16.html

** #17 - slime

http://gazette.call-cc.org/issues/17.html

** #18 - exception handling

http://gazette.call-cc.org/issues/18.html

* Parsing mboxes for fun and profit

ftp://lists.gnu.org/emacs-devel/
http://wiki.call-cc.org/eggref/4/mbox
http://userguide.icu-project.org/boundaryanalysis#TOC-Sentence-Boundary

Purpose: Getting random emacs-devel quotes or markov-chain quotes.

Problems:
- Removing blockquotes (RMS has a very unusual style), attachments, signatures
- Detecting sentences

Alternatively:

Search debbugs for reports with patches and check how many of these
are closed/open

https://lists.gnu.org/archive/html/bug-gnu-emacs/2016-02/threads.html

* Write a tool for c4 remote control

See http://autoc4/mate/ and https://github.com/Gordin/c4-mate

- List all the names
- Search for an user
- Alter your own balance
- Transfer money to someone else
(- Alter someone else's balance)

* Experiment with 9P as API

- http://wiki.call-cc.org/eggref/4/9p

* qq

Port bb to CHICKEN (bb as name is already taken by the fltk wrapper egg)

* MELPA

- Alternative MELPA interface
- Define an API for paginated package list, package details, etc.
- Make it fast
- Maybe extend it to all archives?

* MELPA Stable

** Design

- http://python3wos.appspot.com/
- http://py3readiness.org/

** Naming

- ms-nss.brause.cc - MELPA not-so-stable
- ms-wos.brause.cc - MELPA Stable - Wall of Shame
- ms-ready.brause.cc - MELPA Stable Readyness

** Idea

- Display MELPA Stable packages
- Highlight installable ones green and the others red
- Sortable and pagination and toggling dependency view and stuff
- A "MS Ready" badge to put on a package that can be safely installed?

* Twitter Bot

- Reuse SUPREMEGENTLEMAN (with Elliot Rodger + X?)
- Alternatively/additionally make one based on John[Lisbeth] and
  classic buddhist texts, maybe erowid
- Implement Markov
- https://github.com/n3mo/clucker/blob/master/clucker.scm <- missing writers
- https://dev.twitter.com/rest/reference/post/statuses/update
- http://wiki.call-cc.org/eggref/4/rest-bind

* Write my own text editor

** Names

- RAGE (RAGE Ain't GNU Emacs)
- ELLE (ELLE Looks Like Emacs)
- TEM (TEM exasperates me) <- add undertale easter egg

** Design

- Modal editing
- Entirely written in Scheme

** Runtime extensibility

This could be tricky (or not).  Time to write a PoC...

- load works on Scheme files, compiled libraries and input ports
  (which allows loading from a string)
- Write program that defines a function, allow redefining it at
  runtime, check whether redefined function behaves differently, time
  performance differences
- What about modules/namespaces/other separation?

** Basis

- SDL2, self-made widgets and pango?

* Write another text editor

- Goal: How well does 9p work for configuration?
- http://wiki.call-cc.org/eggref/4/9p

* DONE EPUB viewer

- solved with the webkit egg!
- feature: double page spread?
- feature: progress indicator (current chapter / total chapters)
- feature: vim-style navigation (by reimplementing existing navigation)
- feature: outline view (see ncx for that and transformation to
  suitable HTML)
- bug: flash of styled content (solvable by preprocessing)
- bug: sometimes the mimetypes file is unreadable due to permissions
  being wrong
- wishlist: write libzip binding for not depending on unzip, inspect
  basics for validation, extract files when everything is ok to /tmp

* Jamming station

A project with Vim-like keybindings for synthesizing and playing
music.  Ideally a custom thing written with SDL2 and its mixing
support.  Alternatively go for the direction of [[https://github.com/ctford/klangmeister][klangmeister]] and use
the webkit egg and webaudio...

- Synthesize some analog chip sound (SID?)
- Provide it as preset, maybe others (like the NES, GB, Genesis,
  whatever else chiptune artists use)
- Allow playing with that preset on the keyboard or via MIDI/OSC (for
  instrument/automation support)
- Write sdl2_mixer or [[http://libsound.io/][libsoundio]] or [[https://xiph.org/ao/doc/][libao]] bindings if no SDL2 (what else then,
  GLFW or maybe even one of my own GUI bindings?)
- Allow recording what you've played and sequencing your output (not
  sure whether sequencer/tracker-style...), ideally with Vim-style
  bindings if the interface is going to be textual

* Implement a C-style printf

- no idea about the name (tarmoff?)
- steal semantics from a simple libc, like musl
- check SRFI-28 and SRFI-48 (and eformat?) for implementation hints

* Porting old eggs

- aalib
- syntax-case (for fame?)
- eprintf/eformat

* Writing a meme generator with imlib2/cairo

* Writing an userbar generator with imlib2/cairo

http://www.userbars.be/resources
http://images.google.de/imgres?imgurl=http%3A%2F%2Ffindicons.com%2Ffiles%2Ficons%2F2807%2Fmetronome%2F128%2Ffirefox.png&imgrefurl=http%3A%2F%2Ffindicons.com%2Fsearch%2Ffirefox%2F5&h=128&w=128&tbnid=RWfMxjdzIvidKM%3A&docid=gfIDnCoopxMOVM&hl=en&ei=7MN1V43eGsvXgAb0iqLYBg&tbm=isch&iact=rc&uact=3&dur=533&page=1&start=0&ndsp=37&ved=0ahUKEwiNhLqmi9HNAhXLK8AKHXSFCGsQMwhjKCAwIA&safe=images&bih=865&biw=1440

* Write blog post on combining pango and cairo

https://dthompson.us/font-rendering-in-opengl-with-pango-and-cairo.html

* Write blog post on combining sdl2 and cairo

- first step: using common pixel format, then render with cairo on sdl2 texture
- <wasamasa> apparently both have rgb16_565/rgb565, rgb24 and argb32/argb8888
- https://www.cairographics.org/SDL/
- http://stackoverflow.com/questions/24316393/can-cairo-use-sdl-texture-as-a-render-target

* Write blog post on combining glfw3 and nanovg

- https://www.upyum.com/cgit.cgi/poulexagone/commit/?id=1567b1df26999275c48bc6e65669d2828f077cac

* Write a irregex visualizer

Not sure whether to console (with fmt?), a web thing (with spock?) or with emacs

* Blog post on REBOL and GUI

- rebol3 doesn't support it, red only supports windows, so rebol2 it is...
- downloaded from http://www.rebol.com/download-view.html
- requires lib32(!) stuff, including xt, xmu, xaw
- crashes with http://www.rebol.com/article/0545.html
- solved with installing xorg-fonts-type1
- to use these right away, use =xset +fp /usr/share/fonts/Type1= and
  verify with =xlsfonts | grep courier=
- alternatively, just reboot your xorg session and these should be found
- execute =rebol= and witness double-free crashes!
- alternatively, =./rebol -v= for console

* Port cool nuklear/dear imgui stuff

- https://github.com/fabioarnold/TwoTriangles

* NFO/ASCII/ANSI art viewer

http://sixteencolors.net/
http://pelulamu.net/unscii/

* Optimize blog source rendering throughput

Currently I'm using =rst2html= which requires pygments for syntax
highlighting.  Starting a process over and over again takes time
(TODO: benchmark whether that's really the bottleneck), perhaps a
better solution is needed

** DONE Run =rst2html= code with pyffi egg

Just forget about it, that thing isn't even close to bullet-proof.

** TODO Write a persistent process converting stuff

I've had some sort of success with communicating by sending a line and
receiving output, integrating that into hyde poses some challenges (as
it's sending the whole file contents, not just a file name).  In other
words, I need to think up a better protocol than that.

* NES emulator

** Names

- boNkErS
- erNESto
- magNESium
- miNEStrone
- NEScaline
- NEStination
- noNsEnSe
- TripleNES

* GB emulator

** Names

- buGBear
- GumBall
- oGBo
- raGBag

* Programming in Scheme [0/9]

** TODO Chapter 4: Making Choices [0/1]

- [ ] days alive

** TODO Chapter 5: Recursion [0/1]

- [ ] random walk

** TODO Chapter 7: Sample Projects II [4/8]

- [ ] Color mode
- [ ] B/W mode
- [ ] Color bias
- [ ] Different formulas
- [X] Batch drawing
- [X] Interactive mode
- [X] Combined program doing both
- [X] Screenshots in interactive mode creating new files

Exercise text:

5. Try other formulas instead of $x^2 + y^2$ as the basis of your
   test. What happens if you try, say, $x^2 - y^2$? Or $x^3 + y^3$? Or
   $x^2 + y^2 + |x| + |y|$?
6. Change the =choose-pen-color= procedure so that it's more likely to
   choose black (as suggested earlier). Suppose, for example, that you
   truncate $x^2 + y^2$ and take the remainder when divided by 7
   instead of 4; and if the result is 3 or less the pen color will be
   set to 0 (if the result is 4, 5, or 6 we set the pen color to 1, 2,
   or 3, respectively). How does this change the pictures drawn?

** TODO Chapter 8: Pairs, Lists and Symbols [0/3]

- [ ] molecular weight
- [ ] DNA translator
- [ ] sentence generator

** TODO Chapter 10: Sample Projects III [0/1]

- [ ] Who's on first?

** TODO Chapter 12: Procedures [0/3]

- [ ] deal cards
- [ ] lissajous plotter
- [ ] cryptograms

** TODO Chapter 13: Altering Bindings, Altering Projects [0/6]

- [ ] turtle
- [ ] hangman
- [ ] c-curves
- [ ] dragon-curves
- [ ] slot-machine
- [ ] hotel

** TODO Chapter 15: Sample Projects IV [0/1]

- [ ] pattern matching

** TODO Chapter 16: A Sampler of Advanced Topics [0/2]

- [ ] solitaire
- [ ] calendar

* DONE Git selfhosting

- Bare repository with index.html and some extra git files
- https://gist.github.com/joahking/780877
- https://mirrors.edge.kernel.org/pub/software/scm/git/docs/user-manual.html#public-repositories

* TODO Image optimization for manga

Several interesting constraints here:

- Detect whether optimization is worth it
- Estimate how much could be saved
- Use below settings for conversion (via IM? OIIO?)

https://mangadex.org/thread/7214/1/#post_1228151
https://www.insidescanlation.com/etc/how-to-edit-manga-with-photoshop/rawr/editguide/3