[Cenaurille]

Tiaga said:

My guess is if they have a problem with it, they'll just change how the garbling is done.

Or they could simply deem it a bannable offense and be done with it. In the mean time I'll just screenshot all the reds who throw up some smack talk after they gank me 4v1.

[Tiaga]

Cenaurille said:

Or they could simply deem it a bannable offense and be done with it. In the mean time I'll just screenshot all the reds who throw up some smack talk after they gank me 4v1.

So you and none of your legion puts up private stores to say things to the other side? And I assume you screenshot anyone who does that as well...

[JKMB]

Spirit said:

if you are fighting an enemy, and see him say "Hgkls fdkaslj fdskel"

then open the program, go to the system tray on bottom right, flip the translator to translate into YOUR FACTION. then

and go into notepad, type "/tr Hgkls fdkaslj fdskel" , press control-a, control-c, control-v

then you will see what he said in english. (like they said above, if the enemy uses caps it screws up translation slightly - but its still understandable)

Ah thank you very much.

Top quality thread, belongs in the mechanics archive imo, and a brilliant piece of software.

[Elumine]

Cenaurille said:

Or they could simply deem it a bannable offense and be done with it. In the mean time I'll just screenshot all the reds who throw up some smack talk after they gank me 4v1.

Opened a support ticket inquiring about this today, and good news - it's not a bannable offense. As long as the translation is taking place completely independently of the Aion client (which it is) we shall be just fine.

[Tiaga]

Thanks for posting the results of that Elumine! That is about what I expected - but there is nothing that says they won't change how things are done in the future to prevent it.

[Platinum]

i was leveling a 26 temp alt on crushed souls and this asmo assassin appears behind me and i freak and pop my run speed+60 spell and make a break for it, but then i went back(don't ask why) and i stared at him and then he(surprise) started talking to me. and i quickly opened the translator program to tell him that he surprised me.

after a few minutes a lvl 26 elyos cleric appears and sees the asmodian and starts shouting in untranslated text "go **** your mother you *****" to which the asmo says "please translate" i let him know what is being said and he ran over and killed the cleric. many lolz were had.

then i rezzed the cleric and told him to play nice. =3

[Elvengard]

I can't them to mke it work. The program is installed, but I don't know what to do now.

Is there a tutorial to use this? Do we need to type commands?

I saw that there is a /tr. Tried using it, but won't work...

Thx!

[nicodega]

Elvengard said:

I can't them to mke it work. The program is installed, but I don't know what to do now.

Is there a tutorial to use this? Do we need to type commands?

I saw that there is a /tr. Tried using it, but won't work...

Thx!

- Select the language you want to translate to on the app. (popup menu on the tray bar)
- Write /tr what you want translated on the chat window, select the line, copy and paste.

When pasted you'll see the translated text instead of the original.

[SandsJP]

Could someone please walk my dumb person through this?
I downloaded th App, extracted and ran it. I can translate through notepad but nothing pops up after running the app so when ppl talk about switching the race in a drop-down, there is none for me. What am I doing wrong?
Thanks in advance.
:what:

[nicodega]

SandsJP said:

Could someone please walk my dumb person through this?
I downloaded th App, extracted and ran it. I can translate through notepad but nothing pops up after running the app so when ppl talk about switching the race in a drop-down, there is none for me. What am I doing wrong?
Thanks in advance.
:what:

When you run the app, look fot it on the tray bar (where the clock and the icons are on the taskbar)

[Isceroth]

Your Elyos -> Asmodian cipher is flawless. The Asmodian -> Elyos is not. It should be as follows:

First Letter
Elyos Output:	a b c d e f g h i j k l m n o p q r s t u v w x y z
Asmodian Input:	j k h i n o l m r s p q v w t u z G x y J a f g d e
		1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 1 1 3 2 2 2 2 2


Group 1
Elyos Output:	a b c d e f g h i j k l m n o p q r s t u v w x y z
Asmodian Input:	e f c d i j g h m n k l q r o p u v s t y z w x I J
		1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 3

Group 2
Elyos Output:	a b c d e f g h i j k l m n o p q r s t u v w x y z
Asmodian Input:	f g d e j k h i n o l m r s p q v w t u z G x y J a
		1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 2


Group 3
Elyos Output:	a b c d e f g h i j k l m n o p q r s t u v w x y z
Asmodian Input:	g h e f k l i j o p m n s t q r w x u v G H y z a b
		1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 3 1 1 2 2

Confirmed and tested on my Asmo, with my Elyos friend present, where we could actually look at one another's screens. Please credit for the fix. It was a well spent two hours.

[Tiaga]

Thanks for testing it!

Can you double check your group 2 v=G? I couldn't think of any words that actually use the letter combinations to get there, but "Zva" as a test set should hit it. According to my simulations, the original table should be correct with "eGg" while your modifications would have it typed as "eGe" which my simulation says the Elyos would read as "Zvc".

The changes to "s", "t", "w" and "x" are cases of different letters producing the same result. (To see what I mean, try having an Asmodian say "fbb fbx fxa fxw" and an Elyos should see "Www www www www".)

The remaining three are indeed mistakes in my tables, verified with my formula based test program as well as against my program which takes the formula and produces the tables out of it. I think the mistakes slipped in while I was adding the color codes to the post, and were missed in spot testing because not many words use those combinations. (Jazz was the only one I could think of off the top of my head, though I'm sure there are plenty others. Using those tables it would have been seen as "Jazy".)

Let me know what you find if you get a chance to double check that. Until then I've modified the tables to reflect the three mistakes I verified in my simulation.

This post has been edited by Tiaga: 20 December 2009 - 06:32 AM

[Isceroth]

Yeah, I'm aware that some different letters will produce the same output for the Elyos. I was using my own tests, because I had something like this going before stumbling across your thread, so I may not have used all of the same letters that you did. All I know is that the tables that I posted, as they are, are flawless.

I only did a few short tests with yours to see if it worked, and when I used it to say a very essential word (that being, "Elyos") from Asmodian to Elyos, my Elyos friend on vent said it was completely incorrect. Now, I had been using "nlIqs" to say that term before, and noticed that your tables weren't necessarily off, but I think after the "I" (which would be the y to the Elyos), I was referred to the wrong table.

I'll test out that v=G thing when next I can.

[Tiaga]

Aside from that one little thing your tables look perfectly fine as well. I haven't had anyone comment that what I said was garbled, but I've been translating with my test program which uses the underlying formulas rather than tables. (Same program that created the tables, which is really just applying the formula to all the letters in turn.) The full table I posted near the bottom of page 1 should be more accurate as it was closer to a copy&paste than the first one. Mostly I just did spacing to make it look nice.

For grins, I can even translate something so a Mau, Balaur or Krall could say it to someone who isn't fluent and they would understand - not terribly useful since you'll never be in that position. (Or translate what Mau, Krall or Balaur say without being fluent, which is only mildly more useful.) It's all the same formula, just different values.

[CraiZE]

Does someone have the table (all of it) and could explain it? I would love to code a translator :)

What does Group 1, Group 2 and so forth signify?

[Tiaga]

The post explaining the underlying mechanism the game uses can be found here. To be honest, a good translation program should use the formula rather than tables, to avoid cases like my little mistake in formatting things.

But what the different groups are... If you don't want to look at the underlying formula, the basics are that what a letter is influences how the next letter is translated. With the formula, that influence is limited to a small set of variations. The initial letter has no variation, then letters after that have one of three variations. The different groups/tables represent how a letter is being influenced by the prior letter.

If you want more specifics, or an explanation of the formula and how I use it in reverse, feel free to give specific questions either in this thread or in PMs.

[Nya]

My Nya's Aion translator uses the formula, not the tables anymore. Hopefuly it does it correctly, havent met any problems with talking with elyos and even backtranslating worked pretty well.

[Tiaga]

Nya said:

My Nya's Aion translator uses the formula, not the tables anymore. Hopefuly it does it correctly, havent met any problems with talking with elyos and even backtranslating worked pretty well.

Ah you caught me on a flaw in my posted formula. I actually fixed it an hour ago right before your post. The flaw is that the language modifier wasn't being incorporated into the carry calculation, where it should be. Basically the carry is the division by 26 (Plus 1) and the actual letter is the remainder. The bad translation doesn't always come out because it's just a few cases that it effects, and it's usually just a single letter error. It also only effects Asmodians. The Elyos translation does not have a problem. (For the same reason Elyos always has exactly 2 possible translations and Asmodian can have 1 or 3 in a few cases. In fact it specifically comes out after a letter which has only one possible translation.)

To illustrate this, the phrase "Elyos are playing jazz" should be said by Asmodians as "nlIqs jvi uleIorg seJb". With my old incorrect formula, it comes out as "nlIps jvi uleInrg seJa", or "Elyjs are playlng jazy". As I was directing someone else there right before your post, it occurred to me to check that the formulas were correct, and I found I forgot to update them with the little corrections I made after that post but before my initial tables were posted.

I take full responsibility for that translation error too. :)

Come to think of it... I might have made the tables before that fix and forgot to update it, resulting in the error I just fixed.

This post has been edited by Tiaga: 21 December 2009 - 05:49 AM

[Nya]

Tiaga said:

Ah you caught me on a flaw in my posted formula. I actually fixed it an hour ago right before your post. The flaw is that the language modifier wasn't being incorporated into the carry calculation, where it should be. Basically the carry is the division by 26 (Plus 1) and the actual letter is the remainder. The bad translation doesn't always come out because it's just a few cases that it effects, and it's usually just a single letter error. It also only effects Asmodians. The Elyos translation does not have a problem. (For the same reason Elyos always has exactly 2 possible translations and Asmodian can have 1 or 3 in a few cases. In fact it specifically comes out after a letter which has only one possible translation.)

To illustrate this, the phrase "Elyos are playing jazz" should be said by Asmodians as "nlIqs jvi uleIorg seJb". With my old incorrect formula, it comes out as "nlIps jvi uleInrg seJa", or "Elyjs are playlng jazy". As I was directing someone else there right before your post, it occurred to me to check that the formulas were correct, and I found I forgot to update them with the little corrections I made after that post but before my initial tables were posted.

I take full responsibility for that translation error too. :)

Come to think of it... I might have made the tables before that fix and forgot to update it, resulting in the error I just fixed.

Thanks a lot, fixed it in my algorithm.

[Tiaga]

Nya said:

Thanks a lot, fixed it in my algorithm.

Seems to work fine in the quick spot testing I did with a few generic phrases with my simulator.