Date
1 - 14 of 14
Newbie - using FoxReplace to learn to read Quikscript
|
Hi all,
I'm new to quikscript and just want to share my experience learning it. With any new script (learning languages etc.), I think it's very useful to learn to read before writing so I have prioritised that first. It doesn't mean that I am not learning to write at the same time, it just means I am trying to carve more opportunities for reading. The goal is to try and convert web pages to quikscript and right now I am not aware of any browser extensions that do that.
For a quick win, I have downloaded FoxReplace (a Firefox addon) and started adding a few of the most common words myself. It's not perfect and certainly need to tweak some things and add regex in certain cases. Refer to the images to get the idea. I am hoping to add a few words each day so it's in my mental vocabulary. When I journal, there are inevitably words that I will visually remember and can summon at once. I never force myself to write a word that I am not 100% sure of. Initially, I couldn't get the webpages to show the quikscript fonts. I had to download them from http://friedorange.xyz/quikscript/otherfonts.html. I then installed them onto my OS but that was not enough for Firefox to show them. I had to go into the preferences and select Quikscript Sans as the default font and that fixed the problem. My coding skills are weak but if I was better, I think the main thing we would need as a prerequisite for such a project is some kind of Quikscript dictionary. When that is sorted, an API can be created for it so I don't have to maintain the above table of definitions myself in FoxReplace. I don't know if Quikscript will have different spellings based on accent. If it does then possibly multiple dictionaries are required (one for British English, one for US etc). This is possibly a massive project in itself. Might be able to speed this up by translating all the IPAs directly to Quikscript. Just sharing my thoughts. I've attached the FoxReplace file as well in case you want to have a go. |
|
|
|
Brad Neil
Hi Bob,
I actually took a similar approach to generating reading material, too. I wrote a rough-and-ready Lua script that would transliterate an input text, word-for-word, into Quikscript using a user-defined dictionary. That allowed me to read things entirely in Quikscript for the first time, without seeing the original text beforehand, similar in principle to what you've been doing. So I can certainly see value in your project. I'm currently working on an improved, browser-based version of that tool that would actually be user-friendly and fit for public release. Over time, a sizable dictionary could be built up, (but not without considerable initial effort). From there, it would be a fairly trivial task to convert the dictionary for use with other tools like FoxReplace (or vice-versa). Unfortunately, browsers and OSes don't automatically know which fonts to use for arbitrary characters - they have a predefined list of fallback fonts which doesn't react to newly installed ones. That would be why you had to manually choose Quikscript Sans. I never force myself to write a word that I am not 100% sure of.What do you mean by this? Are you unsure of the pronunciation, or are you trying to find 'proper' Quikscript spellings? If the latter, my advice would be to aim for familiarity with all the letters and their speech values, which should allow you to write anything you can pronounce (albeit, in your dialect/accent with all its local and individual peculiarities :-) ). I don't know if Quikscript will have different spellings based on accent.Some of the most obvious pronunciation differences between those dialects are what linguists call the Trap-Bath Split and Non-Rhoticity; where most English speakers (including most parts of England and Wales) would pronounce "pastor" as /pæstər/, but those from southern England, Australia, New Zealand, South Africa, India etc. would say /pɑːstə/. Words affected by these differences are to be spelt as for the majority pronunciation in Quikscript (that is, use the 'short a' vowel and include the R), eliminating the majority of potential variant spellings. Beyond those differences, don't worry so much about finding canonical spellings, since these will always vary for certain words: tomato (təmɑːtoʊ/təmeɪtoʊ); suggest (sədʒest/səgdʒest); debris (debrɪ/dəbriː) and so on. Those examples are British-Australian-etc/American, by the way. I hope this answers your uncertainty about spelling differences between accents - the short answer is yes, differences certainly can and do exist, but those speakers in the minority are advised to make the conscious effort to standardise. After this, a relatively small number of varying words will remain and it should be no more difficult to read them than to hear them spoken in the 'other' accent. Is it strictly necessary to write differently from our speech? Of course not, just as the language police won't kick your door down for making 'misspellings' in the Latin alphabet. But it's nice to have a baseline of consistency for the 99% of words where it's easy to do so. I'm a relative newcomer myself, having discovered Quikscript about 6 years ago, so it's really great to see other folks starting out now. I'd love to participate if you wanted to discuss anything about Quikscript :-) Brad |
|
|
|
Benjamin Bruce
This is a cool idea, and I've thought about the dictionary idea before too. The main hurdle I've wondered about is how to handle heteronyms: words like "read," "live," "tear," etc. which have two pronunciations and two meanings but are spelled the same in traditional orthography. At the very least it would require some programming logic to figure out which word is meant in order to translate it into QS, and given the complexities of language I figure it would be hard to come up with an algorithm that gets it right every time. Any thoughts? Ipan ne 4 tonal tlen metstli 07 tlen xiwitl 2020, ipan 11:55 kawitl, chickensaregreat@... kiihkwilo:
|
|
|
|
John Cowan
On Tue, Jul 7, 2020 at 8:37 AM Benjamin Bruce <benjamin@...> wrote:
Translation is always imperfect. Keep both in the dictionary, and when you see "read", split out "{reed|red}" in QS and leave it to the user to fix them up. They can be found by searching for any of {, |, or }, which are all rare in ordinary text. It also won't happen often. <http://www-personal.umich.edu/~cellis/heteronym.html> is surely not complete, but lists 77 heteronyms: affect, alternate, are, attribute, axes, bass, bow, bowed, buffet, close, combine, conduct, conflict, console, content, contest, contract, convert, converse, convict, crooked, deliberate, desert, digest, do, does, dove, drawer, excuse, house, incense, intern, invalid, laminate, lather, lead, live, minute, moderate, mow, multiply, number, nun, object, pasty, pate, perfect, periodic, permit, polish, present, primer, produce, project, pussy, raven, rebel, record, recreation, refuse, relay, rerun, reside:, resign, resume, row, sake, secrete, secreted, separate, sewer, slough, sow, subject, tear, wind, wound. (I have removed the ones that differ in case, like ares/Ares and august/August, since QS has the namer dot.) |
|
|
|
Benjamin Bruce
Good points, and I can see how that would be a good solution if you are just outputting a text. I guess I was thinking more along the lines of a browser plugin transcribing text on the fly like chickensaregreat is wanting to do. Thanks for the heteronym link by the way, very interesting! Ipan ne 7 tonal tlen metstli 07 tlen xiwitl 2020, ipan 08:45 kawitl, John Cowan <cowan@...> kiihkwilo:
|
|
|
|
Hi Brad,
Good to hear that you took a similar approach even though I haven't a clue what Lua scripts are. It's been a week and I can read all the letters now but obviously it's not as fluent as my usual reading speed. The only way forward is more immersion. Yes you were right about my question on spelling, I was looking for 'proper' spellings which don't yet exist since there is no official site of some sort. I will go by my own speech. On the topic of the addon thing, I myself tried the Firefox tutorial on creating an extension and got as far as making the border of the mozilla webpage's border red. Beyond that, my javascript skills are limited so I was going to attempt and just make small snail progress a day at a time. But if you're already doing something similar, I wouldn't want to start from scratch. Will I be able to help with the browser-based tool that you are working on? Or will it be open up to more of the group as a whole? I think it will certainly speed things up if we can have the community contribute their Quikscript 'spellings' or split it so we have custom dictionary entries if needed (US, UK, Aussie etc). Do let me know if there is anything I can do to help. |
|
|
|
Sounds quite complex, I have no knowledge in that area so if it was me developing it, I would probably leave those questionable words until I can figure it out.
There's also other exceptions I thought of like, what if you wanted to keep the initials of the normal alphabet. For example, 'If you pick A or B, you will lose.' , you wouldn't want to replace 'A' with the usual 'a' (in the sense of 'a pen'). |
|
|
|
Ph.D.
Lua is a programming language. It's designed to be embedded into
other applications.
toggle quoted message
Show quoted text
https://www.lua.org/ --Ph. D. On 7/7/2020 11:47 AM, wrote:
Hi Brad, |
|
|
|
jcdietrich@...
See https://groups.io/g/QuikScript/files/International%20Quikscript/Malt%27s%20International%20Quikscript%20Project for a good starting point for a dictionary. Jonathan C. Dietrich On Tue, 7 Jul 2020 at 10:15, Benjamin Bruce <benjamin@...> wrote:
|
|
|
|
John Walker <ac7gkjohn@...>
You folks are making too much of a problem here. The phonetics take care of your heteronyms. Just use the phonetics, the way we do in oral conversations. ToneText makes it clear, simple, neat, and easy to use.
toggle quoted message
Show quoted text
John Peace comes from accepting people as they are, not trying to make them into what we want. On Jul 7, 2020, at 11:46 AM, jcdietrich@... wrote:
|
|
|
|
Brad Neil
Regarding Latin-spellings with multiple pronunciations (read, use, conflict, etc. etc.) I always thought the word for it was 'homograph'. I just looked it up on Wikipedia and it turns out that 'heteronym' is actually more technically correct for what we're talking about. So thanks for bringing that up!
Anyway, a few years ago I found this list of 647 of them: https://web.archive.org/web/20180816004508/http://www.minpairs.talktalk.net/graph.html?showpage=true It was one of the first things I incorporated into my personal transliteration dictionary. My Lua program was designed to stop and prompt the user to choose one of the two possible QS spellings for each one found in the text during conversion (you would have to have the original text open, ready to search for the words to identify each one manually). I figured that would be much easier and more reliable than trying to find some kind of algorithm to do it! And as John C. said, they don't come up terribly often so it doesn't really create unnecessary tedium :-) |
|
|
|
Instead of an algorithm or singular words, how about we do it by phrases?
Examples: I read that book yesterday would definitely be 'RED' so we can replace that whole phrase with the Quikscript equivalent. Granted, this workaround is not ideal because we will be clogging up the dictionary with arguably unnecessary bloat but it's a workaround nonetheless. |
|
|
|
John Cowan
Alas, that's hopeless: you underestimate the productivity of language. Consider all the words for things that can be read besides "book": magazine, newspaper, article, post, comment, tweet, page, paragraph, (stone or clay) tablet. Then consider all the possible adjectives that could apply to any of them. Mere phrase substitution will not work. Consider "The more the merrier." That doesn't look like a grammatical sentence, and most people simply understand it as a fixed phrase. But it is indeed a type of sentence and just as productive as any other sentence type: "The deeper I breathe, the more often I cough" or "The more bored you are in class, the more difficult it'll be to enjoy the subject and learn" or even "The less coffee I drink, the sleeper I seem to feel." On Wed, Jul 8, 2020 at 10:16 AM <chickensaregreat@...> wrote: Instead of an algorithm or singular words, how about we do it by phrases? |
|
|
|
Nathan Galt
I did something like this in adaptor raptor. One base dictionary (50.yaml), combined with a document-specific dictionary (04.yaml) is how I’m* adapting “Omnilingual”. The document-specific dictionary gets ambiguous and document-specific words/phrases like:
toggle quoted message
Show quoted text
- “wind” (in this story, there’s a breeze, but no cranking anywhere) - “Wilhelm II” (unlikely to be useful elsewhere) - “what I object” (because the spelling of “object” is generally ambiguous) - “object after object” (uses the other spelling of “object”, because we’re talking about Arean junk lying around) - “Tony” (elsewhere in English, “tony” is an adjective, too, in which case it doesn’t get the namer dot) Of course, reading something like “wind” and remembering that the other pronunciation exists can be a bit of a challenge. I’ve probably misclassified at least one by now, and I’m not even done with the short story.
|
|
|
