2024-01-02 00:29:09 is "only" not supposed to reset current? 2024-01-02 00:30:07 well, now that i think of it, i guess it can't. otherwise you couldn't define vocabularies without them also being in the search order 2024-01-02 00:30:16 well you could use previous to that, i guess 2024-01-02 01:01:33 ACTION looks at https://axleos.com/writing-a-truetype-font-renderer/ 2024-01-02 03:24:07 forth.org is back up now 2024-01-02 03:28:04 crc: thanks :) 2024-01-02 03:28:28 crc: are you the admin or are you just the messenger? 2024-01-02 03:29:16 just a messenger for this one 2024-01-02 19:52:11 Out of curiousity, any example of memoization in Forth? 2024-01-02 19:53:17 I've thought of it for a bit and I guess it could be done with CREATE DOES>, allocate a block with CREATE and do the computations with DOES> 2024-01-02 20:08:57 Forth really leaves it to you to build such things yourself, but I do think there are some systems out there that use memoization to construct a hash table for dictionary searches. It's certainly something I've thought about doing in cases where I have a lot of unused RAM. So you'd have the usual linked list, but each time you search the list for a word you make an entry in a temporary hash table for it. 2024-01-02 20:08:58 That's an example of memoization. Subsequent searches for the same word would find it in the hash table and thus totally bypass the linked list search. 2024-01-02 20:12:10 it's common in algos where you hit the same result over and over and over 2024-01-02 20:25:00 Yeah, they talked about the Fibonacci sequence in Algo class today :) 2024-01-02 20:25:24 You only need the previous value for Fibonacci though 2024-01-02 20:28:27 Yeah, at first they started with a recursive approach then changed to the more sensible loop approach. 2024-01-02 20:29:09 I like this way of generating the sequence in Haskell: fibs = 0 : 1 : zipWith (+) fibs (tail fibs) 2024-01-02 20:29:19 Saying 'Yeah,' twice in a row seems awkward.. so.. happy new yeah :P 2024-01-02 20:30:27 Also, x86 has an instruction called xadd which can be used for Fibonacci 2024-01-02 20:31:12 https://www.felixcloutier.com/x86/xadd 2024-01-02 20:35:32 https://ideone.com/JGuSuG 2024-01-02 20:37:29 I'm reading https://stackoverflow.com/questions/6273621/understanding-a-recursively-defined-list-fibs-in-terms-of-zipwith now 2024-01-02 20:38:46 Haskell can define infinite lists 2024-01-02 20:39:12 zipWith combines two lists with a function, in this case + 2024-01-02 20:39:30 code for "infinite" lists was also shown in PAIP 2024-01-02 20:46:46 https://ideone.com/ekOHu8