Kurkistan

I have to admit that that goes so far over my head that my neck hurts looking at it. Perhaps this is another example of me failing physics forever, as you so helpfully pointed out last time. Not that I doubt you, but could we call in another mathy-person to take a look at these numbers? Barring that, could you use dumb-person-speak to explain each step of this process. Where exactly you get the number 100 at the beginning there? Also, given this: doesn't this: mean that dv/dm ~= 0, or just about? AKA, an exceptionally small energy gain?

*Pulls vacuum-tube monitor out of basement* Ah ha! Your evil scheme has failed! I think we may actually be reaching the endgame here. Someone who isn't me needs to go over a relatively long and diverse text with a fine-toothed comb to find phonetic errors, and we'll probably need to rearrange the order of the grammars at some point before the day is over, but barring massive oversights on my part, I think we may have most everything we need written down in the program. Updated "-ly" suffixes to cover all "-y" suffixes, reordered grammars, added various new grammars.

Better late than never, that's what I always say. 1. If the object is wholly contained within the bubble upon "bubbling", then no part of it need enter at any point, resulting in pure gain. As for lowering the energy of objects inside, this is only a temporary phenomenon which is reversed upon either leaving or dropping the bubble and which we ultimately gain energy from due to differentials as objects exit the bubble. 2. I don't see why you would think bullets should stop completely. Lose some energy, yes, but not stop completely. My head isn't in the math right now, but off the top of my non-mathing head, only an object with infinite velocity would be stopped dead when leaving a bendalloy bubble. I don't know if I can speak to air temperature, but my hazy understanding is such that, from the perspective inside a cadmium bubble, temperature stays the same while the outside heats up, but any air molecules crossing over will immediately assume the temperature of the proper environment because there are no indissoluble connections to transfer energy between molecules inside and outside of the bubble. EDIT: Although, you wouldn't know that the other side of the bubble is a different temperature, since you can't meaningfully interact with it. I wonder what a laser thermometer would say if you pointed it at the other side of a bubble, given the weirdness of light's interactions with time bubbles.

I've reached the conclusion that we deserve medals for this. No rush, but it would be nice. Substantial reordering of C category to simplify debugging, substantial number of grammars added to C category, various bugs squashed, -ed, -ly, -ish, -able, and -er suffixes added, timer added to show how long the transliteration took, various touch-ups. As always, but particularly with so many changes, comments are welcome. EDIT 2: added tests for y\n, not sure if I might want to put them in replace(). EDIT: Just ran the Odyssey again, 8 minutes, 7 seconds with no misses on 'c' except for names and weird compounds (washingcistern, panicstricken).

*Opens .doc with Alethi script at 72* *Screen explodes* Ah, I see what you did there. I assumed that "AlethiTS" was simply the old version. My mistake. I prefer no lines between words, and that's how it is in the notebook pages, so I think it's better to leave it as optional. *Goes back to work*

As far as I've seen, it doesn't take away line breaks. Could you link in a .txt and it's output for me to test? I'm in the middle of making it compatible with -ed, -ly, -ish, and -able endings, as well as hunting down some fugitive c's. I'll try to get to your bug by the end of the night. Thanks for bringing it to my attention. EDIT: Thanks for posting a new version, although I would suggest renaming the upload-version to just AlethiTS, since we want it to be compatible for the documents and posts of people who upgrade.

I use Google Chrome's MagicScroll eBook Reader app, or you can use a free program called Calibre to convert from .ePub to .pdf.

New version up with gerunds, proper permutations of ise\n, and easier to use debugging instructions (just change the two booleans at the top of the program). EDIT: Added rules for nge\n, fixed issue with gerunds, error-proofed plurality checks @FeatherWriter Sorry about all the trouble you went through. I tend to take these things for granted. I'm sure you're not computer-illiterate . File size is an issue. My program isn't exactly the most efficient one on the block right now because of the resource-intensive nature of the replace() function, and the sheer number of times that it's called. It should be able to chew through any file size if you give it time, although I would suggest breaking something like the WoK into more manageable chunks so that you don't have to wait so long that you're computer goes unattended should something start to burn. I wouldn't suggest large-scale transliteration just yet, either, since I'm sure that we still have some letter combinations to work on. If you want to help, you could transliterate a few pages of the WoK to start with and do a close reading for accuracy. Be sure to get rid of any apostrophes and replace exclamation points and question marks with periods, though. @Turos -Thank you for helping people out there. P.S. And, as FeatherWriter said, the period character is supposed to have a line as well. EDIT: I just ran it on the full text of the Odyssey, and it took just about 7 minutes. Stats: 597165 characters, 594 KB. I was running at 50% resources the entire time, with no unmanageable heat.

First of all, you should change the file extension to .java instead of .txt. The forum hates me, so it doesn't let me upload java files directly. This site shows you how to run a java file off of Command Prompt. I'm using a free program called BlueJ to compile and test the code because it's what I'm used to, but Command Prompt works fine for Turos. Put the .txt that you want to convert in the same directory and then run the file. When prompted, enter the full file name (I've been testing with Test.txt, if you can believe it). Any and all feedback would be greatly appreciated. I've been staring at this thing too long, so have a very tenuous grasp on reality when it comes to any flaws.

Yes, double post, but I think this one deserves it's own post. I added some debugging tools at the end of the main() and replaceLetters() functions, and used them to derive a large portion of the rules for 'e' modifiers at the end of words. I also moved the c-replacements to the bottom of the replaceLetters() function, so we'll see what happens there. As always, comments from a fresh set of eyes are more than welcome.

Yeah, I'm a rather big offender in this regard. EDIT: A good post is never truly finished. EDIT 2: So such a feature would be useful.

.it uorX alToh Te size Te forum goes up to is a bit underuelming "It works although the size the forum goes up to is a bit underwhelming." EDIT: And now the size has been fixed in the new version. And now they overlap vertically.

That's my bad. I've just revamped the period-moving process. Also, you're not supposed to be using any punctuation besides periods, oh Alethi-writer. We have a stable (to my knowledge) version now, and I need to incorporate some mildly significant changes to the existing code, so here's an attachment/spoiler of the newest version. EDIT: Small change to allow for no-period headers and generalize twistRight(), as well enable proper replacement at the end of the file. EDIT 2: Smaller change to make sure ces->seez works even if S->Z conversion happens. Also changed name from "AlethiTranslator" to "AlethiTransliterator." EDIT 3: Made a few changes to the "sc" conversions. EDIT 4: Misc. fixes, tinkering with "sc" and s->z conversions EDIT 5: Added more specifications for words beginning with 'c'.

Let's not get too far ahead of ourselves! We probably still need some detail work, especially on the -gh's, so we might want to bring in a fresh set of eyes to double check everything. That's why I moved all of the "necessary" changes down to the bottom: so I wouldn't have to worry about stuff like not putting W's into phonetic spellings. On that note, I did forget to throw in a simple w->u conversion if all else failed. You're welcome, once again. It's no problem. Have you been able to actually use the program? 1.3.6.2 (I might want to amend my naming protocols. . .), end of segment:

Thanks for the triple plus-sing, and done, although "tow" will just have to cry in a corner by itself for now. I think I may have implemented all of this correctly. It's a thorny issue though, so keep a sharp eye on it. Towards the end of my changes, I thought to look at Dictionary.com. It's phonetic spellings are quite helpful. It wouldn't have occurred to me otherwise to use y so freely. I should probably take a few minutes to look over all of my changes again, but I need a break, and want to get some feedback before I get too caught up in a Kurkistan-only thought loop. Changed "eses" to "seas." Don't know where my mind was there. The Z problem is difficult. There are some soft S's in normal words, like Peas, that would be compromised. How about ses -> sez? My goal for the program is just to enable your implementation of the font to the largest extent possible, so I think that a personalized version might undermine that a bit . I'll let you know if there are any irreconcilable issues, but I haven't found any yet. Thanks for the offer. I did underestimate those -gh's. That section definitely needs work. 1.3.6.1 code:

Double post ftw! Looking more in-depth at the problem of systematized transliteration, I'm considering taking this on as a mini-project. That 'e' modifier is annoying. From this point on, I'll only update the replaceLetters() function, where the replace() function is formatted (Entire text, Section to Replace, Replacement). Barring mistakes and inefficiencies in the original code, of course. P.S. You guys get to watch (and hopefully contribute to) my process, as opposed to me sitting alone in the dark for a few days and then throwing a list at you out of the blue (dark blue, presumably).

I didn't have much to do today, so I whipped up a quick Java program to convert from normal text assuming Roman letters into Turos's convention. An example of this would be converting from "This is a test. Chh." into ".Tis is a test .Ch" EDIT: And then just copy the modified text into a word document for font-ificatin. I'll spoil the code here as well as attaching a text file (the forum doesn't like .java files). I haven't gotten my hands dirty actually coding in a while, so I apologize for any mistakes/laughable inefficiencies present in this first iteration. I invite anyone who knows what they're doing to suggest/submit improvements. Instructions: Place the text you want to transliterate into a .txt file. It's easiest to put it in the same directory as the java file is being run from, with any sub-directories requiring that a path be given. Simply type in the file name (i.e. Example.txt) and a new file called Alethi_<YourFile>.txt will be created in the same directory as the original. EDIT: Sorry, didn't see you there for some reason. The reason I ask is because this (admittedly not the Coppermind) wiki provides a chart that shows two different shapes for "U." EDIT 2: Tell a lie! that other "U" is a V. Any other specifications you can think of besides c->k, ck->k would also be useful. Do you know what's going on with "Wh", "gh," etc?

Nice. What about J sounding like a Y? Also, how do you differentiate long and short U's?

No worries. I'll be the same in a week when mine starts up again.

That statement is possible evidence against the possibility of objects partially transiting the boundaries of time bubbles, but does not apply to how it would function if partial transitioning exists. If it is the case that objects, in their entirety, retain their velocity perfectly upon entering/exiting time bubbles, then my own theory does not work. I hold that they do not, as whole objects, maintain their initial velocity. Each constituent part attempts to accelerate/decelerate to its proper velocity in its new space-time, but is still attached to the rest of the object. So, to summarize, my own theory holds that we do not, in fact, know that objects, in their entirety, maintain the same velocity before and after transitioning. In fact, I hold that some change in velocity is necessary for anything larger than whatever the "minimum transition size" is which does not have exactly equal proportions of it's mass entering and exiting a time bubble.

We'll just have to agree to disagree about the entire object transitioning at the same time. I believe that it doesn't make sense for an arbitrarily defined "object" to transition all at the same time after some proportion of its mass exits/enters the bubble. Given my assumption of partial transition, though, I believe that I've shown that the energy gained from exiting cadmium bubbles can be harnessed, looking both at my example in the Assumptions section of my OP and at my initial response back in the original thread. To reiterate, the energy is gained when the portion of an object outside of the bubble attempts to accelerate to its "proper" speed, but is still attached to a segment inside the bubble. Because of this attachment, a large part of the energy which the bullet, in this case, gains from the perspective of the section inside the bubble is fed back into the bulk of the bullet. To use a simple example, let's picture a 2 meter pole traveling at 1 m/s within a cadmium bubble with a compression factor of 10. This magical pole transits the edge of the bubble 1 meter at a time. We will assume an outside perspective at all times. Step 1: Entire pole within bubble, traveling at 0.1 m/s. Step 2: First half of pole exits bubble, attempts to travel at 1 m/s while rest travels at 0.1 m/s. Step 2.5: Energy divided evenly between halves, both travel at (1 + 0.1)/2 = 1.1/2 = 0.55 m/s Step 3: Second half of pole exits bubble, attempts to travel at 5.5 m/s while rest travels at 0.55 m/s Step 3.5: Energy divided evenly between halves, both travel at (5.5 + 0.55)/2 = 6.05/2 = 3.025 m/s This 3.025 m/s compared to the 1 m/s that we would expect the pole to be traveling if we simply dropped the bubble without any transiting of the edge. I had to ponder this a bit to be sure I was actually right, but this math makes a fair amount of sense to me. Note: I changed the example in the Assumptions section as "Edit 2." Thanks for pointing that out, and I can now understand your confusion. EDIT: I'm not positive about how the energy transfer should work between "spaces." It seems as if 0.1 m/s worth of energy from real space should be enough to accelerate an equal mass by 1 m/s within a x10 cadmium bubble, but I may be reading it wrong.

We've seen allomancers simultaneously Push/Pull on objects directly in front of and behind them, a la Vin opening the Kredik Shaw storage cache/Well by pulling on the door and an object behind her at the same time. The "blue lines" that allomancers see are probably just a mental crutch and/or make it easier to differentiate targets. I doubt that Inquisitors are limited to the normal human field of view. All iron/steel allomancers can probably still "sense" metals out of their line of sight, especially when they already know it's there.

Thanks for going through the trouble of fixing it. I would feel sad if I had a minus rep.

It's my understanding that it takes some as-yet-unknown subtlety on the part of a Seeker to detect Feruchemy, so you might be better off with a more active Allomantic power, like, say, Brass to lull suspicions. Electrum to look out for sudden encounters? Actually, I really like electrum for this. An infiltrator could determine a signal to make (to themselves) ahead of time whenever they're spotted. Maybe even make it more complex to give details. Electrum/Copper? Electrum/Chromium?

I do believe that Dahak is referring to the total number of spikes for all inquisitors, using Marsh's spikes as an upper limit.