Clojurians Log v2

New releases of ClojureCUDA, Neanderthal, and Deep Diamond, with CUDA 11.2 support. http://clojurecl.uncomplicate.org, http://neanderthal.uncomplicate.org, and http://github.com/uncomplicate/deep-diamond

CIDER 1.1 is out! More details here https://github.com/clojure-emacs/cider/releases/tag/v1.1.0 Cheers! :cider:

@cassiel I deleted your sweet reply CIDER glass. It’s because the content at the root level of this channel needs to be announcements. There are bots and such publishing it around the interweb. Totally agree with the sentiment!

:cider:

My bad - no problem. I put it in the wrong place.

when could we expect it in https://stable.melpa.org/

No idea. I assume it will land there fairly soon, but I'm not familiar with the build schedule and I don't know how often the packages are rebuilt.

There's also the chance that the extra tag I added to fix the docs version broke MELPA in some way. 😄

For anyone doing language-related work, I just released a library that converts text to phonetic sounds and does syllabification of those sounds. https://github.com/eihli/phonetics https://clojars.org/com.owoga/phonetics

(get-phones "alaska")
  ;; => [["AH0" "L" "AE1" "S" "K" "AH0"]]

  (syllabify (first (get-phones "alaska")))
  ;; => [["AH0"] ["L" "AE1" "S"] ["K" "AH0"]]

  (syllabify (first (get-phones "foobarbazia")))
  ;; => [["F" "UW1"] ["B" "AA1" "R"] ["B" "AA1"] ["Z" "IY0"] ["AH0"]]

  (get-word ["AH" "L" "AE" "S" "K" "AH"])
  ;; => ["alaska"]

  (get-word ["N" "IY" "S"])
  ;; => ["neice" "neece" "niece" "nice(1)" "kneece" "kniess" "neiss" "neace" "niess"]

Very cool! How does the approach compare to something like soundex?

I wasn't familiar with Soundex. Looking at it now, it looks like Soundex is like a hash function that hashes words (specifically names - proper nouns) to a 4 character code. Robert and Rupert both map to "R163" in Soundex. But the phonetics are different. I'm guessing Soundex is used as a hashing function in database indexes? The phonemes this produces are for "human" consumption.

> A future TODO and a request for contribution would be for a function that converts a made-up set of phonemes into a somewhat appropriate spelling of a word. Haha, this was my immediate thought "Oh cool, a way to generate words" :P

A quick search didn't reveal any existing implementations of the reverse of https://freetts.sourceforge.io/javadoc/com/sun/speech/freetts/lexicon/LetterToSoundImpl.html.

Looks like this paper describes an implementation. https://www.cs.cmu.edu/~hovy/papers/07IJCAI-spelling-variants.pdf

> I’m guessing Soundex is used as a hashing function in database indexes? Yes, that’s how I remember it. Long time since i used it. Back in the mid 90's. I adapted it for Swedish names for a CRM I was working with.

Pathom Viz 2021.04.22 is out! The changes: - UI improvements on Request tab, now the recent items will be at the top. The max size for a request row can now fit 5 lines of query, and I add a border on the right side to make easier to scroll the whole list in the presence of many scrollable requests. - Now when seeing node details in the graph, there is a button to show the whole graph data, this complements the data from individual nodes. - Now when you log a graph you can click on nodes to see their details. Another great news is that thanks to @ales.najmann now Pathom Viz also creates Windows builds! Now you can download Mac, Linux and Windows builds at the Pathom Viz releases, download 2021.04.22 at: https://github.com/wilkerlucio/pathom-viz/releases/tag/v2021.4.22

in addition pathom-viz is newly available via Scoop installer on Windows.

scoop install pathom-viz