User talk:Linas/Archive17

Archives
(Jan-May 2005) (May-June 2005) (June-Aug 2005) (Sept-Oct 2005) (Nov-Dec 2005) (Dec 2005-Jan 2006) (Jan-Mar 2006) (Mar-May 2006) (May-July 2006) (July-Aug 2006) (Aug-Oct 2006) (Oct 2006-Jan 2007) (Feb-May 2007) (May-Dec 2007) (Jan-June 2008) (July-Dec 2008)

"Was this reviewed?"

On Wikipedia:Village pump (proposals) you wrote:

... much of the burden of revieweing edits could be improved with better tools. For example, I would love to know if one of my trusted collegues has already reviewed the same edit I'm reviewing. This would greatly reduce my review burden, and allow me to monitor many, many, many more articles. linas 23:35, 30 August 2005 (UTC)

Fantastic idea. Do you know whether there is some ongoing discussion on such things? (Feel free to reply here; I'm watching this page.) — Nowhither 18:36, 31 August 2005 (UTC)

I suspect there is, but I know not where. I have noticed that the wikimedia software made an attempt at implementing something like this, but it was either a hack or mis-designed or incomplete. You can see this on newer wikimedia sites, for example [1]. If you look at edit histories, you'll see red exclamation marks denoting unreviewed pages. But you'll also notice that any sockpuppet can reset them, ... so it really doesn't work correctly. So it seems someone thought about it, but I don't know what the status is, or where its going, or who is doing it. You'll have to look up the wikimedia folks.

Anyway, what I really want is actually fancier than what I wrote at the village pump, but I thought I'd keep it simple. I'd happily engage in a conversation with the wikimedia developers if you can locate them. linas 04:01, 1 September 2005 (UTC)

To clarify: This site runs the latest version of the wikimedia software, but the review system is turned off because it hurts performance. -- Jitse Niesen (talk) 11:23, 1 September 2005 (UTC)

Hmm, yes, it could be written as a fancy SQL query, and that would make the lights dim. Is this MySQL or Postgres? I'm guessing there are ways to make this more efficient, by using status bits of various kinds, requiring table redesigns. No matter, I didn't like the way the red exclamation marks worked anyway; they weren't really useful. linas 14:20, 1 September 2005 (UTC)

The WikiMedia sites are using MySQL. I was wrong by the way: the feature that you described is called "RC patrol", it's described on m:Help:Patrolled edit, and it seems that it was turned off because anybody could mark an edit as patrolled (as you already noticed, see also this mail and replies). I was confusing it with the m:Article validation feature, which is a more elaborate scheme that is disabled for performance reasons. -- Jitse Niesen (talk) 15:05, 1 September 2005 (UTC)

Hmm, thanks for the links, I'll have to prowl around there a bit. My other bit of patrol paranoia is that it is easy to review only the most recent change; thus a "bad edit" could be hidden in the history and overlooked. Thus, I'd prefer to see *all* changes since I last looked. linas 04:01, 2 September 2005 (UTC)

Santara-Šviesa

Dear Linas, would you consider writing this article? It has popped up a few times, most recently while researching the Vytautas Kavolis article, and probably will do so again. BTW, I think I met your mom (six degrees of Santara-Šviesa), she's a friend of my sister's, and am hoping to see her famous garden sometime. Sincerely, Novickas 12:55, 15 February 2007 (UTC)

I've wondered whether I should. I'm glad you asked. Despite my Santara-Šviesa number of 1, it may take me a while to gather up material. linas 13:43, 16 February 2007 (UTC)

That would be great. Hope you have some photos too. Thanks, Novickas 14:19, 16 February 2007 (UTC)

I'm still planning on doing this, but it may take a while. linas 04:13, 19 March 2007 (UTC)

Maybe I should ask my mom to do it. Also missing: Metmenys, although I see that Lituanus, Draugas are no longer red links. And going to Lituanus, and clicking on "what links here" seems to call up a fairly impressive list of other pages.linas (talk) 05:24, 23 January 2009 (UTC)

WikiProject History of Science newsletter : Issue V - January 2009

It's here at long last! The January 2009 issue of the WikiProject History of Science newsletter is ready, with exciting news about Darwin Day 2009. Please feel free to make corrections or add news about any project-related content you've been working on. You're receiving this because you are a participant in the History of Science WikiProject. You may read the newsletter or unsubscribe from this notification by following the link. Yours in discourse --ragesoss (talk) 03:02, 11 January 2009 (UTC)

SCM/AccuRev

Hi Linas, I was wondering if you would update your http://linas.org/linux/cmvc.html web page that includes references to commercial version control products, to include AccuRev SCM? It support LINUX. Thank you in advance and my apologies if this is the wrong place to inquire on your wiki. User:Vcwizard on 21 March 2007

Yes, I will try to get around to this soon (and also the wrike request below) linas 04:00, 24 July 2007 (UTC)

Linas,

I manage corporate communications for AccuRev. When you are able to add AccuRev to your listing of commercial version control tools listing, would you kindly use: AccuRev Version Control and hyperlink to this page: http://www.accurev.com/accurev-version-control.html

Thank you, Alex —Preceding unsigned comment added by 12.197.13.46 (talk) 22:33, 22 January 2009 (UTC)

Done. Sorry, I created that web page in 1996-1998, and have made almost no changes since then. I've added AccuRev, but have not attempted to otherwise revive that page. linas (talk) 05:10, 23 January 2009 (UTC)

GCC on S/370

Hi Linas, Paul Edwards here, maintainer of GCCMVS (now a sourceforge project). I was trying to document the history of GCC on S/370 (I picked up the code from Dave Pitts) and thought I'd make contact with you. It all works now as a native application (from MVS 3.8 to z/OS), generating HLASM and using IBM's assembler, but there are some bugs which no-one has the expertise to fix, so will thus remain forever. Please contact me at mutazilah at gmail.com if you'd like to help or comment. Note - this is the i370 port, not s390, since I needed HLASM. —Preceding unsigned comment added by 124.171.34.90 (talk) 21:30, 7 February 2009 (UTC)

Responding by email. I fixed a few bugs in the i370 backend to gcc; one had something to do with comparison of signed (or unsigned?) shorts; another was some inability to make register assignments when using some of the string copy instructions. I submitted these patches to GCC, but don't know if they ever were accepted or not. (I think the first ones were, but the later ones were not.) I'm sure I still have the patches lying about on a website somewhere, if you're interested in trying them out.

As to the "history", you'll have to ask pointed questions. linas (talk) 22:00, 9 February 2009 (UTC)

Reference for Eisenstein series recurrence relation please?

Hi, Do you have a reference for the Recurrence relation section of Eisenstein series? (You added that material back in 2005 if I'm reading the editing history correctly.) I'm looking for a similar recurrence relation for modular forms on the congruence subgroup $\Gamma _{0}(2)$. Must be standard but I'm just starting to learn about modular forms and haven't found that recurrence relation in any of my sources yet. Thanks! Typometer (talk) 14:53, 22 February 2009 (UTC)

Sure, its Tom M. Apostol, Modular Functions and Dirichlet Series in Number Theory, Second Edition (1990), Springer, New York ISBN 0-387-97127-0 I guess its basic, since its chapter 1, page 13 linas (talk) 14:21, 12 March 2009 (UTC)

Thank you very much! Typometer (talk) 23:10, 20 March 2009 (UTC)

Category:National electric power policy

Hi, Linas. I proposed to rename Category:National electric power policy to Category:Electric power policy. You could comment it here. Beagel (talk) 18:50, 21 March 2009 (UTC)

On transactional interpretation

Hello Linas. I saw you had edited also transactional interpretation article, so could you comment on the apparent paradox following from Cramer's proposal. Talk:Transactional_interpretation#Problem_with_the_transactional_model. In the simple beam splitter experiment TI leads to non-sense, could you comment on it. Regards, Danko Georgiev MD (talk) 12:32, 25 March 2009 (UTC)

Danko, Wikipedia is not the right place to perform original research, and is not really a very good place to discuss it either. You should find some mailing lists where you can discuss your ideas; you might try publishing your ideas in various forums. Wikipedia is meant to review established concepts and ideas. linas (talk) 16:39, 25 March 2009 (UTC)

Lidstone series

Can you address the questions I raised at Talk:Lidstone series? Michael Hardy (talk) 02:17, 3 April 2009 (UTC)

Arbatsky's principle

Hi, Linas, As regards your explanation of the "certainty" principle.

• There's a typo. "LatTer", not "later".
• It's not the best style to give references to articles not directly devoted to the topic and not to give references to the original sources. Possibly, a link to Arbatsky's later review would be appropriate there.
• It remains unclear, whether you agree that Arbatsky proved some UNcertainty relations for coordinate and momentum and for angle and angular mometum, that were not known before (AFAIK, for angle and angular momentum there was no simple relation at all).

Hunshi (talk) —Preceding undated comment added 14:22, 14 April 2009 (UTC).

Hi, Thanks for the note, I made the corrections. In the heat of the moment, I must have forgotten the references, I added the one.

Is the relation "unknown"? The derivation that I present is not difficult or deep; that is, it takes very few steps to obtain the result. Because of this, it seems possible that others took similar steps to obtain a similar result. Normally, when a derivation takes only a few steps to obtain, it is not considered to be "unknown", even if it has not been published in a journal. So, for example, the solutions to homework exercises in a book are "unknown", but are not publication-worthy precisely because they are not hard to obtain. The derivation that I give is about the same length/difficulty as some "average" homework problem, so it doesn't seem all that new or original.

On the other hand, the relation is interesting, and thinking about it does help shed light on the nature of tangent motion in projective spaces. It is possible that Arbatsky worked quite hard to obtain this result; it is possible that he has much more to say, or to clarify, than what I've said. I cannot say whether his papers are suitable for publication in some journal .. maybe they are. My goal was to provide a simple and direct derivation of his primary result, cutting through some of the fog and confusion of the various arguments.

Certainly, much more could be said about the relation. Questions pop into mind: how does it change on finite-dimensional projective manifolds? What about Hamiltonian manifolds (i.e. things that use the Poisson bracket instead of the commutator?) What about motion on Lie groups? How about quantum Lie groups? Integrable manifolds in general? What about plain-old real-valued manifolds? Compact manifolds? When are the operators Hermitian, and when not? Can anything be said about the C*-algebras? What about bases? And so on. I think it would be a worth-while (and very publishable) exercise to explore the above areas in detail. linas (talk) 20:38, 21 April 2009 (UTC)

These questions are all interesting. But let us discuss them later, because I do not understand some more simple things here.

Is your formula for the perpendicular velocity correct? Check, how WP defines the dot product. In complex case the order of arguments is important.

Also, in the definition of the angle, do you mean that arccos is a function of the complex argument? Is the angle theta complex? Hunshi (talk) 18:12, 22 April 2009 (UTC)

I attempted to present the derivation so that it would work in general for both real and complex generators, for a variety of different manifolds and tangent spaces, and not just for quantum mechanics. For the case of Hilbert spaces, where the by convention the generators are taken as Hermitian, there is an extra factor of i floating around in the definition of exp. I think that my argument also works for symplectic manifolds, in which case there would be an extra factor of $\Omega$ floating around. For Hermitian manifolds, this would be an extra factor of J. No, I have not double-checked all of these different special cases. Yes, the dot-product is understood to be a contraction with i, or J, etc. as appropriate for the given manifold. Yes, the order of index contraction is important, as is whether a given quantity is contra- or covariant. No, the angle theta is never complex.

If this generalized notation makes you uncomfortable, then I suggest that you should go through the detailed calculations for each of the different types of manifold. If you don't think you are able to do this, then I suggest that you should spend some time studying symplectic systems (I assume you already know quantum mechanics). There is a reason why quantum mechanics resembles classical mechanics, but is different in subtle ways, and by studying symplectic systems, you will better be able to understand how they are the same, and how they differ. Also, the notion of generators, tangent spaces, and manifolds are *far more general* than what is merely in quantum mechanics, and it is good to understand how this is. Indeed, I think Arbatsky's argument is far more general than just quantum mechanics -- the perpendicular velocity has the flavor of some aspects of sub-Riemannian geometry, since it splits the motion into parallel and perpendicular components, and then throws away the parallel part. Now, this is natural for QM, since the natural setting for QM is a projective space. So perhaps that setting would be the best setting for the argument. Just an idea.

Anyway, my argument is just a simple sketch, not a finished proof. The various generalizations I allude to above might not be possible. I am not planning, at this point, to further articulate the matter. linas (talk) 17:26, 25 April 2009 (UTC)

• If you try to "unmask Arbatsky's principle", your explanation should work at least in the particular case of quantum mechanics. But, in fact, it doesn't...
• In your article you explain the equality of angular speed and of standard deviation of generator. Why do you call this "principle"? And why do you associate it with Arbatsky's name? Arbatsky's paper is about something different. The "certainty principle" is expressed by INequalities (either (5) or (10)). And, IMHO, importance of those inequalities comes exclusively from their physical interpretation.
• I checked, which pages link to your article and found an interesting discussion. There you call those inequalities "golden truth" and allege that they are present in standard textbooks. But this is not so. If they were known, there would be no reason to read Arbatsky's paper, IMHO.
• Currently, your article works as search engine spam. I found your article because I want to read something about Arbatsky's principle. But, instead, I found here a discussion of some intermediate result from Arbatsky's paper. I am not even sure that Arbatsky pretends to be the original author of that result.
• You have not answered the last question from my first post. Do you agree that Arbatsky proved some UNcertainty relations for coordinate and momentum and for angle and angular mometum, that were not known before? Hunshi (talk) 18:43, 30 April 2009 (UTC)

Whatever. I gave you honest, truthful answers to all of your questions. I am sorry that you are so unhappy. linas (talk) 19:01, 30 April 2009 (UTC)

Hunshi invited me to express my opinion about this discussion. Ok, let it be so.

• Professional scientists are smart enough to see that Linas is not an expert in the topic. So, if he really wants to keep the article in this form, there is no problem in that.
• On the other hand, the problem is that Linas uses Wikipedia space not only for the discussion of a purely scientific question, but also to spread insinuations about supporters of the certainty principle. This is against WP:NOT policy, especially WP:NOTMYSPACE.
• So, I would recommend Linas to remove the text, starting from "There's been argumentation..." and finishing by "Now, on to the actual math and physics behind it".
• If Linas ignores this recommendation, I would suggest to nominate the article for deletion. Rcq (talk) 19:53, 31 May 2009 (UTC)

Random Message

Hi, I'm a stranger. I'm shamelessly editing your Talk page because I wanted to say hi, and your website said this was the way to do it. I found you while looking for at people in the UChicago WP users group, and I see that you're working for Novamente. That is awesome. How is it? I've seen a few talks by Ben at various H+/AGI conferences, and I'm rooting for you guys. It's always great to meet people who care about AGI. Hope you're doing well :-) Take care! --Parijata 12:25, 27 April 2009

Hi, Thanks, its going well! Just this week, I wired up a little chatbot that hooks up the core NLP processing subsystem to OpenCog. It doesn't do any reasoning yet, but it can remember, and answer questions about, certain simple kinds of statements. Visit the IRC chat channel #opencog on the IRC network freenode.net; I'm there most of the time. linas (talk) 20:58, 27 April 2009 (UTC)

BTW, behind the now-demolished Woodward, maybe half-way from there to Ida Noyes, there was/is a little student coffee shop in a university building -- what is it's name? Was it called 'Frog and Peach'? (per Streeb-Greebling) linas (talk) 21:07, 27 April 2009 (UTC)

Your exponential generating function

Linas,

I just wanted to point out that your exponential generating function using a sum over natural numbers and the Möbius mu function is connected to the Riesz function - as published in 1916 by Marcel Riesz. In particular, your function - sorry. I don't have time to do thid properly in TeX - so in Maple notation: f(x) =sum(mu(n)* exp(-n*x), n=1..infty) would appear to be big-OH(x6(1/2+\epsilon)) if RH is indeed true -the argument is essentially that of Riesz - and is a straightforward consequence of of vrious properties of the Mellin transform. See also: Wilf. I hope that this is both comprehensible and helpful. Hair Commodore (talk) 12:45, 5 May 2009 (UTC) I would usually send this kind of thing by email - but I did note the warning on your Home Page!

Hi, Thanks. Sounds plausible to me ... Remind me again, where did I write about this generating function? I know that, in the mid-late 19th, early-mid 20th century, Riesz and dozens of cohorts explored all kinds of sums of this kind, transformed and mangled every-which way. At the moment, I can't remember where I would have discussed this particular one, although I'm sure I contemplated it somewhere ... As to email: you just need to know my secret addr, which is linasvepstas at gmail dot com. linas (talk) 02:59, 6 May 2009 (UTC)

Sorry Lnas - on the discussion page for Moebius mu function Hair Commodore (talk) 10:36, 6 May 2009 (UTC) titled "xponential generating function?".

Another Random Question

Hi Linas. I am interested in some information related to your text about the "Non-Computability of Thought", specifically in the sentence of the paragraph:

- The novel twist to this game is this: suppose I have two computers, connected with some communications channel. Suppose the two computers run at clock speeds that are independent of each other (they are not synchronized). Lets call the ratio of the clock speeds R, which is a real number. Suppose R is in fact one of these uncomputable numbers. Then these two computers can compute things that are traditionally non-computable. The most trivial of these are R itself (both computers generate the strings of alternating ones and zeros 010101010101... one of the two computers interleaves these. The interleave is R itself (actually, its 4/5th's of R, but lets not quibble))

How do you arrive to the conclusion that the interleave is 4/5 of R? Could you give me some clue about it? A paper, a reference,... Or simply some explicit way to view the reasoning behind it, the mathematical proof...

Thanks for your atention

greebzz@gmail.com —Preceding unsigned comment added by 83.44.190.100 (talk) 10:34, 24 May 2009 (UTC)

I'm not sure how I got 4/5th's, I assume it was an easy calculation. Try working some examples by hand ... fix r to be small .. say 1/8 or 1/10 or 1/16 and just work the examples w/ pencil & paper, and see what you get. (there's some chance that 4/5 is erroneous.)

I wrote that when I was young and naive. I now think that investigations along the lines of quantum finite state machine (QFA) are far more fruitful -- it gives something concrete to work on. The connection to distributed processing is given by the history monoid or trace monoid -- I think that these, appropriately tweaked for non-rational clock frequencies, give something similar to the QFA (although this alone would probably be a publishable result). linas (talk) 17:01, 25 May 2009 (UTC)