The Orville star and creator Seth MacFarlane dropped a bomb at San Diego Comic Con – the show will move off of Fox onto the Hulu streaming network for season 3. In a statement MacFarlane said, “The Orville has been a labor of love for me, and there are two companies which have supported that […]

Don't expect a pack of running fanatics swarming to the finish line at road races this year. But that doesn't mean that participants don't have options.

The IOC Session -- an annual meeting of approximately 100 members -- will be held in July via a video conference rather than the originally scheduled gathering in Tokyo prior to the Summer Olympics.

Říká se, že po krizi spojené s onemocnění covid-19 se probudíme do jiného světa. Otázka je, zda to bude platit i pro realitní trh. Podle odborníků s největší pravděpodobností ano. Což platí jak pro prodejní ceny nemovitostí, tak pro ceny nájemního bydlení.

Sněmovna schválila návrh zákona o některých opatřeních ke zmírnění dopadů epidemie koronaviru. Jaké úlevy se dočkají dlužníci a jaké jsou podmínky pro zastavení exekuce, vysvětluje právník Adam Stawaritsch.

V době koronavirové pandemie nás bombardují ze všech stran nejrůznější informace. Informační přetížení je tak značné, že v mnohých lidech vyvolává strach a stres. Jak tomu čelit, komu věřit a jaké informace ignorovat, se v komentáři zamýšlí psycholog Jan Urban.

Současná koronavirová krize otřásla finančními trhy, zlevněné akcie jsou příležitostí, jak zhodnotit investici. Jenomže kdy je správná chvíle investovat? Už se trhy odrazily ode dna, nebo přijde další propad? A jak snížit riziko špatného načasování? Nad tím se zamýšlí Michal Valentík, investiční analytik společnosti Broker Trust.

Do pracovních i osobních životů vstoupila pandemie, jakou nikdo nepamatuje. Krize, která nastává, se dotkne všech. Nad aktuální situací a tím, co nás letos čeká a co je klíčové pro rozhýbání ekonomiky, se v komentáři zamýšlí personalista Tomáš Surka.

Mohlo by se zdát, že pojišťovny budou v koronavirové době zaznamenávat zvýšený zájem o sjednání životních pojistek. Zejména kvůli krytí rizika pracovní neschopnosti, což se vztahuje i na onemocnění infekcí covid-19. Praxe je ale zatím opačná.

Práci z domova u zaměstnanců soukromých firem nelze jednostranně nařídit, vzniká jen na základě dohody obou stran. U pedagogických pracovníků a zaměstnanců škol však platí trochu jiná pravidla. Ve spolupráci s právníky Bořivojem Líbalem a Markem Polonim přinášíme odpovědi na nejčastější dotazy pedagogů.

Как меняется политическая повестка во время эпидемии? Сколько денег нужно на поддержку граждан и бизнеса? Какие активы падают, а какие, наоборот, растут? Ольга Орлова обсудила эти животрепещущие вопросы с экономистом Александром Аузаном в программе «Гамбургский счет» на Общественном телевидении России. Публикуем отредактированную расшифровку беседы.

Каким образом вирус SARS-CoV-2 корректирует ошибки при транскрипции своего генома? Как эта способность связана с безуспешностью клинических испытаний ремдесивира? И можно ли считать эти испытания окончательно проваленными? На вопросы Юлии Черной отвечает Дмитрий Жарков, член-корр. РАН, директор Центра перспективных биомедицинских исследований НГУ, зав. лабораторией геномной и белковой инженерии Института химической биологии и фундаментальной медицины СО РАН.

• 1 avocado, chopped ($1.25)
• 1tablespoon finely chopped celery ($.05)
• 1 tablespoon finely chopped onion ($.05)
• 2 teaspoons mustard ($.05)
• 2 teaspoons mayo  (use something like Veganaise or make your own) ($.10)
• 1/2 teaspoon salt
• 1/2 teaspoon pepper
• 1/2 teaspoon paprika
• Lettuce and optional tomato for serving

• Coarsely chop the avocado, add the remaining ingredients, then stir gently until evenly incorporated. 
• Store leftovers for up to a day or two in an air tight container.
•  

Normal life up to February or March 2020 was clearly the Before Times. (I like that better than calling it BC for 'before corona', because the latter seems slightly offensive to Christians, and Before Times is ironic but immediately transparent.)

ironed_orchid pointed out that the period we're in now is clearly the Time of Isolation.

So what are we going to call the post-Covid future? Maybe we'll never really be post-Covid, any more than we are post-flu or post-TB, but I am daring to hope for a time when it's not the dominant feature of everybody's lives. Anyone heard a term as memorable and fitting as Before Times and Time of Isolation?

comments

Well, here we are on the eve of SIGBOVIK 2019. I'm in the midst of a long day of video-making for one of my projects, but I can get 2 for the price of 5 by posting now about four of my papers. This year I've been on a chess kick, which I think I've successfully gotten out of my system by writing all these (previous posts alluded to there being five, but one of them didn't really go anywhere and/or just became part of the other(s)). They are sort of intertwined:

Survival in chessland is about how to stay alive if you are being a chesspiece to the death

Color- and piece-blind chess is about, among other things, playing chess without being able to tell what the pieces are (only where they are)

Elo World, a framework for benchmarking weak chess engines is about exploring the full spectrum of computer chess play

CVE-2018-90017117 #KingMe is just a short joke, but based on a true story


My last paper is on a different (maybe even weirder?) topic, and I'm putting together a video for it now, so I should be uploading that tomorrow some time. It's been a bit rough going, though, since I replaced my computer a few months ago and forgot that I hadn't actually set stuff up for this kind of work; I'm experiencing small problems like custom key commands aren't set, and bigger problems like audio drivers acting crazy. Looks like I will be able to finish with some vacation time, at least.


Speaking of vacation, this month we also went to Belize, which was pretty cool. The highlight for me was swimming/scrambling 1km into a cave ("Actun Tunichil Muknal") to access an approximately 1000 year-old Mayan site where they performed human sacrifices; it's remarkable because almost all of the artifacts are still in situ, including a number of calcified human skeletons. Was pretty wild. I got some good running done, found some New Haven-style pizza (!?), and wrote papers about chess (?!).

Well, April felt simultaneously short and long! I should have just posted these at the beginning of the month, my SIGBOVIK papers from 2020:

Is this the longest chess game? is another needless chess paper, here trying to figure out the longest possible legal game. There are several rules that make sure games can't go on forever, and some surprisingly subtle details/ambiguity to those rules. The whole game is of course included in the paper (17697 moves), but I was far from being the largest waste of space in this year's proceedings, as one provocateur had a paper with 150 pages of citations. Mathieu made a 5-hour video of the chess game I computed for his companion blog post.

What is the best game console? A market-based approach is a silly idea taken too far. It was a year in the making (mostly waiting) and didn't quite turn out the way I was expecting due to world events, but that's part of the "fun" I guess!

Conditional Move For Shell Script Acceleration was another collaboration with Jim (mostly his doing, but I like to lather on an additional patina of absurdity).

This month I have mostly been trying to keep sane and healthy during the shelter-in-place order. It's been harder than usual to find the energy to be creative, but I have had some spurts. I basically only leave the house to run (not going anywhere near other people). But I have been doing that pretty regularly, so between that and the prohibition against going out to bars and ice cream, I'd say I'm currently in the best I have been in ~6 years. Yesterday I claimed some course records for some Strava segments in my neighborhood! I also finished up Doom: Eternal, which was good but you pretty much already know what it's like and I'm playing Animal Crossing and haven't yet gotten sick of that. The timing for the release of that latter game couldn't have been more perfect, huh? Sometimes I need something with a little challenge, so I just started Nuclear Throne. I'm liking it but not sure if I have decided whether it's good enough to invest the time in to win (I almost always play games to the end but these randomized roguelikes demand a certain kind of potentially infinite investment. Like I never did beat the last boss in Wizard of Legend, and even in Dead Cells, which I loved, I had to settle for some modest personal criteria for "winning.") Any other recs? Could use a good Metroidvania perhaps?

(in-package #:cl-pestilence)

;;   This is _not_ a simulation. It's just a game. And any resemblance
;; to any world, real or imaginary, is entirely coincidental.

;;   You can copy/paste this post in its entirety into a Common Lisp
;; REPL and play around with it if you like. I'm documenting it where
;; possible, but it's just a small toy to poke at for the moment.

;;   I've been thinking a lot about asymmetric multiplayer games and
;; <gestures wildly to world at large> all this.
;; I'm not actively _trying_ to model it accurately, but it's probably
;; obvious what's been consuming my thoughts lately.

;;   Let's get right into this. I'll explain as I go, and tie a few things
;; together neatly at the end. I hope. Regardless, there will absolutely
;; be a repo sometime fairly soon.

;; A place can be tagged arbitrarily, and can contain occupants.
;; They also collect points.

(defclass place ()
  ((tags :initarg :tags :initform nil :accessor tags)
   (occupants :initarg :occupants :initform nil :accessor occupants)
   (points :initform 0 :accessor points)))

(defun place? (thing)
  (eq (find-class 'place) (class-of thing)))

(defun place (&key tags occupants)
  (make-instance 'place :tags tags :occupants occupants))

(defun gen-place ()
  (let ((tag (pick '(:apartment-building :house :cottage
		     :office-building :factory :store
		     :cafe :lounge :theater))))
    (place :tags (list tag))))

(defmethod details ((place place))
  (format nil "====================~%~a {~{~a~}}~%~{  ~a~^~%~}~%"
	  (first (tags place))
	  (rest (tags place))
	  (mapcar #'details (occupants place))))

(defmethod show ((place place))
  (format nil "~20@a ~5a [~{~a~}]~%"
	  (first (tags place)) (points place)
	  (mapcar #'show (occupants place))))

;; A peep goes places.
;; They have
;;  - their daily routine (a list of places to visit)
;;  - their todo (the part of their routine they still need to do;
;;                they are currently at the first place in this list)
;;  - their health (a number from 0 to 100)
;;  - a list of plagues
;; Finally, they _also_ collect points.

(defclass peep ()
  ((routine :initarg :routine :initform (list) :accessor routine)
   (todo :initarg :todo :initform nil :accessor todo)
   (health :initarg :health :initform 100 :accessor health)
   (plagues :initform nil :accessor plagues)
   (points :initform 0 :accessor points)))

(defun peep? (thing)
  (eq (find-class 'peep) (class-of thing)))

(defun peep (&key places)
  (make-instance 'peep :routine places :todo places))

(defun health->string (health)
  (cond ((>= health 90) "@")
	((>= health 80) "0")
	((>= health 70) "O")
	((>= health 50) "o")
	((>= health 30) ":")
	((>= health 1)  ".")
	(t "☠")))

(defmethod details ((peep peep))
  (format nil "[~a ~3d [~{ ~a~^ ->~}]]"
	  (health->string (health peep)) (health peep)
	  (mapcar
	   (lambda (place) (first (tags place)))
	   (routine peep))))

(defmethod show ((peep peep)) (health->string (health peep)))

;; A world is a list of places, occupied by peeps. The world we start
;; peeps in also determines their routine.

(defun gen-world (&key (num-places 20) (num-peeps 100))
  (let ((places (loop repeat num-places collect (gen-place))))
    (loop repeat num-peeps
       do (let* ((routine (loop repeat 5 collect (pick places)))
		 (peep (peep :places routine)))
	    (push peep (occupants (first routine)))))
    places))

(defmethod details ((world list))
  (format nil "~%~{~a~}~%" (mapcar #'details world)))

(defmethod show ((world list))
  (format nil "~%~{~a~}~%" (mapcar #'show world)))

(defmethod all-peeps ((world list))
  (loop for place in world append (all-peeps place)))

(defmethod all-peeps ((place place))
  (loop for o in (occupants place) if (peep? o) collect o))

;; `tick!`ing a world means moving every peep through their routine once.
;;   We `tick!` each peep, then `tick!` each place until all the peeps are
;; done. Then we reset their routines.
;; You can think of this as a turn in the game.

(defmethod tick! ((world list))
  (let ((peeps (all-peeps world)))
    (loop while peeps
       do (setf peeps
		(loop for p = (pop peeps) while p
		   for res = (tick! p)
		   if res collect res))
       do (mapc #'tick! world)
       do (format t "~a" (show world)))
    (loop for p in (all-peeps world)
       do (setf (todo p) (routine p))))
  world)

;; Don't worry about the details of how to `tick!` peeps or places yet.

;;   Ok, here's where it gets a bit darker. Although we _did_
;; foreshadow this in the definition of `peep`. And also in the title
;; of the accompanying blog post.

;; A plague is another living thing.
;; It has
;;  - a host (a peep that it's infecting)
;;  - a signature (a token representing its lineage and strain)
;;  - health (how well it's doing inside its host)
;;  - virulence (how likely it is to spread to another host)
;;  - efficiency (how efficient they are at feeding)
;;  - reproduce (a function that returns a new instance to push into a new host)
;;  - and a strategy (a function, possibly closed, that takes
;;    itself and its host peep and mutates)

;; Plagues do not collect points; they score differently.

(defclass plague ()
  ((host :initarg :host :initform nil :accessor host)
   (signature :initarg :host :initform "SIG" :accessor signature)
   (health :initarg :health :initform 10 :accessor health)
   (virulence :initarg :virulence :initform 10 :accessor virulence)
   (efficiency :initarg :efficiency :initform 0.2 :accessor efficiency)
   (reproduce
    :initarg :reproduce
    :initform
    #'plague
    :reader reproduce)
   (strategy
    :initarg :strategy
    :initform
    (lambda (plague peep)
      (feed! plague peep 30))
    :reader strategy)))

(defun plague ()
  (make-instance 'plague))

;; Plagues can `feed!` on peeps or plagues. To feed means to
;; take away some of the targets' health and add some to your own.

(defmethod feed! ((self plague) (peep peep) (amount integer))
  (decf (health peep) amount)
  (incf (health self) (* (efficiency self) amount)))

(defmethod feed! ((self plague) (plague plague) (amount integer))
  (decf (health plague) amount)
  (incf (health self) (* (efficiency self) amount)))

;; Plagues can also `infect!` peeps by `reproduce`ing into them.

(defmethod infect! ((self plague) (peep peep))
  (unless (infected-by? self peep)
    (let ((child (funcall (reproduce self))))
      (setf (host child) peep)
      (push child (plagues peep)))))

(defmethod infected-by? ((self plague) (peep peep))
  (member (signature self) (mapcar #'signature (plagues peep))
	  :test #'string=))

;;  `tick!`ing a plague causes it to weaken and also carry out its strategy.
;; This models the background effect of the immune system of its host.

(defmethod tick! ((plague plague))
  (decf (health plague) 1)
  (funcall (strategy plague) plague (host plague))
  plague)

;;  `tick!`ing a peep means moving them to their next place, and also
;; `tick!`ing any plagues they may have contracted. Also, peeps are
;; resilient; they heal a small amount each time they tick (to a
;; maximum of 100).
;;  If a peep dies, they no longer move. And their plagues probably
;; won't do well. Peeps like to go places. They score points for each
;; place they go to.

(defun dead? (thing) (>= 0 (health thing)))

(defmethod tick! ((peep peep))
  (unless (dead? peep)
    (let ((location (pop (todo peep))))
      (incf (points peep))
      (setf (occupants location) (remove peep (occupants location)))
      (push peep (occupants (or (first (todo peep)) (first (routine peep)))))
      (setf (health peep) (min 100 (+ 5 (health peep))))
      (mapc #'tick! (plagues peep))
      (unless (empty? (todo peep))
	peep))))

;; `tick!`ing a place causes it to score for each `peep` present. And it causes
;; any `plague`s on present `peep`s to try to `infect!` other nearby peeps.
;; Places also lose points for each dead peep they contain.

(defmethod tick! ((place place))
  (incf (points place) (length (occupants place)))
  (loop for peep in (all-peeps place)
     if (dead? peep)
     do (decf (points place) 2)
     else do (loop for plague in (plagues peep)
		do (loop for victim in (remove peep (all-peeps place))
		      if (>= (virulence plague) (random 100))
		      do (infect! plague victim))))
  place)

;;  So, now we've got the basic framework of the game in place. There are three
;; players in this game: places, peeps and plagues.
;;   A plague player automatically loses if they are completely cured, and
;; automatically wins if they manage to kill everyone. That's fairly simple.
;;   A place player wins if they manage to cure the plague. They automatically
;; lose if all the peeps die. Also, fairly simple.
;;   A peep player is trying to survive. If they manage to make it some numer
;; of turns before dying, then we have to score the game instead of declaring
;; an outright winner regardless of game state.

;;   A peep player's score is the total number of points plus remaining health
;; on all of their peeps, minus the number of active plagues on said peeps.
;;   A plague player's score is the total number of health of their plagues,
;; with a multiplier equal to the number of places fully infected by
;; their plague.
;;   A place player's score is the total number of points in their places.

(defun score (world)
  (list :peep (let ((score 0))
		(loop for p in (all-peeps world)
		   unless (dead? p)
		     do (incf score (+ (health p) (points p)))
		   do (decf score (length (plagues p))))
		score)
	:place (let ((score 0))
		 (loop for p in world
		    do (incf score (points p)))
		 score)
	:plague (let ((score 0))
		  (loop for victim in (all-peeps world)
		     do (loop for p in (plaguesvictim)
			   do (incf score (max 0 (health p)))))
		  (loop for target in world
		     if (every
			 (lambda (victim)
			   (not (empty? (plagues victim))))
			 (all-peeps target))
		     do (setf score (* 2  score)))
		  score)))

;;   I think that's all I've got for now. This is definitely an idea I want
;; to run with. At the moment, it's just a tiny, in-repl proof-of-concept,
;; and not particularly fun, but I'm going to try developing it further with an
;; eye towards turning it into an actual web game playable from this site.

;; As always, I'll let you know how it goes.

(defun pick (lst)
  (nth (random (length lst)) lst))

(defun empty? (lst)
  (null lst))

As any Common Lisp coder knows, a REPL is an incredibly useful tool. It can be used not just for development, but for running all sorts of tasks. Personally, I don't bother making my Lisp tools into executable scripts and just run them directly from SLIME. As such, any operation that requires leaving the REPL is quite inconvenient. For me, one such operation was viewing image files, for example in conjunction with my match-client:match tool. So lately I've been researching various methods to incorporate this functionality into the normal REPL workflow. Below, I present 3 methods that can be used to achieve this.

Open in external program

This one's easy. When you want to view a file, launch an external process with your favorite image viewer. On Windows a shell command consisting of the image filename would launch the associated application, on Linux it's necessary to provide the name of the image viewer.

(defvar *image-app* nil) ;; set it to '("eog") or something

(defun view-file-native (file)
  (let ((ns (uiop:native-namestring file)))
    (uiop:launch-program (if *image-app*
                             (append *image-app* (list ns))
                             (uiop:escape-shell-token ns)))))

Note that uiop:launch-program is used instead of uiop:run-program. The difference is that launch- is non-blocking - you can continue to work in your REPL while the image is displayed, whereas run- will not return until you close the image viewer.

Also note that when the first argument to run/launch-program is a string, it is not escaped, so I have to do it manually. And if the first argument is a list, it must be a program and a list of its arguments, so merely using (list ns) wouldn’t work on Windows.

Inline image in REPL

The disadvantage of the previous method is that the external program might steal focus, appear on top of your REPL and disrupt your workflow. And it's well known that Emacs can do everything, including viewing images, so why not use that?

In fact, SLIME has a plugin specifically for displaying images in REPL, slime-media. However it’s difficult to find any information on how to use it. Eventually I figured out that SWANK (SLIME’s CL backend) needs to send an event :write-image with appropriate arguments and slime-media's handler will display it right in the REPL. The easiest way is to just send the file path. The second argument is the resulting image's string value. If you copy-paste (sorry, "kill-yank") it in the repl, it would act just like if you typed this string.

(swank::send-to-emacs '(:write-image "/path/to/test.png" "test"))

You can even send raw image data using this method. I don't have anything on hand to generate raw image data so here's some code that reads from a file, converts it to a base64 string and sends it over SWANK.

(with-open-file (in "/path/to/test.png" :direction :input  :element-type '(unsigned-byte 8))
                (let* ((arr (make-array (file-length in) :element-type '(unsigned-byte 8)))
                       (b64 (progn (read-sequence arr in) (cl-base64:usb8-array-to-base64-string arr))))
                  (swank::send-to-emacs `(:write-image ((:data ,b64 :type swank-io-package::png)) "12345"))))

Note that the first argument to :write-image must be a list with a single element, which is itself a plist containing :data and :type keys. :data must be a base64-encoded raw image data. :type must be a symbol in swank-io-package. It’s not exactly convenient, so if you’re going to use this functionality a helper function/macro might be necessary.

Image in a SLIME popup buffer

Inline images are not always convenient. They can’t be resized, and will take up as much space as is necessary to display them. Meanwhile EMACS itself has a built-in image viewer (image-mode) which can fit images to width or height of a buffer. And SLIME has a concept of a “popup buffer” which is for example used by macroexpander (C-c C-m) to display the result of a macro expansion in a separate window.

Interestingly, slime-media.el defines an event :popup-buffer but it seems impossible to trigger it from SWANK. It is however a useful code reference for how to create the popup buffer in ELisp. This time we won’t bother with “events” and just straight up execute some ELisp code using swank::eval-in-emacs. However by default, this feature is disabled on Emacs-side, so you’ll have to set Emacs variable slime-enable-evaluate-in-emacs to t in order for this method to work.

Also Emacs must be compiled with ImageMagick for the resizing functionality to work.

Anyway, the code to view file in the popup buffer looks like this:

(defun view-file-slime (file &key (bufname "*image-viewer*"))
  (let ((ns (namestring file)))
    (swank::eval-in-emacs
     `(progn
        (slime-with-popup-buffer (,bufname :connection t :package t)
          (insert-image (create-image ,ns))
          (image-mode)
          (setf buffer-file-name ,ns)
          (not-modified)
          (image-toggle-display-image))
        ;; try to resize the image after the buffer is displayed
        (with-current-buffer ,bufname (image-toggle-display-image))))))
    ))

Arriving to this solution has required reading image-mode’s source code to understand what exactly makes image-mode behave just like if the image file was opened in Emacs via C-x C-f. First off, image-mode can be a major and a minor mode - and the minor mode is not nearly as useful. slime-with-popup-buffer has a :mode keyword argument but it would cause image-mode to be set before the image is inserted, and it will be a minor mode in this case! Therefore (image-mode) must be called after insert-image.

Next, the buffer must satisfy several conditions in order to get image data from the filename and not from the buffer itself. Technically it shouldn’t be necessary, but I couldn’t get auto resizing to work when data-p is true. So I set buffer-file-name to image’s filename and set not-modified flag on.

Next, image-toggle-display-image is called to possibly resize the image according to image-mode settings. It's called outside of slime-with-popup-buffer for the following reason: the buffer might not yet be visible and have any specific dimensions assigned to it, and therefore resizing will do nothing.

Here’s an example of how calling this function looks in Emacs.

The position of the popup buffer depends on whether the original Emacs window is wide enough or not. I think it looks better when it’s divided vertically. Use M-x image-transform-fit-to-height or M-x image-transform-fit-to-width to set up the auto-resizing method (it gets remembered for future images). Unfortunately there’s no way to fit both height and width, at least with vanilla Emacs. I prefer fit-to-width because in case the image is too tall, it is possible to scroll the image vertically with M-PgDn and M-PgUp from the other buffer. Unlike other image-mode buffers, this buffer supports a shortcut q to close itself, as well as various SLIME shortcuts, for example C-c C-z to return to the REPL.

That's it for now, hope you enjoyed this overview and if you happen to know a better way to display images in Emacs, I would be interested to hear about it.

So recently, I had to use zippers at work. Specifically, the Clojure implementation. There were some close-to-arbitrary transformations I needed to do with some close-to-arbitrary trees and it turned out that zippers were more efficient than the alternatives¹.

Using them this way, combined with the general state of the world and my free time, finally tipped me into doing some more Common Lisp development. Before, I go any further, let me be clear about something.

I Like Clojure

Seriously.

Its logo is up top in the language bar, I was one of the inaugural members of the Toronto Clojure User Group, I recommend it as a first lisp you should learn, and have for about six years now. I'm also painfully aware of the shortcomings of Common Lisp, and make no excuses for them.

However.

• I don't like the JVM. It's slow as balls, its' deployment options are less than ideal for my purposes, its' error system is at best useless, and Clojure without it is unlikely.
• Clojurescript build incompatiblities are, if anything, worse².
• I don't like the underlying licensing decisions.

These are deep reasons to stay away. They're not the sort of thing I can paper over with a library or two. Fixing them would mean a superhuman amount of work poured into the underlying technical and social infrastructure, and I'm not into it. I wouldn't be into it even if the community was interested in heading that way, and near as I can tell, they're not particularly.

Whether or not I think you should learn Clojure as your first³ lisp, it definitely wasn't my first lisp. The more uniform, mostly-better-thought-out interface, lack of historical baggage and functional data structures are not enough to pull me all the way over.

It is enough for me to start plotting a smash-and-grab of as much of the stuff I like as I can carry. Which is exactly what clj represents. As of this writing, it defines and exports exactly four symbols: if-let, when-let, -> and ->>. This is a tiny beginning of the list, and I fully plan to put something more substantial together using cl-hamt, named-readtables, test-utils and possibly optima. Stay tuned to that repo if you're interested, but it's not the focus today.

cl-zipper

The thing that percipitated this thought was having used the Clojure Zipper implementation. So, obviously, this is something I want next time I need to manipulate trees in Common Lisp. The paper is here, and unless you have a terminal phobia of datastructures⁴, you should go read it. It's six pages, they're light, and one of them taken up by the intro and references.

The operations defined in the paper are left, right, up, down, insert_right, insert_left, insert_down and delete. There's a few conveniences defined for the Clojure version, and I've implemented some of my own stuff too. Lets go through the main file in almost-literate style.

First up, we have constructors.

(defstruct path
  (left) (path) (right))

(defstruct loc
  (node)
  (path)

  (fn-branch?)
  (fn-children)
  (fn-make-node))

;;;;;;;;;; Constructors
(defun zipper (branch? children make-node root)
  (make-loc
   :node root
   :fn-branch? branch? :fn-children children :fn-make-node make-node))

(defmethod make-zipper ((thing list))
  (zipper #'listp #'identity (lambda (node children) (declare (ignore node)) children) thing))

(defun make-node (zipper children)
  (funcall (loc-fn-make-node zipper) zipper children))

You can see influence from both clojure.zip and the paper here. I'm taking the lead from the paper by explicitly separating the path triple our from the loc definition. However, I'm not explicitly defining my own type tree the way that Huet does. Instead, I'm going to be dealing with assorted lisp trees. These could be implemented as lists, vectors, hashes, or any number of other formats. I'm going to implement a few type-distpatching built-ins, including the make-zipper list method above, but the basic zipper function just needs to take an interface as input in the form of branch?, children and make-node arguments. This is the same solution that the Clojure implementation went with, and I see no reason to go a different way. The only material difference is that theirs uses the Clojure metadata system, while I explicitly define slots in the loc structure.

Now that we can construct, we need to be able to select.

;;;;;;;;;; Selectors
(defun branch? (zipper) (funcall (loc-fn-branch? zipper) (loc-node zipper)))
(defun children (zipper)
  (funcall
   (loc-fn-children zipper)
   (loc-node zipper)))
(defun node (zipper) (loc-node zipper))
(defun path (zipper) (loc-path zipper))

(defun lefts (zipper)
  (when (loc-path zipper)
    (reverse (path-left (loc-path zipper)))))

(defun rights (zipper)
  (when (loc-path zipper)
    (path-right (loc-path zipper))))

The basic navigation is four functions; down, up, left and right

;;;;;;;;;; Navigation
;;;;;;;;;;;;;;; Basic navigation
(defun down (zipper)
  (when (children zipper)
    (let ((fresh (copy-loc zipper)))
      (setf (loc-node fresh) (first (children zipper))
	    (loc-path fresh)
	    (make-path
	     :left nil
	     :path (loc-path zipper)
	     :right (rest (children zipper))))
      fresh)))

(defun up (zipper)
  (when (path zipper)
    (let ((fresh (copy-loc zipper)))
      (setf (loc-node fresh)
	    (make-node
	     zipper (append
		     (reverse (path-left (path zipper)))
		     (cons (loc-node zipper)
			   (path-right (path zipper)))))
	    (loc-path fresh) (path-path (path zipper)))
      fresh)))

(defun left (zipper)
  (when (and (path zipper) (path-left (path zipper)))
    (let ((fresh (copy-loc zipper)))
      (setf (loc-node fresh) (first (path-left (path zipper)))
	    (loc-path fresh)
	    (make-path
	     :left (rest (path-left (path zipper)))
	     :path (path-path (path zipper))
	     :right (cons (loc-node zipper) (path-right (path zipper)))))
      fresh)))

(defun right (zipper)
  (when (and (path zipper) (path-right (path zipper)))
    (let ((fresh (copy-loc zipper)))
      (setf (loc-node fresh) (first (path-right (path zipper)))
	    (loc-path fresh)
	    (make-path
	     :left (cons (loc-node zipper) (path-left (path zipper)))
	     :path (path-path (path zipper))
	     :right (rest (path-right (path zipper)))))
      fresh)))

The main difference between this and the paper is that I've chosen nil as my Top representation, which lets me pull the trick of using when to check for the presence of a path, and its' non-Top-ness at the same time.

The bad news is that since Common Lisp doesn't have pervasive functional data structures, I have to explicitly copy locs while moving through a tree. The good news is that the copy is fairly light weight. Effectively, I'm copying out a set of 5 pointers, and could get that down to 3 by defining an intermediate struct.

Hm.

Which I probably should do. Note to self.

Out of those, we get three compound navigation functions. With more probably coming soon. Specifically, I found find useful for the work I did. It's easily externally definable, but would be even easier to bundle along. The ones I've already implemented are root, leftmost and rightmost.

;;;;;;;;;;;;;;; Compound navigation
(defun root (zipper)
  (if-let (z (while zipper #'up))
    (node z)))

(defun leftmost (zipper) (while zipper #'left))

(defun rightmost (zipper) (while zipper #'right))

...
(defun until (zipper f)
  (let ((z zipper))
    (loop for next = (funcall f z) while next
       when next do (setf z next))
    z))
...

That's the traversals done. Next up, we've got modification, without which this library is fairly useless. The basics are replace, delete and the insert/child twins.

;;;;;;;;;; Modification
(defun replace (zipper node)
  (let ((fresh (copy-loc zipper)))
    (setf (loc-node fresh) node)
    fresh))

(defun delete (zipper)
  (when (path zipper)
    (let ((fresh (copy-loc zipper))
	  (fresh-path (copy-path (loc-path zipper))))
      (cond ((rights zipper)
	     (setf (loc-node fresh) (pop (path-right fresh-path))
		   (loc-path fresh) fresh-path))
	    ((lefts zipper)
	     (setf (loc-node fresh) (pop (path-left fresh-path))
		   (loc-path fresh) fresh-path))
	    (t (setf (loc-path fresh) (path-path fresh-path))))
      fresh)))

(defun insert-child (zipper node)
  (replace
   zipper
   (make-node
    zipper
    (cond ((not (branch? zipper))
	   (list node (node zipper)))
	  ((children zipper)
	   (cons node (children zipper)))
	  (t (list node))))))

(defun append-child (zipper node)
  (replace
   zipper
   (make-node
    zipper
    (cond ((not (branch? zipper))
	   (list (node zipper) node))
	  ((children zipper)
	   (append (children zipper) (list node)))
	  (t (list node))))))

(defun insert-left (zipper node)
  (let ((fresh (copy-loc zipper))
	(fresh-path (copy-path (loc-path zipper))))
    (push node (path-left fresh-path))
    (setf (loc-path fresh) fresh-path)
    fresh))

(defun insert-right (zipper node)
  (let ((fresh (copy-loc zipper))
	(fresh-path (copy-path (loc-path zipper))))
    (push node (path-right fresh-path))
    (setf (loc-path fresh) fresh-path)
    fresh))

The paper defines an insert_down function. It fails on a Leaf node, and otherwise inserts a singleton branch at the given location. The insert/append child functions above also insert nodes at a lower level at the current loc. They give you a choice about whether to insert the new node as the leftmost or rightmost child, and additionally succeed on Leaf nodes by including the leaf value as a child of the new branch.

There are, thus far, three compound modification functions; edit, splice-left and splice-right.

(defun edit (zipper f &rest args)
  (replace zipper (apply f (node zipper) args)))

(defun splice-left (zipper node-list)
  (reduce #'insert-left node-list :initial-value zipper))

(defun splice-right (zipper node-list)
  (reduce #'insert-right (reverse node-list) :initial-value zipper))

edit takes a function instead of a new node, and replaces the node at loc with the result of running that function on the existing node. The splice-* twins are fairly self-explanatory; they're like insert-left/insert-right, but work on multiple nodes rather than single ones.

I haven't yet implemented next, prev and remove because these might relate to the different representation of the traversal end? state. The reason for this seems to be that next/prev/remove assume a depth-first traversal. The reason I'm being weasely here is that I haven't thought about it hard enough to be sure that the end? marker is really necessary. It also seems odd to privilege depth-first over breadth-first traversals; ideally, I think you'd want to be able to support either. Possibly interchangeably.

Minor Housekeeping

That wraps it up for this edition. My immediate intention is to do more work on the cl-zipper and clj libraries, as well as that game I mentioned last time. Ideally, I'd like to up my blogging output too. Probably not to the same volume as I had at my peak, but it was definitely helpful to keep some sort of written journal around for a while. The current state of the world is, hopefully, going to make it easy for me to get more programming time in. All things considered, I'd count that as a win.

1. Although admittedly, it does require me to explain the concept of zippers to a few other people for maintenance purposes. So ironically, this adds complexity despite being much more technically elegant than other options.↩
2. There's a reason that langnostic.js is a raw JS file, rather than compiled from clojurescript source, and that reason is like 90% that the compilation process is nontrivial to set up.↩
3. "First", not "only". You can probably make educated guesses about which other ones I think you should learn.↩
4. In which case, why are you here? This blog could kill you accidentally with an errant click or two. You should probably just go do something else.↩

It was all I could do, when I was reading the Futel Party Line zine, to not just call up their operator and say “I want a free pay phone. In Randolph.” Because, hey, that would be cool. But realistically, someone would have to manage that project. And, if I am being honest, I am […]

So hey I think I managed to break a leisure record this Virgo Month of Leisure! The bad news is that’s because I’ve been feeling borderline punk for a few months now. I always thought of myself as someone in pretty good shape, but every so often I just come down with… sinus nonsense, or […]

To all local bodies within U.S. Grand Lodge: For all official gatherings, please observe all guidance issued by the CDC, and by state or local health authorities, pertaining to sanitation, hygiene, and event attendance related to the new Coronavirus.

Event cancellations at local bodies due to COVID-19 will reduce revenue, but not rent and other fixed expenses. Many local bodies already operate on extremely tight margins and struggle to pay their bills even during normal times. All members are encouraged to continue to financially support their local bodies, even if there are no events …

Continue reading