Jump to content
Dante Unbound: Share Bug Reports and Feedback Here! ×

Tennotyper - An English To Warframe Text Translator


Clarvel
 Share

Recommended Posts

b87J2ov.png

oh, and yes, i still write my history notes in the Tenno language (https://twitter.com/NearlyDedicated/status/519826650725621760). now i can type my reports in it too! you just gave me the tool by which all my teachers will get headaches until graduation.

thanks a ton, man.

Edited by (PS4)NearlyDedicated
Link to comment
Share on other sites

Get ready for some theory, because it's late and I want to explain how things work! (or just complain about things not working)

 

Great work, but it still has a bit of polish missing.

 

The "ng" rune is replaced by "n" and "g", and "law" is missing the "h" for placement.

Yes, it doesn't translate perfectly, mainly because there are several words that are almost impossible to distinguish based on the letters how to phonetically pronounce them, ie:

danger -> d ae n j i r

longest -> l o ng e s t     (I think the orthos is wrong on this one, should be 'longist', not 'longest', but this may just be because of how I pronounce it)

or:

steal -> s t ea l

stealth -> s t e l th

or:

frame -> f r ae m

the -> th u h

law -> l a oo h

are -> a r     ( woo, silent 'e')

(and lots, lots more!)

 

The silent 'h' after vowels is interesting and something I didn't catch, as there are conflicting examples in the given translations(you->youh and are->ar and the->thuh). I suppose I could add the 'h' after if the word is 4 or less letters and ends in a vowel that is not a silent 'e'. 'the' would have to be a special case, as would any other words that don't fit the mold.

 

This is one of the problems of using a phonetic translation, not to mention accents or patterns of speech, no translation will be perfect without having word to word conversion, there will always be exceptions to the rules and words that make no sense phonetically.

As is, I'm not sure how to fix a lot of these problems without implementing some form of dictionary for each word, which is just too much work. I'm not writing a complete dictionary by myself.

 

All consonants in the tennobet are aligned from the top. In your case the letters for 'h', 'l', 'm', and 'n' are lower than the rest.

 

Vowels/accents are closer together and in general, notice how they are often touching the consonants.

 

I suggest putting the cheat sheet somewhere and have users refer to it when typing, instead of trying to make the app do it for them, because the English language makes no sense and it will probably take you months to take everything into account. Also because of this I can't find a way to type 'ch' as a part of a word. And if you include dashes and dots or other symbols the custom words you have for "you" and "the" return to how they used to be.

Kethus here, pretty much points out all the things I don't like about my TennoTyper, but those were all concessions made to facilitate finishing the project.
 
Because consonants are saved un-rotated and rotated afterward, I can have them but right up next to each other while having them overlap vertically (when typing, press 'option' or 'alt' to see canvas, line, word, and letter bounding boxes and the draw line all the letter positions are referenced from). To align them from the top, I'd have to store where the top of the symbol actually is somehow, the easiest way I can think of is to have 2 sets of images for each symbol, one rotated and one not, but that can run into problems with memory and what happens when one image gets updated but not its sibling. 
 
I plan on fixing vowel spacing later, likely over christmas break, and should be easy once I fix how the program adds letter spacing (currently there is none for tenno, but there is for corpus and grineer, see bounding boxes above, because it's not working right yet for tenno)
 

because the English language makes no sense

English is such a weird language! (I am not fixing, and you can't make me fix homonyms)

 

In any case, it's missing polish because there are some things just not worth fixing, and I think good enough for most cases is better than never released and stuck in development hell.

Link to comment
Share on other sites

Do you mind adding a switch to turn off the phoneticizer? And add a reference key somewhere? Since English is a mess, might as well let people do the work if needed.

 

For the alignment thing why don't you manually figure out the anchor point and then store the XY coordinate as part of the data? It's only a handful of letters and they are definitely never changing again.

 

(BTW I've noticed DE isn't entirely consistent in how they phoneticize things, and I think they sometimes make mistakes)

Edited by Kyte
Link to comment
Share on other sites

Do you mind adding a switch to turn off the phoneticizer? And add a reference key somewhere? Since English is a mess, might as well let people do the work if needed.

I can, and have done so!

 

Added checkbox to toggle program automatically guessing phonetic translation. Added cheatsheet button to open language specific reference sheet in new window/tab.

 

IE saving fix possibly incoming soon, I don't have access to IE, so I can't test this on my own.

Link to comment
Share on other sites

Because consonants are saved un-rotated and rotated afterward, I can have them but right up next to each other while having them overlap vertically (when typing, press 'option' or 'alt' to see canvas, line, word, and letter bounding boxes and the draw line all the letter positions are referenced from). To align them from the top, I'd have to store where the top of the symbol actually is somehow, the easiest way I can think of is to have 2 sets of images for each symbol, one rotated and one not, but that can run into problems with memory and what happens when one image gets updated but not its sibling. 

 
I plan on fixing vowel spacing later, likely over christmas break, and should be easy once I fix how the program adds letter spacing (currently there is none for tenno, but there is for corpus and grineer, see bounding boxes above, because it's not working right yet for tenno)

 

This is fantastic. I'm currently designing a badge for my clan and I'm going to incorporate the clan name in Tennobet as part of the badge as soon as the spacing fixes are in. Great job!

Link to comment
Share on other sites

This is fantastic. I'm currently designing a badge for my clan and I'm going to incorporate the clan name in Tennobet as part of the badge as soon as the spacing fixes are in. Great job!

Sadly I'm not sure how much I can work on this until ~the 18th, finals are looming surprisingly fast.

Link to comment
Share on other sites

 
Tenno: program's translations do not match all examples shown here.
 

 

 

you wouldn't happen to be able to explain why they don't quite match up do you?

 

after looking though a few of them intensely; I found that for one tribe, Hayden Tenno, and evolution were spelled correctly. unless of course it was a case of aesthetics then say no more. 

Link to comment
Share on other sites

Tenno: program's translations do not match all examples shown here.
 

you wouldn't happen to be able to explain why they don't quite match up do you?

 

after looking though a few of them intensely; I found that for one tribe, Hayden Tenno, and evolution were spelled correctly. unless of course it was a case of aesthetics then say no more.

 

That's because the Tenno alphabet is phonetic.

It doesn't say "You are the hope and future of the Tenno", it says (if you 1:1 translate using the cheatsheet)

 

eeooh awr dhuh,hop and feeoochoor,

awf dhuh tehnoh.

sWOtAxa.png

 

Clarvel made an admirable attempt at machine-transforming written english to phonetic english, but it's literally impossible to make a perfect phonetizer.

If you want to write Tenno properly you'll probably need to hand-convert it into phonetic english before sending it through TennoTyper. This is why it has a "manual override" checkbox.

 

Did nobody read the newspost explaining the tenno language?

Edited by Kyte
Link to comment
Share on other sites

Clarvel made an admirable attempt at machine-transforming written english to phonetic english, but it's literally impossible to make a perfect phonetizer.

If you want to write Tenno properly you'll probably need to hand-convert it into phonetic english before sending it through TennoTyper. This is why it has a "manual override" checkbox.

 

That's basically what I did when I tried it to see what my clan's name would look like. It includes the spanish word "pollo" but a double-L sound in English is the same as a single L, whereas in Spanish it's pronounced with a sort of "Y" sound. Phonetically for the Tennotyper, I had to type it as "poaeo" to get the right sounds across.

Link to comment
Share on other sites

What I learn abiut the comments are that its based on phonetics... Well phonetics, visualise, imagery, teevision farvision..... SPONJEBOBBE

hue

det iz kul

think about rainbows, think, alitnin. tausend sans....pahweh.gad.hi bin wid as de hol taim, bat hev wi bin der, didyu rimimber abaut it wen u ar bon? didje chuz wat u wer? Wat u ar now?

Link to comment
Share on other sites

The Grineer "Q" on the tennotyper is using the same symbol as "K".

 

EDIT: Also, you have the Corpus "L" listed wrong. It looks like the "S", but there's a line through it.

 

As far as I can tell, the Grineer don't have a 'q', but I seem to remember general consensus wanting 'k' to replace it.

fixed missing visual distinction between corpus 's' and 'l', thanks for finding this, the image on the corpus language wiki doesn't have the best resolution.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...