Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
Double sharp, Oschkar, and I have been developing a spreadsheet out in the Cloud, which can automatically generate descriptions of various Day-Gravity-Water based coherent metrologies, utilizing Quantitels, Systematic Numeric Nomenclature, and alphanumeric encoding of alternate radixes. We are developing this spreadsheet (and at least one variation on it) using Google Sheets, storing it out on Google Drive, so that it is accessible to anyone with a browser and the link to it.

Unfortunately, this conversation has ranged over two other threads, into which others have seen fit to inject unhelpful and mostly off-topic diatribes, incoherent nay-saying, arrogant self-promotion, sophistry, prejudices, dogmatism, and other irrelevant claptrap. I'm starting this thread in order to focus the conversation on actually developing the spreadsheet and supporting DGW metrologies.

(Note: In order to get this all into a single post, images have been turned into links, and emoticons have been replaced by plain text.)

EDIT: Additional Note: Much of the work on trans-dozenal digits supporting larger bases has been done in the Systematic Vigesimal Nomenclature thread.

From here (Oschkar actually got the idea to do this first):
Oschkar @ Dec 7 2017, 09:13 PM wrote:Anyway, have a spreadsheet:
From here (excerpt):
Double sharp @ Dec 7 2017, 11:58 PM wrote:
Oschkar @ Dec 7 2017, 09:13 PM wrote: Anyway, have a spreadsheet:
Brilliant! I think this is exactly what we've been talking about at Wendy's thread:
icarus @ Dec 5 2017, 04:07 PM wrote: I also think that a true octonary primel system would be interesting as a study (but so would many other studies). Further, I think an attempt at writing an algorithm to automatically generate a primel-style system - true to its tenets and given objectives that the user could set (i.e., based on gravity, etc.) would be very interesting if possible. It would be a massive demonstration of the flexibility of the directives Primel sets up.
Of course, this is just a starting point.

...
From here:
Double sharp @ Dec 8 2017, 04:03 PM wrote:
Oschkar @ Dec 7 2017, 09:13 PM wrote: Anyway, have a spreadsheet:
Some interesting possibilities for timels in some bases of interest, found after experimenting with Oschkar's very useful spreadsheet:

Quinary: 5^-9 (timel 44.24 ms, lengthel 19.17 mm, massel 7.048 g); 5^-8 (timel 221.2 ms, lengthel 479.3 mm, massel 110.1 kg)

Senary: 6^-8 (timel 51.44 ms, lengthel 25.93 mm, massel 17.42 g); 6^-7 (timel 308.6 ms, lengthel 933.3 mm, massel 813.0 kg). Oschkar's senary Ashtrian metrology uses 6^-8.

Septenary: 7^-7 (timel 104.9 ms, lengthel 107.8 mm, massel 1.254 kg)

Octal: 8^-7 (timel 41.20 ms, lengthel 16.63 mm, massel 4.599 g); 8^-6 (timel 329.6 ms, lengthel 1.064 m, massel 1206 kg). My Xing metrology uses 8^-6.

Nonary: 9^-6 (timel 162.6 ms, lengthel 259.0 mm, massel 17.37 kg)

Decimal: 10^-6 (timel 86.40 ms, lengthel 73.14 mm, massel 391.2 g). This is essentially Donald Sauter's system, although he uses slightly different values for gravity and water density, making it a little off.

Undecimal: 11^-6 (timel 48.77 ms, lengthel 23.30 mm, massel 12.66 g)

Dozenal: 12^-5 (timel 347.2 ms, lengthel 1.181 m, massel 1648 kg); 12^-6 (timel 28.94 ms, lengthel 8.203 mm, massel 552.0 mg). The first is Sunny's metrology; the second is Primel.

Tridecimal: 13^-5 (timel 232.7 ms, lengthel 530.5 mm, massel 149.3 kg)

Tetradecimal: 14^-5 (timel 160.6 ms, lengthel 252.9 mm, massel 16.17 kg)

Pentadecimal: 15^-5 (timel 113.8 ms, lengthel 126.8 mm, massel 2.040 kg)

Hexadecimal: 16^-5 (timel 82.40 ms, lengthel 66.52 mm, massel 294.3 g)

Octodecimal: 18^-5 (timel 45.72 ms, lengthel 20.48 mm, massel 8.595 g)

Vigesimal: 20^-5 (timel 27.00 ms, lengthel 7.142 mm, massel 364.4 mg)

Tetravigesimal: 24^-4 (timel 260.4 ms, lengthel 664.4 mm, massel 293.3 kg)

Trigesimal: 30^-4 (timel 106.7 ms, lengthel 111.5 mm, massel 1.385 kg). This one might prove interesting as the smallest number with 3 prime factors, and thus the one that manages to keep as many nice sizes as possible.

Sexagesimal: 60^-3 (timel 400.0 ms, lengthel 1.568 m, massel 3852 kg). I think Oschkar did a sexagesimal Ashtrian metrology some time back; I can't find the details, but given the ridiculous results of all other choices, this must be it.

Tetroctogesimal: 84^-3 (timel 145.8 ms, lengthel 208.2 mm, massel 9.024 kg)

Centovigesimal: 120^-3 (timel 50.00 ms, lengthel 24.49 mm, massel 14.69 g). Likewise, this is Oschkar's current centovigesimal Ashtrian.

Trecentosexagesimal: 360^-2 (timel 666.7 ms, lengthel 4.354 m, massel 82565 kg). Definitely monumental, but might work out.

Dumillequingentovigesimal: 2520^-2 (timel 13.61 ms, lengthel 1.814 mm, massel 5.965 mg). Okay, this is starting to get a little ridiculous.

Base 720720 (skipping over lots of SHCNs until we get one that actually gives one reasonable power): 720720^-1 (timel 119.9 ms, lengthel 140.8 mm, massel 2.791 kg). Surprisingly convenient units for a completely ridiculous base.

Base 1441440 (the last SHCN that actually works out): 1441440^-1 (timel 59.94 ms, lengthel 35.20 mm, massel 43.62 g). Okay, I'll stop now before I accidentally recreate The Funniest Joke in the World.
From here:
Kodegadulo @ Dec 8 2017, 11:49 PM wrote:Here's a couple more spreadsheets. Oschkar beat me to it, and his does an excellent job, but I was working up something a bit more elaborate. (smiley)

DGW-Primel.xslx
DGW-Xing.xslx

These are actual Excel 2016 workbooks, using the version for Windows, including use of the BASE() function (which supports the "Computerese" base style using uppercase letters, up to trinilial base). Perplexingly, this function does not seem available on the Mac OS versions of Excel I've had access to in the past.  I haven't played with Google Doc's spreadsheet software much yet, so I don't know how much of this could be reproduced there.

On the main Metrology tab of these docs, the first section defines a few parameters:

http://i67.tinypic.com/2i135hx.png

You can enter a different (decimal) number for the Radix and it will automatically generate an appropriate RadixAbbrev based on the "Computerese" base annotation convention, which is used later to mark based values e.g. 100[8] = 54[C].  It also automatically generates the appropriate Radix Syllable for use in Systematic Nomenclature prefixes, but this is somewhat limited in that I really only have worked up decent syllables for octal ("os"), decimal ("des"), unquadral ("tes") and unoctal ("vig").  We would have to do something creative to provision more bases with such syllables. Dozenal or unqual base of course omits such a syllable. I haven't yet worked this up so it could handle Ashtrian's unqua·decimal prefixes.

There is a drop-down selector to choose the timel for your metrology:

http://i67.tinypic.com/dxlc5.png

This then automatically populates the next section in an appropriate way:

http://i64.tinypic.com/9puopf.png

There are also drop-downs to select from various choices for accelerel:

http://i68.tinypic.com/30kszti.png

and for massic·heatability (Joule constant):

http://i63.tinypic.com/akap0g.png

Excel forces me to sort these alphabetically in order for the selection-and-lookup to work properly, but I wanted them sorted from smallest to largest values, so I had to prefix each name in these lists with these numeral indexes.

The bulk of the Metrology tab then lists the various quantitels and their values in coherent SI units. For each generic quantitel, it generates the target-system's quantitel with brand mark as well as the pronunciation using the brand prefix.

In some cases, you can see other unit equivalences to the right of the SI coherents, and some of those are even implemented as drop-down choices, so you can play with different customary units. I've also fleshed out some systematic scalings of the quantitesl.  I haven't finished doing these things for all the quantitels yet, but I may add more.
From here:
Kodegadulo @ Dec 8 2017, 11:50 PM wrote:Continued from previous post (since number of images is limited):

http://i66.tinypic.com/xpr38.png

http://i64.tinypic.com/1609382.png

http://i66.tinypic.com/288cv49.jpg
From here:
Oschkar @ Dec 9 2017, 02:09 AM wrote:
Kodegadulo @ Dec 8 2017, 11:49 PM wrote: I haven't yet worked this up so it could handle Ashtrian's unqua·decimal prefixes.
Come to think of it, multiple-subdigit prefixes would be very rare in Ashtrian, and would indicate that the scales being dealt with are so large or so small that they're far beyond what is intuitive. Even the masses of galaxies and subatomic particles are easily covered by the "zen"-series of prefixes in your spreadsheet.
From here:
Double sharp @ Dec 9 2017, 03:01 AM wrote:For radix syllables, I suggest continuing the use of argam from "tesqua, tescia":

2: bitqua, bitcia
3: tritqua, tritcia
5: kinqua, kincia
6: senqua, sencia
7: seffqua, seffcia
8: osqua, oscia
9: novqua, novcia
a: desqua, descia
b: ellqua, ellcia
c: qua, cia
d: thysqua, thyscia
e: frasqua, frascia (Oschkar's)
f: mandqua, mandcia (from German Mandel, because otherwise "trick" is to close to "tri" for comfort)
g: tesqua, tescia
h: zotequa, zotecia
i: dynqua, dyncia
j: axqua, axcia
k: scorqua, scorcia
l: tressqua, tresscia
m: dellqua, dellcia
n: florqua, florcia
o: caxqua, caxcia (because dex collides with dec)

There's always Oschkar's and my digit roots if you like one-syllable names for everything:

0: nil
1: un
2: bi
3: tri
5: pent
6: hex
7: sept
8: oct
9: enn
a: dec
b: lev
c: zen
d: cist
e: fed
f: gint
g: kix
h: mipt
i: ruct
j: vinn
k: ald
l: ift
m: wel
n: xor

albeit these only achieve the low bar of starting in different letters, not of sounding like anything familiar, except for c. d might be from argam "thise". There are a few derivation principles, of course: e to j are generated by looking at 4 to 9, keeping the ending, changing the vowel to the next one in the English alphabet, and changing the initial consonant to the earliest possible one. But other than that it is mostly ad-hoc, which I will be the first to admit for my supplemented m and n, which I created just to push this contraption to the double dozen.
From here:
Oschkar @ Dec 9 2017, 06:13 AM wrote:
Kodegadulo @ Dec 8 2017, 11:49 PM wrote: These are actual Excel 2016 workbooks, using the version for Windows, including use of the BASE() function (which supports the "Computerese" base style using uppercase letters, up to trinilial base). Perplexingly, this function does not seem available on the Mac OS versions of Excel I've had access to in the past.  I haven't played with Google Doc's spreadsheet software much yet, so I don't know how much of this could be reproduced there.
Don't have access to Excel 2016 easily. I can confirm that this is completely broken in Excel 2010, but it works in LibreOffice 5.3.1.2 with one bug: because the cell for the dozenal radix syllable is totally empty, it gets converted to "0" upon concatenation. This can be fixed easily by adding a function that evaluates to the empty string in cell D13 of the DigitRoots sheet.

Anyway, you’re dropping “concentration” and “solventage”? I’ll have to fix my Thermodynamics and Chemistry file then...
From here:
wendy.krieger @ Dec 9 2017, 07:14 AM wrote:They look nice, but some comment is useful.

1.  The 'convert to' units are the ones you intend to get accustomed to.  The implication here is to translate the DGW system to SI, means you want the people to think in SI, and regard DGW as a side issue.  All of my versions are *common=target style.

2.  It doesn't handle side-by-side comparisons of different systems.  I've done some of these in Lotus 1-2-3 + Allways which handles several columns side by side.

3.  It doesn't handle x*b^y even, such as 2D5 (tgm) or 6*60^3, or 4d5 or 6d5 or 8d5, all of which have been mentioned in other threads.  The current versions handle only b^n.
From here:
Oschkar @ Dec 9 2017, 07:26 AM wrote:
wendy.krieger @ Dec 9 2017, 07:14 AM wrote: 3.  It doesn't handle x*b^y even, such as 2D5 (tgm) or 6*60^3, or 4d5 or 6d5 or 8d5, all of which have been mentioned in other threads.  The current versions handle only b^n.
TGM was a one-off thing. There’s nothing intrinsically coherent about 2×125 that would make it suitable for Kodegadulo’s spreadsheet; it was selected in order to retain the hour and the foot, the closest things that we currently have to dozenal units. And in any case, TGM units are also Primel auxiliary units, though they are not coherent dozenal powers.

By the way, “4d5” doesn’t mean “four times dozen to the fifth power”; it means “roll four five-sided dice and add their values”, which is somewhat awkward for a number of divisions of the day.
From here:
Kodegadulo @ Dec 9 2017, 11:10 AM wrote:
Oschkar @ Dec 9 2017, 06:13 AM wrote:Anyway, you’re dropping “concentration” and “solventage”? I’ll have to fix my Thermodynamics and Chemistry file then...
Eh, I don't know if you need to do that. I'm experimenting. The terminology around concentration of solutions is awkward. Even the Metric/SI crowd has found it awkward, punting with nonsense coinages like "molar" (but not about dentistry) and "molal" (not about anything).

My original thought was to reserve concentrat(ion)el for "number concentration" (number of individual items (dimensionless) per volumel of solution); and qualify it as substancel•concentratel for "substance concentration" (amount of substance per volumel of solution, traditionally referred to as "molarity" in Metric).

Well, substancel•concentratel is a mouthful, bordering on being just a derived-unit phrase. Perhaps we could used substancentratel as an awkward contraction, but it's still a mouthful. But then it occurred to me that volumic• is (almost) a symonym for concentrationel, so I'm trying that out here. I say "almost" because concentrationel conveys the nuance "of solution".

I notice that you've decided to reserve unadorned concentrat(ion)el to mean specifically "substance concentration" and put aside other measures such as "number concentration" and "mass concentration" as not important enough to worry about. (Similarly for solvent(ag)el.) That becomes another terminology experiment, which is fine. Perhaps yours is the better ... solution. (smiley)

EDIT: What's truly annoying is that the "molar" is not 1 SI•substancel (mole) per one SI•volumel (cubic-meter), but rather a thousand times more concentrated: one mole per liter (milli•SI•volumel).
From here:
Double sharp @ Dec 9 2017, 03:55 PM wrote:Translation for the perplexed: 1 molar is 1 mol/dm3, while 1 molal is 1 mol/kg. While these are commonly encountered quantities, and it's nice to have short names, it would be nicer to have more sensible short names indeed. Even the unit abbreviations are less confusing; at least they don't sound so alike!

I suspect there are two reasons for the issue you raise. First of all is that 1 mol/m3, the millimolar, is a rather low concentration for most laboratory purposes, the cubic metre being so large. (Except for radiochemistry, when it becomes an impractically high concentration: most work on astatine is done with concentrations ranging in the nanomolars and picomolars.) The second reason is that the metric system from the beginning has had this factor of a thousandth built into it: the original French metric system had the litre (the cubic decimetre) as its volumel for liquids and the stère (the cubic metre) as its volumel for solids. From this was derived the gram as originally the mass of a millilitre of water, creating yet another factor of a thousandth. So metric awkwardly has this built into it.

Really, all of this could have been resolved by using the decimetre as the SI-lengthel instead of the metre. We'd then be rid of all those factors of a thousand, and all that would have been needed is for the French to aim for one hundred-millionth of the length of half a meridian instead of one ten-millionth (and then got it slightly wrong, but never mind that). Don't get me wrong, SI is a good system and it works, but there are a few things like that that could have been done better. Something had to be first, though, and SI is very good for a first attempt at a consistently-based metrology put into practice. It still wouldn't bring in gravity, though; if you're keen on that, you'd be starting with the microday and have a lengthel of 73.14 mm and a massel of 391.2 g, which isn't half bad either. In fact, it is just Donald Sauter's system.

Now, following on Sauter's remarks, it is true that while the "water" part of the metrology is constant for us, the "day" and "gravity" parts will change when we colonise another planet. So now I have a lightly edited version of Oschkar's spreadsheet on my computer with 86400 defined in cell I5 (with H5 reading "Day"), and the formula in cell B1 altered to match. So, with that addition, you can just change the day length and gravity length as you like for other celestial bodies. To make your Martian colonists happy, just change 86400 to 88775.244 and 9.79758271961639 to 3.711. The lower gravity indeed makes for a nice Martian xing-lengthel of 425.6 mm and Martian xing-massel of 77.09 kg. On Venus, (use 20997360 and 8.87 respectively this time), you should probably start with the ununosciaday instead of the hexosciaday - albeit the concept of a Cytherean day as a fundamental reality is a little suspect even after terraforming for the colonists, considering that you can't possibly adapt to a 243.025-day sleep-wake cycle like you can for the Martian day.
... skipped more side-conversations about con-words and concentrations ...

From here
Kodegadulo @ Dec 16 2017, 06:32 AM wrote:Double Sharp, if you want to complement the octal Xing metrology with a sibling tessal (unquadral, hexadecimal) metrology based on the penttescia·day, I suggest giving it a name based on the Mandarin for "compass", either &#25351;&#21335;&#38024; zh&#464; b&#283;i zh&#275;n, or &#25351;&#21335;&#38024; zh&#464;nán zh&#275;n. Perhaps reduced to just the final syllable &#38024; zh&#275;n "needle".  This would be an oblique reference to the tess (fourzeen, sixteen) major directions of the compass which the "needle" can point to: N, NNE, NE, ENE, E, ESE, SE, SSE, S, SSW, SW, WSW, W, WNW, NW, NNW.

So let's call it the "Zhen" metrology, with brand prefix "zhen".  For a brand mark, I suggest &#10042;, Unicode Character 'SIXTEEN POINTED ASTERISK' (U+273Ax)

Take a look at this Google Sheets spreadsheet I've just put together.

This spreadsheet documents 6 existing Day-Gravity-Water systems: Primel, Tertiel, Phasic, Pendlebury, Xing, and Zhen.  (I haven't tried to tackle Ashtrian yet, as mixed unqua·decimal base adds a whole other level of complexity.)

On the main "Metrology" tab, the "Metrology Selector" cell (B1) lets you easily choose among them, and the rest of the tab automatically recalculates. It pulls selections for each of these metrologies from the "Metrologies" tab, which in turn contain selector cells pulling choices from other tabs about specific aspects, including "Gravities", "Densities", "Calories", "Impedances".

For experimenting, I've included a seventh metrology called "Adhoc". Fiddle with the selections in its row of the "Metrologies" tab to try out different possibilities.  If you identify additional interesting metrologies that this spreadsheet can model, I'd be happy to include them above "Adhoc" in the "Metrologies" tab.

This spreadsheet generalizes the "Day-Gravity-Water" idea to allow simple reciprocal fractions of the day to act as the "Period" from which a timel may be derived as a pure power of the selected radix. (See the "Periods" and "Period Powers" tabs.) I've included the customary names for such fractions we've identified (clock, shift, phase, watch, vigil, etc.) as well as some of your binary series (whiling, serenade). For other fractions, I just punted and constructed names from Systematic Nomenclature digit roots + "id" as a common suffix to mean "fraction of a day" (e.g. pentid, septid).

Incidentally, I found a way to name digit roots all the way up to trinilial base. (See the "Digit Roots" tab.)  As before, I introduce the syllable "zen" to cover the zeens (dozen-somethings). But I thought another syllable might be introduced to cover the twenzies (two-dozen-somethings) in a similar way.  I'm trying out "yon" (short for "yonder") as that syllable, signifying that the digits in that tier are "farther off" than the zeens.
From here:
Double sharp @ Dec 16 2017, 07:31 AM wrote:"Yon" is an interesting idea, and so is "weg" for three dozen (hmm, is it because it is even more of ein langer Weg from the dozenal digits?). I likewise think "Zhen" is a good name for the sister metrology for hexadecimal, and &#10042; is indeed the obvious brand mark.

The trouble is that I don't like the idea of a hexadecimal metrology with single digit roots up to "zen" (c ), and then going "zenun" (d), "zenbi" (e), "zentri" (f). How exactly are we going to distinguish those from c1x, c2x, and c3x? And, really, what are we doing switching at twelve in a hexadecimal system, anyway? Eight might make more sense. Though that's not a very extensible solution because it leaves us no way to handle prime bases, or bases with no small enough factor pairs (e.g. 26).

"Yon" isn't a bad fit for d, I have to admit: it's on yonder side of the dozen, instead of the left side. So that takes care of tetradecimal, though hexadecimal seems to want additional roots for e and f. We might I guess coopt "weg" too for e (it also sounds like "week", and indeed means that in Frisian, which is half its length), and German Mandel immediately yields "mand" for f. But I also don't like having them mean one thing in a small base and another thing in a big base. Hmm. Omnibase also works, at the expense of being limited to base 36 as well, and having to make casing distinctive ("e" means 9, while "E" means fourteen).

Incidentally, even if you can't tackle centovigesimal Ashtrian, I would think that the old senary Ashtrian (link) would be a good addition, with &#10038; (U+2736x, "SIX POINTED BLACK STAR") as the brand mark.

(P.S. Senary Ash-Ashtrian has now been added, further to some chatting Kode and I have been doing on that spreadsheet about the issues raised above. Replacing "zen" and "yon" with "zennil" and "yonnil" solves the first problem, for example.)
From here:
Kodegadulo @ Dec 16 2017, 08:53 AM wrote:I've included various proposals for digit roots and their abbrevs for reference, but now I've set the ones in force in the spreadsheet to DS's proposal for transdecimals up to hexadecimal, and Omnibase above that: un (u), bi ( b ), tri (t), quad (q), pent (p), hex (h), sept (s), oct (o), enn (e), dec (d), lev (&#8467;), zen (z), yon (y), weg (w), mand (m), golf (G), hoat (H), ind (I), jul (J) ... etc. So up to the near human-scale we have single-lowercase-letter abbreviations, and only get awkward shouting uppercase when we choose to go more extreme.

I did add the Ash-Ashtrian metrology.  Note that I had set the Adhoc metrology to argam "zeffal" base (twenzeen, fourteen). I guess I should now call it "wegal" base. (smiley)  You can now see a little of what SNNw prefixes would look like, if you check the "Period Powers" tab.
From here:
Double sharp @ Dec 16 2017, 03:35 PM wrote:I just realised that we're missing one metrology: Donald Sauter's decimal DGW system, which starts from the microday, and uses the SI gravity. Unfortunately he doesn't seem to have specified it completely all the way to the electrical units, but after all, he says "We'll also need to choose new fundamental units for electric current, temperature, and luminous intensity, that work well with these, but I'll leave that to others when this idea gets rolling", so we can choose those for him. (EDIT: Added!)

And just like we came up with the laevo excuse-etymology for lev, we can use Greek to come up with one for yon: &#7984;&#974;&#957;, present active participle of the verb "to go", &#949;&#7990;&#956;&#953;. Clearly, we are going to the yonder side of the dozen! As for the Germanic weg and mand, I say that once you've accepted zen, from which the first letter comes from the French development of the second d in Latin duodecim, and the rest doesn't even come from a number but the Latin distributive suffix, they're not that much less desirable! It's difficult to argue with mand when there is essentially no choice for a non-decimal name for 15; and while weg is kind of cheating, I'm not sure we can do much better. Anything to do with "fortnight" already has a lot to do with fourteen, and for a number that is twice seven, obfuscating the seven-day week should work. If you can't come up with one good, solid etymology, the next-best thing is clearly to come up with many weak ones! (wink)

I suggest that the final c in "dec" and g in "weg" should not be allowed to soften among any circumstances; the roots should remain constant in pronunciation.

EDIT: While chatting we've come up with some other names for the other human-scale metrologies: senary is Ash-Ashtrian &#10038;, octal is Xing &#10050;, decimal is Sauter
&#9416; (since he wrote up the idea for decimal first, as far as I can tell), dozenal is of course Primel &#8242;, tetradecimal is Fleur &#10087; (because obviously seven needs a flowery mark for it being a special prime, so it might as well get the fleuron), and hexadecimal is Zhen &#10042;.
From here:
Kodegadulo @ Dec 16 2017, 06:32 PM wrote:I've opened up permissions on the Google doc so anyone with the link can edit it. This seems the only way to let people operate the drop-down selectors, as that's apparently considered an edit. This does leave the doc open to possible "hostile" edits, but Google Docs does maintain a version history, so anything destroyed or cobbled up by ill-advised parties ought to be recoverable. Suffice to say, I've made my own (secret) backup copy as well.

As DS has mentioned, the spreadsheet now supports Primel, Tertiel, Phasic, Pendlebury, Sauter, Ash-Ashtrian, and DS's Fleur, Xing, and Zhen metrologies.  The AdHoc metrology is for random play.

I'd like to see someone who fancies themself a "programmer" try to top that with some piddly command-line script that they wrote in some vintage 20thd-century scripting language who-knows-how-many decades ago and barely advanced in all that time. A script not actually provided, I might add, and of little interest to anyone even if.  Who wants to download, install, and learn how to run a scripting language?
From here:
Kodegadulo @ Dec 16 2017, 08:04 PM wrote:By the way, I originally got weg as a corruption of omega > wmega > wmeg > weg. But hey, if week > weg and/or way > weg work better, no problemo. As long as we don't call it a "wig" on the grounds that it's a "fur piece" into the transdecimals. (smiley)

P.S. The "Digit Sequences" tab now generates SNN multiplier, reciprocal, positive power, and negative power prefixes for the radix selected in the metrology. Note that each prefix embeds a "radix syllable" somewhere (except for single-digit multipliers and reciprocals). But if we combine a multiplier, reciprocal, and power into a single prefix, it may be fair to have only one radix syllable per combination.

Of course, dozenal SNNz prefixes don't take any radix syllable. But I never said I wasn't biased. (wink)
From here:
Double sharp @ Dec 17 2017, 03:08 PM wrote:One more feature that would be nice to have in this spreadsheet would be to be able to have not only "Period Powers", but be able to select any unit you want (timel, frequenciel, accelerel, velocitel, lengthel, etc.) from a drop-down list and have a bunch of powers come up, kind of like what I quoted from Oschkar in the OP. Then we can not only systematically see what the lengthel looks like, but also the unqualengthel, biqualengthel, ..., uncialengthel, bicialengthel, etc., and get even more of an insight of how all those units and the base slice up the quantities they are measuring.
From here:
Kodegadulo @ Dec 17 2017, 07:02 PM wrote:
Double sharp @ Dec 17 2017, 03:08 PM wrote: One more feature that would be nice to have in this spreadsheet would be to be able to have not only "Period Powers", but be able to select any unit you want (timel, frequenciel, accelerel, velocitel, lengthel, etc.) from a drop-down list and have a bunch of powers come up,
Great idea! I already have a lot of the machinery to support that. I'll put it together, soon... (smiley)
... skipped useless lecturing from David Kennedy ...

From here:
Double sharp @ Dec 18 2017, 10:04 AM wrote:
Double sharp @ Dec 16 2017, 03:35 PM wrote: EDIT: While chatting we've come up with some other names for the other human-scale metrologies: senary is Ash-Ashtrian &#10038;, octal is Xing &#10050;, decimal is Sauter
&#9416; (since he wrote up the idea for decimal first, as far as I can tell), dozenal is of course Primel &#8242;, tetradecimal is Fleur &#10087; (because obviously seven needs a flowery mark for it being a special prime, so it might as well get the fleuron), and hexadecimal is Zhen &#10042;.
For some strange reason the one character among these which I cannot see on my smartphone is the simplest: the circled S for Sauter. Nonetheless it is such a good fit that I don't think we need to replace it. (Kode's original suggestion of &#963; is I suppose an option, albeit a placeholder-ish one. I think &#960; for Pendlebury TGM makes a lot more sense, because it also references how it divides the day into two first and has a unit named the "Pi".)

I have thought about including the two "higher-natural scale" bases 18 and 20; by then the transition into rather shouty uppercase is rather inevitable. Vigesimal nevertheless seems to strand you between a rock (a massive massel) and a hard place (a tiny massel), and I need to go think up a few more brand marks for those. Not sure if doing these or the odd bases are worth it, though.
From here:
Kodegadulo @ Dec 18 2017, 01:11 PM wrote:
Double sharp @ Dec 18 2017, 10:04 AM wrote: For some strange reason the one character among these which I cannot see on my smartphone is the simplest: the circled S for Sauter. Nonetheless it is such a good fit that I don't think we need to replace it. (Kode's original suggestion of &#963; is I suppose an option, albeit a placeholder-ish one. I think &#960; for Pendlebury TGM makes a lot more sense, because it also references how it divides the day into two first and has a unit named the "Pi".)
Eh, I was considering the circled-S rather a placeholder too if we can come up with better. Although it does have the saving grace of being as easily distinguished from the succeeding text of the unit name, with no additional formatting, as any of the stars or dingbats or punctuation we've been using for brand markers. The disadvantage of something like the Greek &#960; for Pendlebury is that it is an ordinary letter and just washes together with any other text unless one takes pains to set it off by superscripting it, or interposing a center dot as I have in the spreadsheet. I'm thinking a circled P for Pendlebury might be a better bet. Would that there were a circled &#960;! And a circled &#966; for Phasic, for that matter... (smiley)
... skipped more useless lecturing from Daniel Kennedy ...

From here:
Double sharp @ Dec 18 2017, 02:59 PM wrote:
Kodegadulo @ Dec 16 2017, 06:32 AM wrote: (I haven't tried to tackle Ashtrian yet, as mixed unqua·decimal base adds a whole other level of complexity.)
Wait, why is this a problem? Can't we just treat it as pure base 120 with two-part names for the superdigits? So we could have digit roots "nil, un, bi, tri, quad, pent, hex, sept, oct, enn, undenil, undeün, undebi, ..., undeënn, bidenil, ..., enndeënn, decdenil, ..., levdeënn", just like you suggested for Systematic Unquadecimal Nomenclature. To concatenate the digits, I guess we'd have to use "nildenil, nildeün, nildebi, ... nildeënn" for the first ten digits instead, but I can deal with that.

You'd have to extend the "Digit Roots" table to 120, and have some means of marking them via radices (maybe my old "Double-Wide-with-a-Porch Digital" scheme using Japanese kana), but other than that I don't see why it should pose problems. Unless there is anything that requires base conversion into the target base, in which case indeed bases over 36 would likely prove a mess, but are there?

(Come to think of it, that would be a neat "reverse conversion" feature. Because to truly think in the new metrology, you shouldn't be thinking "a fleur-lengthel is about a quarter of a metre"; you should be thinking "a metre is about four fleur-lengthels", for instance.)
From here:
Kodegadulo @ Dec 18 2017, 03:41 PM wrote:
Double sharp @ Dec 18 2017, 02:59 PM wrote:
Kodegadulo @ Dec 16 2017, 06:32 AM wrote: (I haven't tried to tackle Ashtrian yet, as mixed unqua·decimal base adds a whole other level of complexity.)
Wait, why is this a problem? Can't we just treat it as pure base 120 with two-part names for the superdigits? So we could have digit roots "nil, un, bi, tri, quad, pent, hex, sept, oct, enn, undenil, undeün, undebi, ..., undeënn, bidenil, ..., enndeënn, decdenil, ..., levdeënn", just like you suggested for Systematic Unquadecimal Nomenclature. To concatenate the digits, I guess we'd have to use "nildenil, nildeün, nildebi, ... nildeënn" for the first ten digits instead, but I can deal with that.

You'd have to extend the "Digit Roots" table to 120, and have some means of marking them via radices (maybe my old "Double-Wide-with-a-Porch Digital" scheme using Japanese kana), but other than that I don't see why it should pose problems.
{c}

It's more involved than that, because mixed base requires specifying more than just one radix. The spreadsheet would need to make provisions for specifying multiple sub-bases making up a super-base, and in general a semi-arbitrary number of them, e.g. 6×a for sexagesimal aka hexal•decal [6×A]; or 10×a for unqual•decimal aka zenal•decal [C×A]; or 10×13×12 for unqual•untrinal•unbinal aka zenal•mandal•wegal [C×F×E] (remember that one? (smiey) ); and so forth. It's not impossible to work out, but as I said, it's another layer of complexity. Patience, grasshopper... (wink)

EDIT: On second thought, what I cooked up for shockal and hundal bases were rather one-offs, not really generalizable to any mixed base. They in fact rely on having particular radix syllables indicating the superbase itself, so that we can construct -shocqua/-shoccia and -longua/-longia. (In fact, for Ashtrian, Oschkar is moving to replace the latter bisyllabic constructs with the more Germanic monosyllables -hund/-deal. This is still in the spirit of SNN, but it does represent another one-off case adding to the complexity.)

What I had in mind for a truly generic superbase solution would be something like this, e.g., for zenal•trickal•zeffal (zenal•mandal•wegal??):

(Parenthesized portions are elements thatt might be always omitted, or always included, or optional either way, but I don't know what to decide about that yet -- looking for opinions. )

Multiplier Prefixes:
000[C×F×E] = dozterkzef•(nilnil)nili
001[C×F×E] = dozterkzef•(nilnil)uni
...
00D[C×F×E] = dozterkzef•(nilnil)yona
010[C×F×E] = dozterkzef•(nil)unnili
...
0ED[C×F×E] = dozterkzef•(nil)wegyona
100[C×F×E] = dozterkzef•unnilnili
...

BED[C×F×E] = dozterkzef•levwegyona
...

Note the need for an explicit radix syllable for dozenal (doz), distinct from zen used as a digit root.

Power Prefixes:
000[C×F×E]&#8593; = (nildoz-nilterk-)nilzefqua
001[C×F×E]&#8593; = (nildoz-nilterk-)unzefqua
...
00D[C×F×E]&#8593; = (nildoz-nilterk-)yonzefqua
010[C×F×E]&#8593; = (nildoz-)unterk-nilzefqua
...
0ED[C×F×E]&#8593; = (nildoz-)wegterk-yonzefqua
100[C×F×E]&#8593; = undoz-nilterk-nilzefqua
...

BED[C×F×E]&#8593; = levdoz-wegterk-yonzefqua
...
From here:
Double sharp @ Dec 18 2017, 11:56 PM wrote:Remember that one! I should say so, indeed!

Your solution for base c-on-f-on-e looks mighty promising, though I'd want some clarification about what it does when it crosses to 1'000{cfe} and onward. Are the superdigit boundaries explicitly called out? It's also not far from how I think of mixed bases: to me it's not so much "pure base 2520" as "a tetradecimal column overflowing into a pentadecimal column overflowing into a dozenal column".
From here:
Kodegadulo @ Dec 19 2017, 12:39 AM wrote:
Double sharp @ Dec 18 2017, 11:56 PM wrote:I'd want some clarification about what it does when it crosses to 1'000{cfe} and onward.
I think it would just be :

Multiplier: dozterkzef-un-nilnilnili

Pos. Power: unzef-nildoz-nilterk-nilzefqua

Not really calling out the superdigit boundary in either case, unless you interpret the hyphens as pregnant pauses.

Hmm ... I've got two opposing styles there: I've got the multiplier declaring the superbase once, and all-at-once, and you just have to keep a scorecard to figure out which subbase each digit gets. But I've got the power declaring the subbase of each digit along with it, and you just have to inspect the whole to deduce what superbase you have. Let's try using just one style at a time:

Test case:
1234{cfe) and
1234{cfe}&#8593;

Piecemeal subbases:

All-at-once superbase:

I think the all-at-once superbase style is probably the best.

----
Hmm ... I'm a bit dissatisfied that we've wound up having to cook up two distinct series of syllables for the same numbers, one used for digits and one for bases, with the latter unfortunately incomplete. I wonder if we could make this more parsimonious somehow...
From here:
Oschkar @ Dec 19 2017, 01:11 AM wrote:
Kodegadulo @ Dec 19 2017, 12:39 AM wrote: Hmm ... I'm a bit dissatisfied that we've wound up having to cook up two distinct series of syllables for the same numbers, one used for digits and one for bases, with the latter unfortunately incomplete. I wonder if we could make this more parsimonious somehow...
You might consider adding another monosyllabic particle that separates the base from the rest of the number.

From here:
Double sharp @ Dec 19 2017, 02:01 AM wrote:{a}

I'm not sure if Kode intends to do this for just mixed bases or for all of them. The latter runs into the problem that you can't keep the empty base particle for dozenal, because then "unquadqua" is ambiguous between 1216 and 41. And if we keep it that way we need to distinguish that what sounds like the last digit is always actually the radix syllable.

I agree that it's not particularly elegant to have two such parallel series, but it does seem to do the job with the minimum amount of superfluous ballast if we desire backward compatibility. Though the need to switch bases around seems to be forcing all those mixed bases to look as unfamiliar and clunky as they really are.
From here:
Double sharp @ Dec 19 2017, 02:19 AM wrote:
Kodegadulo @ Dec 18 2017, 01:11 PM wrote: Eh, I was considering the circled-S rather a placeholder too if we can come up with better. Although it does have the saving grace of being as easily distinguished from the succeeding text of the unit name, with no additional formatting, as any of the stars or dingbats or punctuation we've been using for brand markers. The disadvantage of something like the Greek &#960; for Pendlebury is that it is an ordinary letter and just washes together with any other text unless one takes pains to set it off by superscripting it, or interposing a center dot as I have in the spreadsheet. I'm thinking a circled P for Pendlebury might be a (smiley) )
I've noticed this too; Greek doesn't really stand out in Latin text unless you do something funny with it. Neither does Cyrillic. Of course, this is only to be expected given the scripts' common heritage; they even have many letter shapes in common. A second problem is that these are characters used in actual scripts; if we were to translate all this text on the metrologies into Greek or Russian, for example, these characters wouldn't stand out at all.

Still, since the Sauter and Pendlebury metrologies are here being named after their inventors (there's nothing better for Sauter, though why not just call Pendlebury's metrology "TGM"?), I would think that some sort of acknowledgement of that ought to be made in their brand marks. We are after all running out of the really apt and relevant dingbats; even the fleuron for the tetradecimal Fleur metrology is a bit of a cheat, if a happy one.
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
Another split necessary:

From here:
Double sharp @ Dec 27 2017, 07:03 AM wrote:Hmm, I see the centovigesimal Ashtrian has disappeared from the current version of the DGW spreadsheet. Oschkar, did you delete it? Was it not working properly? From what little I saw earlier the front page of the spreadsheet, including the back-conversions into decimal, was working splendidly; it's just that some of the less important sheets at the back weren't.
From here:
Oschkar @ Dec 27 2017, 07:05 AM wrote:
Double sharp @ Dec 27 2017, 07:03 AM wrote: Hmm, I see the centovigesimal Ashtrian has disappeared from the current version of the DGW spreadsheet. Oschkar, did you delete it? Was it not working properly? From what little I saw earlier the front page of the spreadsheet, including the back-conversions into decimal, was working splendidly; it's just that some of the less important sheets at the back weren't.
Kodegadulo branched the spreadsheet so that I can play around with it until it’s ready for deployment.
From here:
Double sharp @ Dec 27 2017, 07:10 AM wrote:
Oschkar @ Dec 27 2017, 07:05 AM wrote:
Double sharp @ Dec 27 2017, 07:03 AM wrote: Hmm, I see the centovigesimal Ashtrian has disappeared from the current version of the DGW spreadsheet. Oschkar, did you delete it? Was it not working properly? From what little I saw earlier the front page of the spreadsheet, including the back-conversions into decimal, was working splendidly; it's just that some of the less important sheets at the back weren't.
Kodegadulo branched the spreadsheet so that I can play around with it until it’s ready for deployment.
Oh, I see. Then I'd better also hold off on doing anything to it myself, except maybe toggling the selectors, so that nothing is lost later.

Looking forward to seeing a working version! What little I saw of it earlier looked really spectacular. BTW, would it be possible to get the base-120 values in decimal-coded form as well as (or, in a pinch, instead of) pure-base form? It may be my suggestion, but I can't remember which digit corresponds to which kana without counting up from one of the few I remember.
From here:
Kodegadulo @ Dec 27 2017, 04:25 PM wrote:
Double sharp @ Dec 27 2017, 07:10 AM wrote:
Oschkar @ Dec 27 2017, 07:05 AM wrote: Kodegadulo branched the spreadsheet so that I can play around with it until it’s ready for deployment.
Oh, I see. Then I'd better also hold off on doing anything to it myself, except maybe toggling the selectors, so that nothing is lost later.

Looking forward to seeing a working version! What little I saw of it earlier looked really spectacular. BTW, would it be possible to get the base-120 values in decimal-coded form as well as (or, in a pinch, instead of) pure-base form? It may be my suggestion, but I can't remember which digit corresponds to which kana without counting up from one of the few I remember.
I didn't mean to squelch usage of Oschkar's version, I was just concerned that it was adding complexity to.the point that Sheets might start performing badly, so I wanted to stabilize my previous version. The branch I made is configured for edit by anyone with the url. Oschkar can you please post the url here? I don't have it on my phone.
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
Update: I believe the link to Oschkar's working copy is here. (Oschkar can you confirm that's the one you're currently using?)

I've made backup copies of my original and Oschkar's version.
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

 Posts 775
Dozens Disciple
Oschkar
Dozens Disciple
Joined: Nov 19 2011, 01:07 AM
It looks like it is.

I'll try to program a subdigit-encoded base conversion function sometime soon, in order for the centovigesimal numbers to be readable at least. (If argam were available in Unicode, I'd use those, but 12-on-10 is more readable than kana, I think.)

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
@Oschkar, with your branch version, you have the liberty to break away from the coding techniques I used for the simple-radix metrologies. You can dispense with conditional code and just specialize it for Ashtrian. That may make the problem far less complex.

We could always try to merge the two coding techniques later, but at least you can make progress with implementing subdigit encoding in peace.
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
If the large bases require very different programming from the small ones (the cut-off being of course 36), then I'm wondering if it mightn't be better to put Ashtrian (centovigesimal) and Gesh-Ashtrian (sexagesimal) in their very own spreadsheet catered specifically for this kind of thing, instead of trying to do some complex conditionals contingent on the magnitude of the base. Then we could use that for the case of alternation, though I'm not sure if something like {6:14} would require very different programming from {6:10} or {12:10}. (It's still alternation; it's just that the digit roots need to be aware of what the sub-bases are meant to be.) Things like {12:15:14} may still be too complicated to do, of course.

2df7{g}: tes-bi-yon-mand-sept

1000{g}: tritesqua
1000{k}: tricaitpempqua, triscorqua
1'00'00'00{6a}: trisendesqua, trishocqua
1'00'00'00{ca}: tridozdesqua, trihundqua (trihund)
1'000'000'000{cfe}: tridozterkzeffqua

(I'm trying out cait for quaternary, based on cater-corner, with an attempt at spelling the long a without "silent e". Yes, I know it creates another homophone with a personal name, but we've already got tes for its square. I'm open to other suggestions.)

{ca}

In particular, while 1'00 is undozdesqua (unhundqua), 10'00 is decanundozdesqua (decanunhundqua). In principle the use of "shock" and "hund" instead of "sendes" or "dozdes" should imply a pure base, but in these small examples it doesn't matter. In order to avoid confusion if the number is long, initial zeroes should always be added to pad out the full superdigit.

So, how does that look?

 Posts 775
Dozens Disciple
Oschkar
Dozens Disciple
Joined: Nov 19 2011, 01:07 AM
I have to confess that whenever I actually read power prefixes with more than one subdigit, it’s usually a single syllable. I never actually use “X-de-Y-hund”, it’s always “dechund”, “levhund”, “zenhund”, and argam names after that. The Planck length, for example, is “one point twenty-six fifty-seven tessdail-ashter-lengthels”. I could get used to a new set of digit roots, like your “yon, weg, mand”, but I’d avoid using subdigits in the power prefixes as long as possible.

In practice, power prefixes above 24 would be so rare in Ashtrian that they might as well be missing, but they’re still needed to make the nomenclature fully systematic, which is where the subdigit-encoded prefixes would come in handy.

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
I also have to confess that when I think about how I'd do things for a sexagesimal metrology, it's exactly the same; there might not be enough letters in the alphabet to give sixty different roots, but the argam themselves serve reasonably, and I might even read them as exponents. So I wouldn't so much read "8 gesh-levshocciamassels" as "eight E minus ell gesh-massels", just like E-notation in decimal, with the base of sixty implied by the context; and the reason is that that's how the notation pretty much ends up if you don't want then to confusingly look like the unit is being raised to a power. I'd do the same thing in base 120; it's just that I'm too used to base 60, so base 120 always feels like it's wrong by a factor of two. (I know Wendy uses D for dozenal and H for centovigesimal, but I don't see why we need different symbols for each base. To me, 3E6 just means 3 times the 6th power of whatever base you are in.)

This systematisation might be welcome, of course, but I think what this is telling us is that beyond the human scale (16, maybe 18 and 20), pure argam names work out because it's too much work to memorise all those digits and a character corresponding to each of them.

P.S. Glad to hear you can get used to "yon, weg, mand", even though only the last is from me - it's just that Kode originally had the first two as cadex and exent respectively, and it was my idea to make them mean thise and zeff. Of course, this being his system, his extensions sound entirely natural, something I feel I only managed with "mand". I still think the "x" annotation for hexadecimal half-seriously implies a syllable starting with "x", following my tongue-in-cheek principle that many bad etymologies are as good as one good one, but I can't think of a serious-sounding one.

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
I agree that we should dispense with clipped intermediate syllables (like -de-) to mark the base of every digit (or even every other digit). After all, we don't say 1C2A4C5A6C7A, we say something like 12'34'56CA. It's already complex enough that we need one syllable-series for digits and another for radixes, we don't need an even more tortured one for these clipped forms. Chalk that up as one of my early ill-advised half-baked ideas that I've repented of since.

So all we really need is some way to vocalize those superdigit delimiters (apostrophes). But hyphens indicating slight pauses might be sufficient:

(Actually, I prefer to make that example 12'34'56C×A to emphasize that it's a superbase which is the product of two alternating subbases. This, as opposed to some two-digit superbase encoded in some even larger subbase, such as 12'34'56CA[G], which would be tess-encoded base zentess&ten, i.e., hexadecimally-encoded base twelve sixteens plus ten (where CAG = 14AC = 202A). Remember, those X-encoded Y bases are "ragged", meaning they don't use every value that X subbase can encode within the number of subdigits needed to express the value Y.)
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
Double sharp @ Dec 28 2017, 09:07 AM wrote:P.S. Glad to hear you can get used to "yon, weg, mand", even though only the last is from me - it's just that Kode originally had the first two as cadex and exent respectively, and it was my idea to make them mean thise and zeff.
Chalk up that dozenally-encoded trinilial thing, with zen and yon as tier-markers, as another one of my half-baked-but-fully-repented ideas. Who needs two-syllable, subbase-encoded digits? I'm all on board with yon=ainzeen=thirteen, weg=twenzeen=fourteen, mand=thirzeen=fifteen, etc.
Of course, this being his system, his extensions sound entirely natural, something I feel I only managed with "mand".
Mandel=thirzeen=fifteen was a nice find. I like it. (Still curious about the deep etymology. Did "almonds" have some association with "fifteen" in Teutonic culture?)
I still think the "x" annotation for hexadecimal half-seriously implies a syllable starting with "x", following my tongue-in-cheek principle that many bad etymologies are as good as one good one, but I can't think of a serious-sounding one.
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM

(Actually, I prefer to make that example 12'34'56C×A to emphasize that it's a superbase which is the product of two alternating subbases. This, as opposed to some two-digit superbase encoded in some even larger subbase, such as 12'34'56CA[G], which would be tess-encoded base zentess&ten, i.e., hexadecimally-encoded base twelve sixteens plus ten (where CAG = 14AC = 202A). Remember, those X-encoded Y bases are "ragged", meaning they don't use every value that X subbase can encode within the number of subdigits needed to express the value Y.)
I think that would be 12'34'56{g*c} instead, but otherwise, yes, that's exactly what I intended.

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
Double sharp @ Dec 28 2017, 11:46 AM wrote: I think that would be 12'34'56{g*c} instead, but otherwise, yes, that's exactly what I intended.
Oops that should have been doz·des instead of tes·doz. Fixed.
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
Kodegadulo @ Dec 28 2017, 11:33 AM wrote:
Double sharp @ Dec 28 2017, 09:07 AM wrote:P.S. Glad to hear you can get used to "yon, weg, mand", even though only the last is from me - it's just that Kode originally had the first two as cadex and exent respectively, and it was my idea to make them mean thise and zeff.
Chalk up that dozenally-encoded trinilial thing, with zen and yon as tier-markers, as another one of my half-baked-but-fully-repented ideas. Who needs two-syllable, subbase-encoded digits?  I'm all on board with yon=ainzeen=thirteen, weg=twenzeen=fourteen, mand=thirzeen=fifteen, etc.
Well, I think that by the time you get far enough past the human-scale, it becomes a necessity. You can't possibly have every digit root abbreviated by a different Latin letter past base twenty-six anyway, and running to different alphabets just delays the problem. Base sixteen certainly needs these extensions, and maybe eighteen and twenty do, but more than that and I think mixed-base SNN is the way to go, since bases past 20 are surely not human-scale anyway.
Kodegadulo @ Dec 28 2017, 11:33 AM wrote:Mandel=thirzeen=fifteen was a nice find. I like it. (Still curious about the deep etymology.  Did "almonds" have some association with "fifteen" in Teutonic culture?)
Oschkar @ Dec 18 2017, 09:07 PM wrote:German Wiktionary says:

von mittellateinisch: mandala „Bündel, Garbe“ im 15. Jahrhundert entlehnt; vermutlich mit lateinisch: manus „Hand“ verwandt

borrowed in the 15th century from Middle Latin mandala ‘bundle, sheaf’; possibly related to Latin manus ‘hand’
It seems likely that it came to be fifteen as a quarter of a shock (sixty), since there is a Bauernmandel of 16 corresponding to the Bauernschock of 64: here's a table of traditional German number names from German Wikipedia.
I know that that's the real reason, but since you used SNNw to mean systematic wegal (zeffal) nomenclature here, I thought it could go both ways. ^_^ Mind you, since most acrostics solve the problem of having few words starting with x in English by putting the x in the middle of the word instead, I wondered today if Icarus' "tex" (very reminiscent of "tess") might not be a serviceable digit root for sixteen. Sure, it doesn't start with x, but it ends with x.

Unfortunately this isn't quite enough to reach the next even base, because you'd need a root for seventeen. The Chinese names for their 24 cardinal directions mostly stick out like a sore thumb in the midst of all these Latin, Greek, and Germanic roots, but a few might possibly work out. Especially because it fits the "going backwards through the alphabet" pattern, I wonder if the last one (壬 rén) might not work as a digit root ren, multiplicative form rena; it is one of the few that won't get mangled by the difficulties of romanising Chinese too badly, and ends in a consonant to allow the multiplicative form. Yes, the vowel won't be right, but the Chinese retroflex rhotic matches the American English one pretty well. This is of course a suggestion of desperation and is liable to getting discarded the moment I find something better. ^_^

P.S. I just realised that this is in fact vaguely relevant to 17; the usual Chinese order of listing the compass points in speech is 東南西北 (ESWN), starting from east and going clockwise. Then if east is at 0, 壬 rén appears at 17 on the Chinese 24-point compass rose. So maybe this is vaguely relevant enough to work.

P.P.S. I just test-drove it in the spreadsheet, setting the "Adhoc" metrology to octodecimal, based on the pentdynciaday. I guess we really ought to have a proper name for this octodecimal metrology, but I have no idea what it ought to be. "Ren" works very well until you realise that base 17 would then be "renal", so evidently we need to double the n to keep the pronunciation with [ɛ] instead of [iː], which is at least closer to the Mandarin. (Mandarin /i/ has no allophones but itself, but the vowel /ə/ that  壬 rén correctly has moves freely around the vowel space, getting as far as [e], [o], and [ɤ]. In this case it is [ə], but these other realisations would likely be understood.) So I've changed it from "ren" to "renn".

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
On the original DGW Spreadsheet, I've split the former "Conventional" tab into separate tabs for "Time Units" and "Length Units". I've added a tab for "Velocity Units". So now, on the "Quantitel Powers" tab, you can select Quantity Selection "04:velocity" and then select Equiv Unit Selection from among the choices on the "Velocity Units" tab ... including, if you like, "07:furlong/fortnight".
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
Kodegadulo @ Dec 29 2017, 12:17 AM wrote: On the original DGW Spreadsheet, I've split the former "Conventional" tab into separate tabs for "Time Units" and "Length Units". I've added a tab for "Velocity Units".  So now, on the "Quantitel Powers" tab, you can select Quantity Selection "04:velocity" and then select Equiv Unit Selection from among the choices on the "Velocity Units" tab ... including, if you like, "07:furlong/fortnight". ;)
Well, that seems to be Oschkar's copy of the spreadsheet instead, but I can certainly merge the few changes I made to yours when I'm not on my phone. And yes, furlongs per fortnight are certainly amusing; can we soon expect the mass of a firkin of water from the FFF system when a tab for "Mass Units" appears? ^_-☆

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
Double sharp @ Dec 29 2017, 02:23 AM wrote:Well, that seems to be Oschkar's copy of the spreadsheet instead, but I can certainly merge the few changes I made to yours when I'm not on my phone.
Oops I pasted in the wrong link.  (Fixed it now.)  I did do the change in the original spreadsheet, not Oschkar's copy.  I believe it has all your latest changes also, DS, so no need to edit them, unless you want to go do it in Oschkar's version to keep it up to date. Although I think the name of the game for him would be to simplify and focus on just implementing the 2-subbase case.
And yes, furlongs per fortnight are certainly amusing; can we soon expect the mass of a firkin of water from the FFF system when a tab for "Mass Units" appears? ^_-☆
I think that would be de rigueur. :)
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

 Posts 775
Dozens Disciple
Oschkar
Dozens Disciple
Joined: Nov 19 2011, 01:07 AM

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
I did some cleanups in my version.

In particular I made sure to use two digits of dozenal, rather than the metrology radix, for the indexes in all the selectors. (No need for these to churn just because you select a different metrology with a different radix.) (Did I catch them all?)

Also noticed that the Multiplier and Reciprocal prefixes on the Digit Sequences tab were getting prefixed with spurious dots when dozenal base was selected. (Blank radix syllable, but the separating dot was still there.)

Introduced constants for the USC and BI prefixes. Shortened "Imperial" to "BI".

Normalized all "cubic X" to "X³"
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
I've added a few of my suggested digit roots to both spreadsheets, since no one so far has raised any serious problems; "cait" (from cater-corner) for quaternary (quadral) base, and "tex" (abbreviation x) and "renn" (abbreviation r) for digits g and h to bring us to octodecimal (I've set the Adhoc metrology in both to a pentdynciaday-based one to make it easy to test out). Just a little more to go before the idea I raised early this year of SVN (Systematic Vigesimal Nomenclature) finally works out!

P.S. Come to think of it, veint (abbreviation v) is an obvious answer to the problem of finding a digit root for 20. Now we just need 18 and 19 to close the gap.

Obsessive poster
Obsessive poster
Joined: Sep 10 2011, 11:27 PM
There's always "vig" short for vigesimal. But it's not so critical to have a digit root for twenty, unless we want to support an even larger base. We already have "scor" as a radix syllable.

As for "cait", "tex", and "renn", they're pretty good, and better than nothing, so let's see how they work out. As for what to do for digits sixzeen, sevenzeen (eighteen, nineteen), why not just keep Omnibase-derived "ind", "jul"? (Note that "ind" is a quasi-anagram of "dine" or "dyn".)

So with that, we have something which maintains unique initialisms for all digit roots up to (and just past) vigesimal:

n = nil
u = un
b = bi
t = tri
p = pent
h = hex
s = sept
o = oct
e = enn
d = dec
l = lev
z = zen
y = yon
w = weg
m = mand
x = tex (well, this one's a finalism)
r = renn
i = ind
j = jul
v = vig

Letters used: bdehijlmnopqrstuvwxyz
Letters unused: acfgk

These can the be used in SNN abbreviations. They can also be used as subscript radix annotations in what I've called the "Nominal" style, sans brackets, square or curly:

400h = 220o = 144d = 100z = A4w = 90x = 80i = 74v

Although square brackets are a fallback in typesetting-challenged environments:

400[h] = 220[o] = 144[d] = 100[z] = A4[w] = 90[x] = 80 = 74[v]

This does not conflict with the "Computerese" style, because that uses digits and uppercase as subscripts:

4006 = 2208 = 144A = 100C = A4E = 90G = 80I = 74K

This also falls back to square brackets, again without conflict:

400[6] = 220[8] = 144[A] = 100[C] = A4[E] = 90[G] = 80 = 74[K]

This is what the spreadsheet uses, because that's what the built in BASE function uses.

This also doesn't conflict with DS's annotation style, which uses curly braces and lower case (at first, at least):

400{6} = 220{8} = 144{a} = 100{c} = a4{e} = 90{g} = 80{i} = 74{k}

400{6} = 220{8} = 144{a} = 100{c} = a4{e} = 90{g} = 80{i} = 74{k}

("cait" was offered as a radix syllable rather than a digit)
As of 1202/03/01[z]=2018/03/01[d] I use:
ten,eleven = ↊↋, ᘔƐ, ӾƐ, XE or AB.
Base-neutral base annotations
Systematic Dozenal Nomenclature
Primel Metrology
Western encoding (not by choice)
Greasemonkey + Mathjax + PrimelDozenator
(Links to these and other useful topics are in my index post;
click on my user name and go to my "Website" link)

Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
I see I've overlooked the completely obvious; the impetus for coming up with coinages for g and h was that you can't re-use the letter h (hex), but this is obviously not a problem for i and j. So I see we've managed together to push this contraption up till unvigesimal while keeping some reasons for each coinage that aren't arbitrary mutations, which is really quite extraordinary fruit for our collaboration, you having suggested {d, e, i, j, k} and I having suggested {f, g, h}.

To illustrate this, I've changed the Adhoc metrology to a pentscorciaday-based one.

 Posts 2,343
Dozens Demigod
Double sharp
Dozens Demigod
Joined: Sep 19 2015, 11:02 AM
Incidentally, I'm wondering if we couldn't come up with a general series of DGW metrology annotations. I don't think anyone would want to come up with two metrologies that only differ in something like the value of Earth's gravity used: the units would be too close with the minute differences only being a vague annoyance. To that end, I think that we could simply call out the base (i.e. the timel) and let the rest follow. I'm thinking that we should first write down the base, and then the absolute value of the exponent.

So Primel would be the (c,6)-metrology and Tertiel would be the (c,5)-metrology, while Xing would be the (8,6)-metrology, Shirushi the (8,7)-metrology, Ash-Ashtrian the (6,8)-metrology, Ashtrian the (c×a,3)-metrology, Sauter the (a,6)-metrology, Xing the (g,5)-metrology, and Fleur the (e,5)-metrology. The two Adhoc metrologies I've previously set are the (i,5)- and (k,5)-metrologies respectively, and we can use these as brand-markers for metrologies we are just alluding to and not really using much. An extension to things like Phasic would probably be something like (c,5,4). Then we could talk about things like the (h,5)-timel and the (6×e,3)-lengthel and the (6×f×e,2,2)-massel, although it stands to reason that for metrologies we can conceive of people using (with a reasonable base and a perceptible timel), a real symbol is better.

I should also note that this system is just a suggestion, and that anyone is welcome to suggest improvements.

 Posts 775
Dozens Disciple
Oschkar
Dozens Disciple
Joined: Nov 19 2011, 01:07 AM
I’m going to propose the roots guin, arc and flor for the next three digits. There were twenty-one shillings in a guinea, the Tarot deck has twenty-two Major Arcana cards, and the argam digit for twenty-three is flore. Finally, the letter "c" for cadex can be used for as a tetravigesimal base annotation, where a digit root isn’t actually needed.