Zeta EDOs that subdivide other Zeta EDOs

It's quite handy that 12-EDO is a zeta EDO, and that 72-EDO is also a
zeta EDO. This is nice for several reasons:

- It means that 12-EDO serves really well as a coarse template for a
unit of finer interval measure like cents, since you only need to
subdivide everything into sixths to get a load of more great JI
intervals

- It means that if you're a native 12-EDO listener, 72-EDO will be a
very natural way for you to expand your horizons to understand more of
JI

- It means that if you start with 12-EDO and split the intervals into
common simple divisions, like splitting the minor third into half, you
tend to get other useful JI ratios like 12/11

I note that this relationship also exists with 19 and 152; 152 is 19 *
8. Likewise, 217 is 31 * 7.

This sort of thing must turn up a lot, because in general, if you take
an interval like 6/5 and split it in half, you get something close to
10:11:12 (or if you split it into thirds, you get 15:16:17:18, etc).
So, if you take tunings which approximate JI well and start
subdividing them, you're bound to, in general, hit other tunings which
also approximate JI well, sooner or later. I would expect that there
will often be "harmonics" in these zeta EDO lists, in that if some
number is in the list, then some low-numbered multiple of that number
is also likely to be in the list, and so on.

Does anyone have any more significant instances of this sort of thing
happening, with EDOs greater than 12?

Mike

On Tue, Aug 6, 2013 at 6:59 AM, Mike Battaglia <battaglia01@gmail.com> wrote:
>
> Does anyone have any more significant instances of this sort of thing
> happening, with EDOs greater than 12?

Ones I've found so far:

12*6 = 72
19*8 = 152
27*10 = 270
31*7 = 217
19*12 = 342
19*26 = 494
53*14 = 742
53*18 = 954
22*46 = 1012

There may be more; I did this coarsely by hand here. It's interesting
to me that 19 and 53 turn up so much, though, whereas 12 doesn't seem
to - the next time 12 turns up is 1236. 22 also does quite poorly.

Mike

One that I know by heart is 19*9 = 171.

I think the reason 19-EDO shows up that often is that it approximates one particular simple ratio (6:5) exceptionally well. (The non-Zeta EDO) 9 has a very good approximation for 7:6, so it's not very surprising to me that it shows up often in this list, too (27, 72, 99, 171, ...).

So 171 has an accurate 6:5 and 7:6, and since it coincidentally has an accurate 3:2 as well, it's accurate in the 7-limit.

12-EDO, on the other hand, isn't that accurate with 3:2 (~2 cents off), and I don't know of other simple ratios it approximates within a small fraction of a cent (which doesn't necessarily mean there aren't any).

-Gedankenwelt

--- In tuning-math@yahoogroups.com, Mike Battaglia <battaglia01@...> wrote:
>
> Ones I've found so far:
>
> 12*6 = 72
> 19*8 = 152
> 27*10 = 270
> 31*7 = 217
> 19*12 = 342
> 19*26 = 494
> 53*14 = 742
> 53*18 = 954
> 22*46 = 1012
>
> There may be more; I did this coarsely by hand here. It's interesting
> to me that 19 and 53 turn up so much, though, whereas 12 doesn't seem
> to - the next time 12 turns up is 1236. 22 also does quite poorly.

A single degree of 2460-EDO corresponds to the Sagittal "mina". 2460-EDO also divides 41-EDO into 60 parts.

--George

--- In tuning-math@yahoogroups.com, Mike Battaglia <battaglia01@...> wrote:
>
> It's quite handy that 12-EDO is a zeta EDO, and that 72-EDO is also a
> zeta EDO. This is nice for several reasons:
>
> - It means that 12-EDO serves really well as a coarse template for a
> unit of finer interval measure like cents, since you only need to
> subdivide everything into sixths to get a load of more great JI
> intervals
>
> - It means that if you're a native 12-EDO listener, 72-EDO will be a
> very natural way for you to expand your horizons to understand more of
> JI
>
> - It means that if you start with 12-EDO and split the intervals into
> common simple divisions, like splitting the minor third into half, you
> tend to get other useful JI ratios like 12/11
>
> I note that this relationship also exists with 19 and 152; 152 is 19 *
> 8. Likewise, 217 is 31 * 7.
>
> This sort of thing must turn up a lot, because in general, if you take
> an interval like 6/5 and split it in half, you get something close to
> 10:11:12 (or if you split it into thirds, you get 15:16:17:18, etc).
> So, if you take tunings which approximate JI well and start
> subdividing them, you're bound to, in general, hit other tunings which
> also approximate JI well, sooner or later. I would expect that there
> will often be "harmonics" in these zeta EDO lists, in that if some
> number is in the list, then some low-numbered multiple of that number
> is also likely to be in the list, and so on.
>
> Does anyone have any more significant instances of this sort of thing
> happening, with EDOs greater than 12?
>
> Mike
>

If we look into edts we get the great example of triple BP! 13edt, 26edt, 39edt and 52edt are remarkably all zeta edts. As Paul Ehrlich has pointed out, 39edt (triple BP) does particularly well at approximating just intervals, from 13edt's great 3.5.7 basis to an extremely accurate 3.5.7.9.11.13.15 system! I must get round to writing in triple BP...

Anyway, you also get
15*3=45 and 15*5=75
So of the zeta edts between 20 and 100, 5 out of 9 are multiples of zeta edts 15 or 13!

Since no edt represents the no 2s 7-limit better than 13edt until some much larger number, (forgot and can't find the answer on the web), and multiples of 13edt are also zeta edts, treating 13edt like how we've treated 12edo, dividing each step into 100cents would provide a really good unit of measurement for intervals in the world of edts, though 13 does not share the added bonus that 12 does of being highly composite.
You could also make division of each step into three steps of 39edt more natural with the use of 60 minutes to divide each step rather than 100c, each step of 39edt becoming 20mins. (This idea has similar advantages in 12edo, where each step of 72edo becomes 10 minutes)

Anyway yeah. I haven't found any other examples.

Gareth

--- In tuning-math@yahoogroups.com, Mike Battaglia <battaglia01@...> wrote:
>
> It's quite handy that 12-EDO is a zeta EDO, and that 72-EDO is also a
> zeta EDO. This is nice for several reasons:
>
> - It means that 12-EDO serves really well as a coarse template for a
> unit of finer interval measure like cents, since you only need to
> subdivide everything into sixths to get a load of more great JI
> intervals
>
> - It means that if you're a native 12-EDO listener, 72-EDO will be a
> very natural way for you to expand your horizons to understand more of
> JI
>
> - It means that if you start with 12-EDO and split the intervals into
> common simple divisions, like splitting the minor third into half, you
> tend to get other useful JI ratios like 12/11
>
> I note that this relationship also exists with 19 and 152; 152 is 19 *
> 8. Likewise, 217 is 31 * 7.
>
> This sort of thing must turn up a lot, because in general, if you take
> an interval like 6/5 and split it in half, you get something close to
> 10:11:12 (or if you split it into thirds, you get 15:16:17:18, etc).
> So, if you take tunings which approximate JI well and start
> subdividing them, you're bound to, in general, hit other tunings which
> also approximate JI well, sooner or later. I would expect that there
> will often be "harmonics" in these zeta EDO lists, in that if some
> number is in the list, then some low-numbered multiple of that number
> is also likely to be in the list, and so on.
>
> Does anyone have any more significant instances of this sort of thing
> happening, with EDOs greater than 12?
>
> Mike
>

Since I don't know your actual name, and "Gedankenwelt" is a pain to
type, I'm going to start calling you Geddy, since it's something like
an actual name. This also leaves open the possibility that you might
be the lead singer for Rush, which would be sweet.

> One that I know by heart is 19*9 = 171.

Yes, dunno how I missed that one! Good.

> I think the reason 19-EDO shows up that often is that it approximates one particular simple ratio (6:5) exceptionally well. (The non-Zeta EDO) 9 has a very good approximation for 7:6, so it's not very surprising to me that it shows up often in this list, too (27, 72, 99, 171, ...).

That may be the case. Also, 11-EDO has a great 9/7, which is nice.

As far as I can tell, the zeta function really just focuses on primes.
But, I had this conjecture at one point that if you come up with an
altered zeta function where the product formula multiplies together
all positive rationals rather than all primes, that you'd end up with
something of the form f(zeta(s)), where f is a nonlinear monotonic
function. You can also have it add together the reciprocals of all
positive rationals rather than just all integers and do the same
conjecture. This means the peaks would be preserved, so you'd get the
same results.

(which doesn't necessarily mean there aren't any).

12-EDO does some 17-limit and 19-limit stuff very well.

Mike

Oh yes, this is great! I had no idea. So 2460/41 is a superior highly
composite number (60), which was going to be the subject of my next
post.

Mike

On Wed, Aug 7, 2013 at 3:42 PM, gdsecor <gdsecor@yahoo.com> wrote:
>
> A single degree of 2460-EDO corresponds to the Sagittal "mina". 2460-EDO also divides 41-EDO into 60 parts.
>
> --George

Gareth, sorry your post took so long to get through - I didn't have
tuning-math's preferences to alert me to when a user had posted for
the first time, but now I do. Users are moderated by default here, so
we can avoid spammers; after the first post the moderators manually
unmoderate you. If this ever happens to anyone else, please feel free
to message me offlist and I'll make sure you get approved.

On Sun, Aug 11, 2013 at 11:44 AM, hearneg23 <gareth.hearne@gmail.com> wrote:
>
> If we look into edts we get the great example of triple BP! 13edt, 26edt, 39edt and 52edt are remarkably all zeta edts. As Paul Ehrlich has pointed out, 39edt (triple BP) does particularly well at approximating just intervals, from 13edt's great 3.5.7 basis to an extremely accurate 3.5.7.9.11.13.15 system! I must get round to writing in triple BP...

This is a really great thing to notice. Are you using the modified
no-2's zeta function here? (It's spelled Erlich, btw.)

> Anyway, you also get
> 15*3=45 and 15*5=75
> So of the zeta edts between 20 and 100, 5 out of 9 are multiples of zeta edts 15 or 13!

Very interesting!

> Since no edt represents the no 2s 7-limit better than 13edt until some much larger number, (forgot and can't find the answer on the web), and multiples of 13edt are also zeta edts, treating 13edt like how we've treated 12edo, dividing each step into 100cents would provide a really good unit of measurement for intervals in the world of edts, though 13 does not share the added bonus that 12 does of being highly composite.

Yes, the fact that 12 *itself* is highly composite, or at least has
lots of factors, is definitely something I find musically useful,
though that's a completely different criterion from that of harmony,
unfortunately, and the two don't merge often.

> You could also make division of each step into three steps of 39edt more natural with the use of 60 minutes to divide each step rather than 100c, each step of 39edt becoming 20mins. (This idea has similar advantages in 12edo, where each step of 72edo becomes 10 minutes)

Yeah, that's where I was going with the next post, as I mentioned
offlist... still to come.

Mike

--- In tuning-math@yahoogroups.com, Mike Battaglia <battaglia01@...> wrote:
>
> Since I don't know your actual name, and "Gedankenwelt" is a pain to
> type, I'm going to start calling you Geddy, since it's something like
> an actual name. This also leaves open the possibility that you might
> be the lead singer for Rush, which would be sweet.

Well, it's certainly true that "Gedankenwelt" is a little long, and may be difficult to type / spell for non-German speakers, so I think it's ok.

But funny that you mention Geddy; if my mother wouldn't have been against it, my father would have given me the forename "Jaco", inspired from another famous bassist / multi-instrumentalist. :)
(which pretty much rules out that latter conjecture of yours, though)

> Also, 11-EDO has a great 9/7, which is nice.
[...]
> 12-EDO does some 17-limit and 19-limit stuff very well.

They're good, but by far not as accurate as the approximations I mentioned, or they involve intervals that are even more complex than 18:17 or 19:16.

2\9 is only 0.204 cents smaller than 7:6. The next EDO that does slightly better is 328-EDO with 73\328.

5\19 is even better with approximating the simpler ratio 6:5 only 0.148 cents sharp, though 19-EDO is about twice as large as 9-EDO, and doesn't have such a simple prime factorization.

Best
- Gedankenwelt (or Geddy, or whatever :)

On Fri, Aug 16, 2013 at 11:56 PM, gedankenwelt94
<gedankenwelt94@yahoo.com> wrote:
>
> Well, it's certainly true that "Gedankenwelt" is a little long, and may be difficult to type / spell for non-German speakers, so I think it's ok.
>
> But funny that you mention Geddy; if my mother wouldn't have been against it, my father would have given me the forename "Jaco", inspired from another famous bassist / multi-instrumentalist. :)
> (which pretty much rules out that latter conjecture of yours, though)

Haha, cool. Jaco used to teach at the University of Miami where I went
to school, I had teachers who knew him... they said that he was a
great bassist and composer but kind of out there as a teacher.

> They're good, but by far not as accurate as the approximations I mentioned, or they involve intervals that are even more complex than 18:17 or 19:16.
>
> 2\9 is only 0.204 cents smaller than 7:6. The next EDO that does slightly better is 328-EDO with 73\328.
>
> 5\19 is even better with approximating the simpler ratio 6:5 only 0.148 cents sharp, though 19-EDO is about twice as large as 9-EDO, and doesn't have such a simple prime factorization.

If you look at Euler's product formula, the function is really
focusing mostly on primes, but maybe there's some connection with
simple ratios in general...

Mike

Let's see ...

If we have primes p_k, we can define "secondary intervals" as prime ratios p_i /
p_j, or as

products of two primes p_i * p_j.

A Zeta EDO approximates p_k as log_2(p_k) + eps_k for all k, where eps_k is the
error in

octaves.

That means the error for p_i / p_j is eps_i - eps_j, and the error for p_i * p_j
is eps_i + eps_j.So if min_err is the error with the smaller absolute value, and
max_err the one with the largerabsolute value, then |min_err| = ||eps_i| -
|eps_j||, and |max_err| = |eps_i| + |eps_j|.
One consequence is that if we consider a certain number of primes, and p_i and
p_j are thebest approximated ones, then p_i / p_j and p_i * p_j will do fairly
well:The better one will outperform the second best approximated prime, and the
worse one is stillpretty good.
Another case:If p_i and p_j are the primes for which |eps_i - eps_j| is minimal,
then there will be asecondary interval with an error min_err, where |min_err| =
|eps_i - eps_j|. So if we look atenough primes, it's probable that there's a
secondary interval with a really small error.

Now it would be interesting to know how the primes and secondary intervals
spread over thesub-EDOs that divide the Zeta EDO (but are smaller than the Zeta
EDO). If we look at theZeta EDO representations of p_i, p_j, p_i / p_j and p_i *
p_j, and if there are subdivisions ofthe Zeta EDO, is it guaranteed that at
least one of those intervals occurs in an (actual)sub-EDO?
If so, and if we consider "simple intervals" as either primes or secondary
intervals (for agiven range of primes that we look at), then there have to be
sub-EDOs that approximatesome of these simple intervals "fairly well".(in the
worst case with an error whose absolute value is |eps_i| + |eps_j|, which is
still goodif p_i and p_j are the best approximated primes)

-Geddy
P.S.: What the heck happened to the Yahoo! Groups interface?There's not even a
preview button, and I don't see a tree view once I log in.

--- In tuning-math@yahoogroups.com, <battaglia01@...> wrote:

If you look at Euler's product formula, the function is really
focusing mostly on primes, but maybe there's some connection with
simple ratios in general...

gedankenwelt94 wrote:

< P.S.: What the heck happened to the Yahoo! Groups interface?< There's not even
a preview button, and I don't see a tree view once I log in.This interface
stinks! I wanted to reply to a specific part of your message, but in order to do
that I had to copy & paste it in above.I would also like to read through all the
new messages on the web in chronological order, regardless of subject, but I
don't see any easy way to do that. As a workaround I decided to edit my
membership (which I finally found in a drop-down box at the top right of the
screen) in each group to receive a daily digest.--George

--- In tuning-math@yahoogroups.com, <gedankenwelt94@...> wrote:

Let's see ...

If we have primes p_k, we can define "secondary intervals" as prime ratios p_i /
p_j, or as

products of two primes p_i * p_j.

A Zeta EDO approximates p_k as log_2(p_k) + eps_k for all k, where eps_k is the
error in

octaves.

That means the error for p_i / p_j is eps_i - eps_j, and the error for p_i * p_j
is eps_i + eps_j.So if min_err is the error with the smaller absolute value, and
max_err the one with the largerabsolute value, then |min_err| = ||eps_i| -
|eps_j||, and |max_err| = |eps_i| + |eps_j|.
One consequence is that if we consider a certain number of primes, and p_i and
p_j are thebest approximated ones, then p_i / p_j and p_i * p_j will do fairly
well:The better one will outperform the second best approximated prime, and the
worse one is stillpretty good.
Another case:If p_i and p_j are the primes for which |eps_i - eps_j| is minimal,
then there will be asecondary interval with an error min_err, where |min_err| =
|eps_i - eps_j|. So if we look atenough primes, it's probable that there's a
secondary interval with a really small error.

Now it would be interesting to know how the primes and secondary intervals
spread over thesub-EDOs that divide the Zeta EDO (but are smaller than the Zeta
EDO). If we look at theZeta EDO representations of p_i, p_j, p_i / p_j and p_i *
p_j, and if there are subdivisions ofthe Zeta EDO, is it guaranteed that at
least one of those intervals occurs in an (actual)sub-EDO?
If so, and if we consider "simple intervals" as either primes or secondary
intervals (for agiven range of primes that we look at), then there have to be
sub-EDOs that approximatesome of these simple intervals "fairly well".(in the
worst case with an error whose absolute value is |eps_i| + |eps_j|, which is
still goodif p_i and p_j are the best approximated primes)

-Geddy
P.S.: What the heck happened to the Yahoo! Groups interface?There's not even a
preview button, and I don't see a tree view once I log in.

--- In tuning-math@yahoogroups.com, <battaglia01@...> wrote:

If you look at Euler's product formula, the function is really
focusing mostly on primes, but maybe there's some connection with
simple ratios in general...

Sorry, missed this...

On Thu, Aug 29, 2013 at 10:27 AM, <gedankenwelt94@yahoo.com> wrote:
>
> One consequence is that if we consider a certain number of primes, and p_i and p_j are the
> best approximated ones, then p_i / p_j and p_i * p_j will do fairly well:
> The better one will outperform the second best approximated prime, and the worse one is still
> pretty good.
>
> Another case:
> If p_i and p_j are the primes for which |eps_i - eps_j| is minimal, then there will be a
> secondary interval with an error min_err, where |min_err| = |eps_i - eps_j|. So if we look at
> enough primes, it's probable that there's a secondary interval with a really small error.

Also, even if p_i and p_j have every high error, if |eps_i| ~=
|eps_j|, then either p_i/p_j or p_i*p_j will be good.

> Now it would be interesting to know how the primes and secondary intervals spread over the
> sub-EDOs that divide the Zeta EDO (but are smaller than the Zeta EDO). If we look at the
> Zeta EDO representations of p_i, p_j, p_i / p_j and p_i * p_j, and if there are subdivisions of
> the Zeta EDO, is it guaranteed that at least one of those intervals occurs in an (actual)
> sub-EDO?

So let's say that N is our zeta EDO. If we allow N to be prime, then
the answer to your question is "no," but I'll assume you only meant
composite N. Then N has a set of positive divisors d(N). An interval
i, expressed as a number of steps in N, exists in a sub-EDO of N iff
it's a multiple of something in d(N).

If V is the patent val for N, then your question can be rephrased as
follows: given two primes p_i and p_j, is one of the set {V(p_i),
V(p_j), V(p_i) + V(p_j), V(p_i) - V(p_j)} a multiple of something in
d(N)?

If N is even, it's clear to see that the answer is "yes." If V(p_i)
and V(p_j) are both odd, then their sum is even, and if not, then one
of them is even.

If N is divisible by 3, then the answer is still "yes." If V(p_i) mod
3 = V(p_j) mod 3, then V(p_i) - V(p_j) = 0 mod 3. If either V(p_n) = 0
mod 3, then the problem is already solved. That leaves the case where
V(p_i) = 1 mod 3 and V(p_j) = 2 mod 3 (or vice versa), but this
requires V(p_i) + V(p_j) = 0 mod 3.

If N is divisible by 5, the answer can finally be "no," because you
have the cases like where V(p_i) = 1 mod 5 and V(p_j) = 2 mod 5. If
V(p_i) and V(p_j) both aren't 0 mod 5, and if they both aren't equal,
then the condition where they add up to something other than 0 mod 5
means their difference is also something other than 0 mod 5.

The first composite zeta EDO not divisible by 2 or 3 that I can find
is 217, which is 31*7. The 11-limit patent val for 217 is
http://x31eq.com/cgi-bin/rt.cgi?ets=217&limit=11. This maps 3/1 to 344
and 11/1 to 751: neither of these are divisible by 31 or 7, nor is
their sum. So that's a counterexample to your conjecture.

> P.S.: What the heck happened to the Yahoo! Groups interface?
> There's not even a preview button, and I don't see a tree view once I log in.

You can find messages here -
http://groups.yahoo.com/neo/groups/tuning-math/conversations/topics

Looks like they gave us 2 GB of file space now! Phew.

Mike

On Thu, Aug 29, 2013 at 12:43 PM, <gdsecor@yahoo.com> wrote:
>
> This interface stinks! I wanted to reply to a specific part of your message, but in order to do that I had to copy & paste it in above.
>
> I would also like to read through all the new messages on the web in chronological order, regardless of subject, but I don't see any easy way to do that. As a workaround I decided to edit my membership (which I finally found in a drop-down box at the top right of the screen) in each group to receive a daily digest.

You can do that by clicking on the "conversations" link under the
tuning-math heading. The link is here:
http://groups.yahoo.com/neo/groups/tuning-math/conversations/topics

Mike

Hey :)

Mike wrote:

> Sorry, missed this...

No problem! I was able to answer the question myself, I only didn't find

the time to post about it here. ;)

> If we allow N to be prime, then
> the answer to your question is "no," but I'll assume you only meant
composite N.

Yeah, I assumed composite N - that's what I meant when I wrote

"and if there are subdivisions of the Zeta EDO". Sorry if this wasn't

clear!

The conjecture may be wrong, but since it applies to EDOs divisible by 2 or 3,

it applies to almost exactly 4 out of 6 EDOs in a given range of subsequent
EDOs.

I noticed that also ~2/3 of the first 51 Zeta-EDOs is divisible by 2 or 3,

and only 6 out of those 51 Zeta-EDOs are composite, but notdivisible by 2 or 3.

Those are:

217 = 7 * 31
935 = 5 * 11 * 17
3395 = 5 * 7 * 97
5585 = 5 * 1117
7033 = 13 * 541
34691 = 113 * 307

I think in these cases it's a question of probability whether for two given
primes

p_i, p_j, at least one interval in {V(p_i), V(p_j), V(p_i) + V(p_j), V(p_i) -
V(p_j)}

occurs in a sub-EDO. If N-EDO is our Zeta-EDO, then the probability should be

higher if N is composed of many different / small prime factors.

So the probability for 34691-EDO and a given pair of primes p_i, p_j should

be rather small if we ignore the fact that it is a Zeta-EDOs, which may

increase the probability due to some relations we're not aware of yet.

In 217-EDO, on the other side, 3 and 11 is the only pair of primes in the
13-limit

for which my previous conjecture doesn't hold.

The probability for 3395-EDO and agiven pair of primes should be rather high,

since it's divisible by both 5 and 7.

I wonder if my approach is useful for finding an answer to your original
question.

- Geddy

Mike wrote:

> You can do that by clicking on the "conversations" link under the
> tuning-math heading. The link is here:
> http://groups.yahoo.com/neo/groups/tuning-math/conversations/topics

He asked for new messages in chronological order, regardless of subject,

so here's a better link: ;)

http://groups.yahoo.com/neo/groups/tuning-math/conversations/messages

(you can get there if you click on "messages" after clicking on "conversations")

- Geddy

gedeankenwelt94 wrote:

> He asked for new messages in chronological order, regardless of subject,

> so here's a better link: ;)

>

> http://groups.yahoo.com/neo/groups/tuning-math/conversations/messages

>

> (you can get there if you click on "messages" after clicking on
"conversations")

>

> - Geddy

Yes, I already tried that, and, as you pointed out, it *lists* the new messages
in chronological order, newest first. However, I said that I would like to
*read* the messages in chronological order, but when I click on a particular
message, there is no option to go to the *next* message. One work-around is to
go through the listing, opening each message in a new tab. However, I'm still
faced with the problem of determining which message is the oldest one I haven't
yet read, since there is no color differentiation between the recently read and
unread messages.

Someone on another Yahoo forum pointed out that not everyone is seeing the new
interface at this time, only those members who are being routed to web addresses
containing "/neo/", and if you have the new interface in one Yahoo group, you'll
have it in all of them. (For example, I'm now working in
http://groups.yahoo.com/neo/groups/tuning-math/conversations/topics/
[http://groups.yahoo.com/neo/groups/tuning-math/conversations/topics/]... Yahoo
is slowly switching everyone over to the new interface so it's a matter of time.
I happen to have another Yahoo ID, and when I log into that one I still see
everything in the old interface.

--George

--- In tuning-math@yahoogroups.com, <gedankenwelt94@...> wrote:

Mike wrote:

> You can do that by clicking on the "conversations" link under the
> tuning-math heading. The link is here:
> http://groups.yahoo.com/neo/groups/tuning-math/conversations/topics

He asked for new messages in chronological order, regardless of subject,

so here's a better link: ;)

http://groups.yahoo.com/neo/groups/tuning-math/conversations/messages

(you can get there if you click on "messages" after clicking on "conversations")

- Geddy

George Secor wrote:

> Yes, I already tried that, and, as you pointed out, it *lists* the new
messages in chronological order, newest first. However, I said that I would like
to *read* the messages in chronological order, but when I click on a particular
message, there is no option to go to the *next* message. One work-around is to
go through the listing, opening each message in a new tab. However, I'm still
faced with the problem of determining which message is the oldest one I haven't
yet read, since there is no color differentiation between the recently read and
unread messages.

Sorry, looks like I misunderstood you there.

If you open a message, at the top right there are buttons "<" and ">" to go

to the previous and next message, respectively. So you can open the oldest

message you didn't read, and then click on ">" (next).

Alternatively, you can click on the "<-" (back) button on the left to return to

the message listing, and then select the next message to read, so opening

new tabs isn't necessary.

If you have different colors for read and unread messages, then it should be

because your internet browser memorizes the links you visited, and uses

different colors for new (unvisited) links and links you already visited.

If the color coding doesn't work in your case, then it probably either means

that Yahoo changed the links since you read the messages, or that you

logged in using different Yahoo IDs, and messages you already read using

the old interface appear as "unread" when using the new interface (or
vice-versa),

because the links are different.

I hope I could help!

- Geddy

On Sun, Sep 8, 2013 at 7:31 AM, <gedankenwelt94@yahoo.com> wrote:
>
> Sorry, looks like I misunderstood you there.
>
> If you open a message, at the top right there are buttons "<" and ">" to go
> to the previous and next message, respectively. So you can open the oldest
> message you didn't read, and then click on ">" (next).

Yeah, Geddy's right. People hate this new interface, BTW, and are
revolting en masse, so I can only hope that Yahoo! makes some fixes to
it soon.

Mike

--- In tuning-math@yahoogroups.com, <tuning-math@yahoogroups.com> wrote:
(Unfortunately, the above doesn't identify the poster as "Geddy", as the old
interface would have.
> If you open a message, at the top right there are buttons "<" and ">" to go

to the previous and next message, respectively. So you can open the oldest

message you didn't read, and then click on ">" (next).

Okay, that's good! I didn't see those, so that solves one problem.

> ...

> If you have different colors for read and unread messages, then it should be

because your internet browser memorizes the links you visited, and uses

different colors for new (unvisited) links and links you already visited.

Yes, that's correct, however ...

> If the color coding doesn't work in your case, then it probably either means

that Yahoo changed the links since you read the messages, or that you

logged in using different Yahoo IDs, and messages you already read using

the old interface appear as "unread" when using the new interface (or
vice-versa),

because the links are different.

I've read well over a dozen messages in the new interface over the past week or
so, and none of them are appearing as already read.

> I hope I could help!

Thanks!

--George

On Sun, Sep 8, 2013 at 9:19 PM, <gdsecor@yahoo.com> wrote:
>
> Yes, that's correct, however ...
> I've read well over a dozen messages in the new interface over the past week or so, and none of them are appearing as already read.

This looks like a bug in Yahoo's CSS. They have two CSS selectors
which have the same specificity, such that the thing that's supposed
to make "visited" links change color is overridden by another thing
that makes links always blue.

I've reported the bug here:
http://answers.yahoo.com/question/index?qid=20130909124628AADGEeV

What browser are you running? I could code up a Greasemonkey script to
fix this in like 5 seconds.

Mike

On Mon, Sep 9, 2013 at 3:46 PM, Mike Battaglia <battaglia01@gmail.com> wrote:
>
> I've reported the bug here:
> http://answers.yahoo.com/question/index?qid=20130909124628AADGEeV
>
> What browser are you running? I could code up a Greasemonkey script to
> fix this in like 5 seconds.

Or, you could just type this in the URL field of your browser:

javascript:document.styleSheets[0].addRule(".yg-msglist-title
a:visited","color:#8384c8!important")

That'll fix you up and suddenly reveal which links are visited. You
can make that a bookmark in your toolbar to click on when you first
open Yahoo! Groups.

Mike