Search by:
· starts w/
· ends w/
· contains
· does not contain
· # of instances of letter/cluster/sound
· # of letters
· characteristic of letter/sound (vowel vs. consonant, plosive vs. fricative, shape of letter)
· letter/cluster/sound order
accent placement
· "and/or" selection
· IPA and other alphabet inputs
· language
· synonyms
· translations
· lexical reference
· part of speech

Display:
· word (recognition of characteristics) (recognition of syllables, accent placement)
· IPA pronunciation (and U.S. diacritic pronunciation)
· lexical reference
· part(s) of speech
· definition
· synonyms
· translations to other languages
· picture
· clickable google/wikipedia/web results

*Fill in the Blank / Crossword Puzzle Cheater
I found this out by accident and it's really easy with the grep feature. I can only really see this being used for crossword puzzles. Type in the number of letters of the word and the letters in the word that you already know, then choose the word that best fits. Bam.

*Synonymizer
Once restrictions have been set and a specific word is desired but does not fit the restrictions, the Synonymizer will search through the list of synonyms and translations from other languages of that word to find one that fits. The list of synonyms could be increased by others adding to it (like a wiki) and could provide for some real creativity and expansion of the language. Or maybe I'm just full of myself.

I'd like to thank Matt Finnegan for his continued support in creating a Java version of the Timmer. Currently we have, what Matt likes to call, version 0.1. We can search for words using the "starts with", "ends with", "contains" and "does not contain" variables, but funny enough, only if all the fields have something in them. It's an excellent start.

Here's a concept mockup I made a while back. This doesn't actually work at all. No. I don't know how to make a program...yet. But I know how to fake one.

This is the most basic version of the Timmer that I can create right now. I can't do it through a program that is that pretty, but I can get all of those results (looking for a word based on its beginning, end, what's in the middle, what's not in it), as well as others that I don't know how to articulate in a program yet, using the less-appealing Terminal in Mac.

All the concepts I don't yet know how to get, but would like implemented in the Timmer:

Note: the "number of times", "syllables" and "total letters" would accept numbers like "5", "between 1-5", "less than 5", "5 or more", etc.

Adding part of speech could help for clarity in the writing so it's not just nonsense. When I wrote the "e" poems, I kept looking for "e" prepositions so I could make my sentences more complex. I've actually figured out a way for the search utility to look for the letters in the word as well as the part of speech. It's very crude, but it's headway.

Choose which language it searches through. Ideally, I'd like to be able to use this feature so that different words with similar aspects in different aspects can be used interchangeably in writing (I got this idea from OuLiPo a while ago. Hell, I got the whole idea of this 'device' from them...in order to make experimental writing based more on the sounds and letters and less so on the actual words).

Since not all languages use the Roman alphabet, there could be a switch to move to IPA (International Phonetic Alphabet) so similar sound patters could be searched for in languages like Chinese, Russian (Cyrillic) and Hebrew and give their words. Then the writing could incorporate multiple writing systems but keep similar sounds. The switch could also grow to allow the input of all different writing systems.

From my test for searching for a word and its part of speech, I theorized on the ability to search for all the data related to a word if each entry looked something like this: crash:ipa-kRaeSH:let-5:syl-1:spc1-verb:spc2-noun:lex1-travel:lg-en

Yes, it's a nightmare. Yes, the IPA transcription is wrong. No, that won't be how it'll work at all, but it's just a theory.

word

ipa transcription

number of letters

number of syllables

part(s) of speech

lexical reference(s)

language(s)

The "Fill In The Blank / Crossword Puzzle Cheater" is something I know how to do now, but can't figure out how to represent in the program so it's easy for the user to figure out. Basically, a person would type "ch...a...e" and the result would be "chimpanzee". This is basically only if a person knows the exact number of letters in the word and just wants to fill in some blanks- probably for a crossword puzzle.

I found the lexical references from here and changed it a little. I don't really know which categories I'd want, but this could help the writing make further sense in context, rather than just putting in words based on the letters, and number of, in them.

I'm thinking that if I can get some aspect of this done, I can shop it around and get backing to license big dictionary companies databases and get people who actually know what they're doing to integrate it all. Of course, it's all just a pipe dream.

And yes, 'Timmerize' is a joke. In theory it should be 'Timm', since it's the 'Timmer' and that clearly means it's "one who timms". I'll think about it. I've actually worked on the programming on my computer and have called it "Busco Stick", like a tasty Bosco stick and the Spanish verb 'buscar' ('to search'). Yes, how lame.