User talk:Gah4

I have hatted the discussion with you on the coupling coefficient talk page as it is really tangential to the issues there. We can continue the discussion here. You wrote

This is obviously not true. If you use V=IR, and change the sign of one of V or I, and not both, the sign of R changes. I suspect, though, that people try harder to define V and I such that R is positive.

Let's take this example (please excuse the ugly ascii art)

         I
    o---->----|
              |
              |       /\
              _       |
            |  |      |
            |  | R    | V
            |  |      |
             -        |
             |        |
             |        
    o--------

By Ohm's law, V/I = R. So far so good. If I define I in the opposite direction then I have to write V/(−I) = R because I is no longer the current going into the resistor relative to the direction V has been defined in. This still results in a positive value for R because I will take on a negative value when V is positive. A similar results is obtained if the voltage is defined in the opposite direction. To write V/I = R for these situations is simply wrong. Ohm's law is not the ratio of any old voltage and current, it is the ratio of the applied emf and the current driven by the source. That is, the current has to come out of the source and in to the load. The fact that you obtain a negative result is itself enough to show that that is wrong because it is unphysical (before you point it out, I am aware of negative resistance, but that is a different thing from Ohm's law). SpinningSpark 12:03, 17 January 2017 (UTC)

I don't have much doubt that you understand. Yes, as you note not counting negative resistance, the result has to be physical. The case that I am indicating comes when you first write down the equations, and then put in the values. If, for example, you write a problem in matrix form, you might find that there are negative values in the matrix, though the resistance is positive. In your ascii art, if I ask for V/I then the result is "-R", a negative number representing a positive resistance. Gah4 (talk) 14:03, 17 January 2017 (UTC)

Well if you define a matrix in a random way, then of course you can get negative values, but that doesn't mean that a matrix element actually "represents" a positive resistance. You have just gone and defined something that has no physical reality (but obviously it is related to the physically real quantity). When we set up the matrix for a two-port we always define the currents in a consistent way – going in to the port. A one-port is an even simpler case of that (a resistor by itself is a one-port). Following this rule will always give the correct sign for the port driving point impedance. A result that contradicts Joule's law, conservation of energy, and the third law of thermodynamics just has to be an incorrect result. SpinningSpark 15:20, 17 January 2017 (UTC)

Of course the Xerox Alto is a historically important machine, but is that what retrocomputing means? It's a vague term; usually the definition is somewhere in the vicinity of old computer hardware that people keep running for nostalgia reasons or develop for as a hobby. The Alto doesn't have much hobbyist support, as far as I can tell. Dgpop (talk) 03:47, 21 February 2017 (UTC)

It happens that I know some of them. The http://www.livingcomputermuseum.org has some actual machines, and some emulated machines, connected up to 3 Mb/s ethernet, through a bridge to the Internet. In its prime, the Alto wasn't produced in large numbers, but was very popular with those who had one. As I noted, it was one of the original GUI machines, but also where Ethernet originated. There are two emulators, SAlto, and ContrAlto. All the software was released to the computer history museum, which makes it much more useful for retrocomputing, but that was only a few years ago. I put a note in talk:Retrocomputing about it. Gah4 (talk) 04:24, 21 February 2017 (UTC)

Please see Talk:SHA-1#.22Announced_an_attack.22_vs_.22performed_an_attack.22. Smyth (talk) 14:30, 5 March 2017 (UTC)

The IBM 1401's card reader could be used in Column Binary mode, which stored two characters in every column, or one 36-bit word in three columns when used as input device for other computers. However, most of the older card punches were not intended to punch more than 3 holes in a column.----------- this sentence is probably correct, and refers to machines such as the 514, not keypunches -so it need clarification. IBM didn't generate more than 3 punches/column thus no reason to build machines for lace cards- repeated firing of all 80 relays wasn't anticipated ------

The multipunch key is used to produce binary cards, or other characters not on the keyboard. -- Here's citation; ---- IBM OPerator's Guide, 22-8484-3 1955 Type 24 Card Punch, page 18. " Multiple digits may be punched manually in one column by holding the space bar down while the keys are depressed. On some machines, a multiple punch key is supplied for this purpose".....Enjoy, {{citation needed|date = March 2017 73.71.159.231 73.71.159.231 (talk) 06:15, 20 March 2017 (UTC) — Preceding unsigned comment added by 73.71.159.231 (talk) 06:10, 20 March 2017 (UTC)

You don't seem like the type who would do something like this for no apparent reason. Makes me wonder if your account has been compromised. -- ChamithN (talk) 06:53, 1 April 2017 (UTC)

I don't disagree with your edit, however I believe that in the compiler as distributed the phases were linked as overlays, so loading one overlayed the previous. Actually there were at least two layers of overlays, so there were main overlay phases and subphases that were also overlays of each other. I think it was a usermod to link the compiler as one executable without overlays. Peter Flass (talk) 08:04, 8 June 2017 (UTC)

Having not thought about this in detail for some years, I believe it uses LINK, LOAD and/or XCTL to do what IBM calls dynamic overlays. That is unlike the static overlay used, for example, by Fortran H. As well as I know it, again it is many years, with LINK the OS decides which to keep in memory and which not to. In 44K, I suspect that each one overwrites the previous, but with more memory it doesn't have to. (The article describes this along with the ability to run in 44K.) I suspect that you understand LINK better than I do. They are all separate load modules, maybe 100 of them, not just a single load module with overlays like Fortran H. Gah4 (talk) 15:20, 8 June 2017 (UTC)

Thinking about it some more I believe you're right. It pretended to be overlays. Once you got to larger memories and could putbthe modules in LPA I believe it ran much faster. Peter Flass (talk) 21:03, 8 June 2017 (UTC)

It does LINK and LOAD and DELETE. In some cases, it keeps the disk address for a module, to speed up access. Some might be only needed once, and so are DELETEd. Yes LPA should speed it up. Without them in LPA, I don't know how it does it. It sounds like it knows if it is small memory, and does things differently. What I forget is, with LINK, when the OS removes them from memory and when it doesn't. Gah4 (talk) 22:50, 8 June 2017 (UTC)

Hi again,

https://en.wikipedia.org/wiki/User_talk:Matannnn#patterns Mathematically speaking, a word is also a pattern, but I think using the word "pattern" in the context of introducing aho-corasick is quite almost wrong or misleading.

I'm new to Wikipedia talk so sorry if my message here is posted against convention (the UX here is kind of IRC age quality...).

Matan

(you forgot to sign your message with four tildes)

As well as I know it, the convention in CS is to call it a pattern, and the field is called pattern matching. Word is a special case of pattern matching. The problem I was working on when I learned about Aho-Corasick uses DNA sequences and not words of human speech. The problem is general, though often used for actual words. More specifically, look up combinatorial pattern matching. Gah4 (talk) 12:14, 13 January 2018 (UTC)

Greetings Gah4,

Hazyj and I are having a discussion about this topic on my talk page. Feel free to join. It is the last topic on the page, right now. Constant314 (talk) 23:33, 16 March 2018 (UTC)

Hi. Thank you for your recent edits. An automated process has detected that when you recently edited Electric charge, you added a link pointing to the disambiguation page Mole (check to confirm | fix with Dab solver). Such links are usually incorrect, since a disambiguation page is merely a list of unrelated topics with similar titles. (Read the FAQ • Join us at the DPL WikiProject.)

It's OK to remove this message. Also, to stop receiving these messages, follow these opt-out instructions. Thanks, DPL bot (talk) 09:21, 14 April 2018 (UTC)

Thanks for "produces" heat in electric current -- better than "releases", which was there before. My longer version simply attempted to make the mechanism clear - but no problem with the shorter one.Sdc870 (talk) 09:23, 22 April 2018 (UTC)

Hi... You reverted my change on the grounds that Armstrong invented "wide band FM" which is significantly different from Idzerda's invention. I disagree that it's significantly different. It's the same transmission scheme; wide vs. narrow is a difference in degree, wider bandwidth makes for greater signal quality at the expense of increased spectrum occupancy. But it's FM either way. In addition, as I noted in a comment on the FM modulation article where that same comment was made, I disagree that what Armstrong invented is specifically wide band FM. His earliest patents are clearly narrowband, given that they describe (in a somewhat muddled way) a modulation index near one. Paul Koning (talk) 19:02, 14 June 2018 (UTC)

I also changed the text so that it says that Armstrong invented wide-band FM, instead of just FM. Since wide-band FM is what we call FM today, it seems significant. I didn't try to trace back whether Armstrong knew about Idzerda, which might matter for patent rights. Greater signal quality, as far as I know, is the reason Armstrong wanted FM. I suspect that he didn't know about capture effect, though. The increased spectral width dilutes the effects of noise. Could Idzerda have invented wide-band? I don't know. Find some references about it. This was about the time that the ability to work with frequencies high enough became available. Discuss on the talk page (maybe you did) about this. I suspect, though, that a lot more sources credit Armstrong, even if they are technically wrong. (I have lost many discussions where I knew something was right, but the popular sources disagreed.) It would not have been politically popular to put a wide-band FM signal in place of the AM band. That would reduce the number of stations by a factor of about 20, besides forcing everyone to throw away their radios. Note that the initial band that Armstrong used was taken away, making the radios of early users worthless, but those were much fewer than outstanding AM receivers. As well as I understand it, though I didn't try all that hard, Idzerda relied on existing receivers detuned to demodulate his FM. But the FM advantage as we know it, was already lost. But discuss in talk to see what the consensus says. Gah4 (talk) 20:05, 14 June 2018 (UTC)

Note in addition that the page mentioned is FM broadcasting not Frequency modulation. Wide-band is specifically what is used in the former. You might see about changes to the latter. Gah4 (talk) 20:16, 14 June 2018 (UTC)

Thanks. I have some answers to some of the issues you mentioned; I'll bring them to the article talk page. On spectrum occupancy, remember that Idzerda was running what appears to be the world's first commercial broadcast station (starting 1919). So greater bandwidth was hardly an issue yet. He used existing radios; FM can be received (without the noise benefits, however) using slope detection.

As for popular sources, maybe, but that's not the Wikipedia way supposedly. The historic record is quite clear about what Idzerda did, for one thing he has both Dutch and US patents (from 1922 and 1927 respectively). The interesting question is what the differences are, and -- if it can be learned -- whether they knew of each other's work. So far I see no evidence that they knew. Paul Koning (talk) 01:29, 15 June 2018 (UTC)

I have noticed that many things are invented just a little bit earlier than they should have been. Before the technology was quite ready. Wide band FM below 1MHz might just not be obvious enough, and the available tubes likely didn't go high enough. Armstrong was specifically trying to reduce noise, and believe that a whole new system would be viable, along side the AM broadcast band. It is almost surprising today, to consider that the ancient technology AM band is still around. Why not dump it, the same way as analog TV was dumped? Exactly why Armstrong came up with wide-band, and Idzerda didn't, I don't know. Maybe you can find some WP:RS to explain it. It does seem that it might go into Frequency modulation, though you should probably discuss it in the talk page there. Gah4 (talk) 01:54, 15 June 2018 (UTC)

Special:Search.

-Moxy (talk) 04:41, 19 June 2018 (UTC)

Hello Gah, I thought it best to let you know that I have reported User:FF-UK for edit-warring on Mains electricity by country (see Wikipedia:Administrators'_noticeboard/Edit warring#User:FF-UK reported by User:CplDHicks2 (Result: )), but really this is encompassing all of his past behaviour. Please comment as you see fit. Thanks. CplDHicks2 (talk) 05:47, 9 July 2018 (UTC)

I have had various disagreements, and some have done sock-puppet investigations. Otherwise, I don't know. Gah4 (talk) 06:01, 9 July 2018 (UTC)

You've always been a peacemaker, Gah, which I appreciate; you don't have to comment if you don't want to. I just wanted to let you know that the discussion is there. :) Cheers, CplDHicks2 (talk) 06:07, 9 July 2018 (UTC)

Well, I have had some disagreements with you, too: for example. Statements like As such I removed them, and they will stay removed. Isn't especially conducive to working together. Gah4 (talk) 07:18, 9 July 2018 (UTC)

In the discussion at Alkali metal, you have changed your !vote from 'neutral' (timestamp 07:49) to 'oppose' (22:14). This may be confusing and contradicting. I suggest for clarity, you make clear below the first post that your !vote has changed, possibly with a reference to other posts. You could write, like:

Neutral changed to Oppose, see posts below re Dirac66. ~~~~ [rest of first post unchanged here]

-DePiep (talk) 06:57, 7 March 2019 (UTC)

Thanks. Since they were specifically not votes, I wasn't worried, but I will change as you say. Gah4 (talk) 08:03, 7 March 2019 (UTC)

Inserting WP:COPYLINKs can get you blocked from editing. Earlier editors, such as Dmcq have already opined that the link violates IEEE's copyright. Glrx (talk) 20:45, 8 March 2019 (UTC)

Well I'd certainly count links to the standard as that, they want to be payed for their copyright work and it should not be distributed for free. However I don't know if access to the working party drafts are supposed to be restricted, I think they may not be. People pay for the proper up to date standard rather than something which can differ at odd places from it. I believe Vincent Lefèvre is a participant in defining the IEEE floating point standards, he can probably check to give the definitive answer. Dmcq (talk) 21:10, 8 March 2019 (UTC)

I don't know about Vincent Lefèvre, I heard it from David Hough, 754r@ucbtest.org, who is, or at least was, the chair of 754WG. I sent e-mail to the address on the SOA for ucbtest.org, and he replied. I also wrote to Kahan, but no reply from him. I suppose I could ask for a signed and notarized statement, but I don't think he would be too happy about doing that. Gah4 (talk) 21:25, 8 March 2019 (UTC)

It's normal for the authors to share a copy of the drafts even if IEEE claims exclusive copyright on the final version. As with preprints of academic papers, the authors often place a copy on their personal website but it's better to use an institutional repository for durability (the curators also check copyright status). I see he published as Oracle employee though, so he probably doesn't have one such repository and would need to use either a wider repository like Zenodo or arxiv or the institutional repository of some university co-author. Nemo 06:18, 10 April 2019 (UTC)

It seems that ucbtest.org is owned by either David Hough or 754WG. It looks like its purpose is to distribute IEEE754 test suites, which I suspect gave its name ucbtest. Sun used to have places to distribute public software. I don't know if Oracle kept up with that. Zenodo used to be on the blacklist, rumors are that it isn't any more. Gah4 (talk) 08:13, 10 April 2019 (UTC)

RMaung (WMF) 16:38, 10 September 2019 (UTC)

RMaung (WMF) 15:39, 20 September 2019 (UTC)

RMaung (WMF) 20:40, 3 October 2019 (UTC)

Following our discussion on units: I have put this here because it is completely irrelevant to the talk page discussion.

Which is heavier: a pound of gold or a pound of feathers?

Answer tomorrow or earlier if you answer and I see it. 86.164.61.122 (talk) 14:18, 23 November 2019 (UTC)

I knew about this one a long time ago, I believe in high school years, but had to look it up to see. From troy weight: The British Empire abolished the 12-ounce troy pound in the 19th century, though it has been retained (although rarely used) in the American system. If you read in more detail, the US Troy ounce has the same value (weight) as the British Imperial Troy ounce, but I don't believe the same name. That is, the US version isn't Imperial (in name), but just Troy. So, OK, the pound of feathers is heavier as long as you are in the US, or maybe your gold is more than 200 years old. Thank! Maybe there should be Fun with units! Gah4 (talk) 01:00, 24 November 2019 (UTC)

You got it! The troy system is still very much in use, so the pound of feathers is heavier. All precious metals are still weighed in troy pounds and ounces in places that do not weigh them in grammes (and there are quite a few metric countries that still weigh gold and silver in troy weight - Switzerland springs instantly to mind). I even have some electronic scales that weigh in both avoirdupois ounces and troy ounces (and troy pennyweight). I have little doubt that these same made in China scales are sold in the US, so they will weigh avoirdupois and troy ounces equally differently in the US.

Incidentally, not so well known is that explosives were also weighed in troy pounds and ounces unless you had a large enough quantity in which case you had to revert to avoirdupois tons as there is no troy measure larger than the pound. Nowadays, metric countries weigh them in metric units. 86.164.61.122 (talk) 13:53, 24 November 2019 (UTC)

Hi Gah4, thanks for all you do on Wikipedia, and for all your help at lighting-related articles. My you have a wonderful Christmas and a Happy New Year. (and if you don't celebrate Christmas please feel free to take that as a Happy Hanukkah, a great Dhanu Sankranti, a blessed Hatsumode, or whatever holiday you want to insert there.) Zaereth (talk) 08:55, 25 December 2019 (UTC

The article Color Developing Agent 4 has been proposed for deletion because of the following concern:

Not notable.

While all constructive contributions to Wikipedia are appreciated, pages may be deleted for any of several reasons.

You may prevent the proposed deletion by removing the {{proposed deletion/dated}} notice, but please explain why in your edit summary or on the article's talk page.

Please consider improving the page to address the issues raised. Removing {{proposed deletion/dated}} will stop the proposed deletion process, but other deletion processes exist. In particular, the speedy deletion process can result in deletion without discussion, and articles for deletion allows discussion to reach consensus for deletion. Qono (talk) 04:27, 10 January 2020 (UTC)

The article Color Developing Agent 3 has been proposed for deletion because of the following concern:

Not notable.

While all constructive contributions to Wikipedia are appreciated, pages may be deleted for any of several reasons.

You may prevent the proposed deletion by removing the {{proposed deletion/dated}} notice, but please explain why in your edit summary or on the article's talk page.

Please consider improving the page to address the issues raised. Removing {{proposed deletion/dated}} will stop the proposed deletion process, but other deletion processes exist. In particular, the speedy deletion process can result in deletion without discussion, and articles for deletion allows discussion to reach consensus for deletion. Qono (talk) 04:27, 10 January 2020 (UTC)

So, why did you remove chapter "Danger of IPv6" from IPv6 article? — Preceding unsigned comment added by Acbaile5 (talk • contribs) 22:06, 30 March 2020 (UTC)

Discuss on the talk page, hopefully others will also come along. And don't forget to sign with four tildes, talk page writing. Gah4 (talk) 22:09, 30 March 2020 (UTC)

Interesting, but is this enough for an article? It might be combined with discussion of other techniques such as memory-mapped I/O, Multics "everything is a segment", and any others. Peter Flass (talk) 17:03, 23 April 2020 (UTC)

Locate mode works both at low level and higher level. It is supported by PL/I, I believe COBOL, and, it seems, from CMS Pipelines. I don't know much about Multics at all. I suppose memory-mapped I/O, as it is visible from high-level languages. Gah4 (talk) 18:57, 23 April 2020 (UTC)

From what I know about Multics you don't open a file, you "initiate" the segment, and then after that it's strictly memory-mapped. I wish I had more time to play with Multics.Peter Flass (talk) 20:10, 23 April 2020 (UTC)

So this is the way they do it in high-level languages, or is this the low-level implementation? Since locate mode is visible to the user in high-level languages, it seems more reasonable to have its own page. Do we need a page on, for example, QSAM, which I believe also have an option for locate mode? Gah4 (talk) 08:53, 24 April 2020 (UTC)

OK, there is a page for QSAM but it doesn't mention locate mode. I suspect that locate mode is more interesting, this is, in need of an article, than QSAM. Gah4 (talk) 08:56, 24 April 2020 (UTC)

Maybe in Data buffer I could write something about buffering techniques in OS/360, and include this also. Peter Flass (talk) 13:22, 30 April 2020 (UTC)

I suppose that would be a good place. I was hoping to be more general than just OS/360, but it does seem to be a lot art. Well, it is a little harder with FBA disks, as the block boundaries don't match the record boundaries, but it should still be possible to arrange for one fewer copy operation. OS like Unix and Windows keep, somewhere, a buffer of disk blocks, and then copy data to/from those blocks before/after writing/reading them. And the user program, in the usual case, copies data to/from its own buffer. Gah4 (talk) 16:31, 30 April 2020 (UTC)

It also reminds me of a feature in IP networking that never caught on, called trailers. In the usual case of a TCP/IP stack, TCP fills up a buffer, which it puts a TCP header on. Then it passes it to IP, which puts the IP header on, usually copying. Since buffers usually have space at the end, but not the beginning, there was an idea to put the TCP and IP headers on at the end, instead of the beginning, which would save some copy operations. It might be that someone figured out a way around that, though. Similar in that it tries to avoid excessive data copying. Gah4 (talk) 16:31, 30 April 2020 (UTC)

Does Linux copy? It should be possible to just remap the buffer from kernel space to user space. I keep forgetting to look this up. The C RTL might copy data or just move the pointer, depending on the API. Peter Flass (talk) 21:00, 30 April 2020 (UTC)

I pretty much don't know the internal of Linux at all. You can only remap whole pages on page boundaries. C has the lower level I/O routines fread() and fwrite() which read or write a buffer of a given length. That pretty much implies copying it. At the system-call level, there are the unix (not C library) routines read() and write() that read/write data to/from a given buffer address. In the case of 9 track tape, and other tape systems that allow variable block size, read() and write() will read/write whole blocks. In the case of disks, other than raw disk I/O, they have to go through the file system buffer. There is no alignment requirement for the buffer. I suppose the system could be designed for page-aligned, page-length buffers to remap, and higher level systems that know about it could do that. I believe I remember systems from the 1970's where direct access I/O was always in multiples of 512 bytes, or in the case of the PDP-10, 128 36 bit words, but Unix and Windows allow direct access on arbitrary boundaries, so it must go through a disk block buffer. This is the big advantage of CKD disks. That you can make disk blocks any size you want, and read/write directly from/to disk. Gah4 (talk) 00:02, 1 May 2020 (UTC)

OK, I just found the page Zero-copy which seems to be the modern term for it. Mostly it seems to describe network I/O, and not disk I/O though. Maybe the appropriate description of locate-mode could go there, and a redirect to it? Gah4 (talk) 00:24, 1 May 2020 (UTC)

(aside) Zero-copy says " a program could instruct the channel subsystem to copy blocks of data from one file or device to another without the nucleus having to copy the data to an intermediate buffer." I'm not sure this was ever true. Peter Flass (talk) 16:19, 1 May 2020 (UTC)

I suspect that I don't know the details all that well, but as far as I know, the nucleus never copies data. Access methods run in user space and problem state, so they don't count as nucleus. To avoid as much as possible, consider a program with one buffer, use EXCP to read a block into it, and EXCP to write the block out somewhere else. Now put that into a loop. No copying by the nucleus or user. But I think mostly locate-mode is important, as it can be done at a higher level. Consider a Fortran program to copy some data from one file to another. A read/write loop will copy data from an input buffer into an array in the program, then, to an output buffer. A PL/I program could use locate-mode for one, and move-mode for the other, so only one copy. I believe Fortran uses BSAM and PL/I uses QSAM, if that matters. Does move-mode QSAM do a copy inside the access method routine? Gah4 (talk) 17:42, 1 May 2020 (UTC)

A discussion is taking place to address the redirect Opteon. The discussion will occur at Wikipedia:Redirects for discussion/Log/2020 May 18#Opteon until a consensus is reached, and anyone, including you, is welcome to contribute to the discussion. _signed, Rosguill ^talk 18:35, 18 May 2020 (UTC)

Yeah, no worries re the LIGO edit. It's complicated by both the lack of clarity in the LIGO article - reading deeply, it seems that the 1994 'start' was really several years of administrative wrangling rather than actual construction - as well as the lack of specific details on LIGO's own website. 1999 seems like a reasonable compromise, but a compromise it is. The precise date - or even year - really isn't that important in the overall scheme of things, I think. Cheers. Anastrophe (talk) 19:17, 25 June 2020 (UTC)

Either SLAC or LIGO people can change it. I had thought, from reading LIGO, that it was 1990-1994 building the building, and after that building the insides. (The latter takes longer.) I am in Seattle, and two years ago went out to see LIGO west. I have known about it for a long time, having actually gone to a Kip Thorne lecture about it once, but haven't thought about the building schedule. Gah4 (talk) 19:58, 25 June 2020 (UTC)

Getting pretty far off topic for Talk:Three-phase electric power, but delta secondaries are commonly used in arc furnace transformers, rectifiers, and other places that take advantage of the lower winding current compared to terminal current of a delta connection. Sometimes power distribution transformers have a"buried" delta winding, used to control harmonic currents and to stabilize line to ground voltages on certain faults. But how much of all electrical engineering can we include in this one article? --Wtshymanski (talk) 00:20, 19 August 2020 (UTC)

The reason I thought about it was the mention of parallel transformers. Yes, I don't know how much is needed for the article, which is why I put it in the talk page to discuss. Gah4 (talk) 00:25, 19 August 2020 (UTC)

Posted here as it irrelevant (like most of your discussion) to the Regency TR-1.

Neutralisation is a subject in which I have a good level of experience and design knowledge.

I had the job of designing a variable frequency oscillator operating from 1 to 36 GHz for the purposes of testing a radar for resistance to jamming. For this purpose a large klystron was required (solid state devices not being able to withstand the 2 kW transmitter pulse from the radar). A klystron in this application is fundamentally a disc seal triode operating in a variable sized cavity. It was necessary to neutralise the Miller capacitance, though the phase and magnitude of the signal on the klystron's anode varied widely as the cavity was tuned over the range. The design ended up with a number of cams on the tuning shaft which continuously varied the magnitude and phase of the neutralisation, not to mention the anode voltage and the oscillation mode of the klystron. Setting all this up was a nightmare. Luckily, only the one was required. 86.164.169.96 (talk) 14:44, 20 October 2020 (UTC)

If you are still interested in klystron design, you might see: https://accelconf.web.cern.ch/pac97/papers/pdf/9C004.PDF. Gah4 (talk) 04:04, 2 November 2020 (UTC)

Hi, this is totally random, but I happened to see your response to a question on the talk page at Talk:Hamiltonian where you said "so some say" qualifying it as an improved formulation of CM. Why did you put it that way? Just curious, everyone I know greatly prefers Lagrangian and Hamiltonian mechanics vs Newtonian, it is far simpler once you know what you're doing. The only argument I know of against it is that it is less intuitive. Again, just curious, thanks! Footlessmouse (talk) 02:43, 2 November 2020 (UTC)

Oh, mostly just trying to be WP:NPOV there. I suppose I believe that for many problems plain Newtonian methods are fine, but for some the other ones get the answer faster. Gah4 (talk) 03:56, 2 November 2020 (UTC)

The talk page at Industrial and multiphase power plugs and sockets is for discussing improvements to that article, which is about plugs and sockets. A digression on collections of light bulbs is off-topic there. A little thought and you may recall that a neutral connection is not required if no line-to-neural voltages are used; you could have three loads, say 1 ohm, 2 ohms, and 3 ohms, connected across each pair of A B C phases, the result would be unbalanced loads, different currents in each phase, but no requirement for a neutral at all. For example, I worked in an industrial plant with an abundance of 3-phase outlets which were used indifferently for cooling fans (nearly balanced motor load) or single-phase welding machines, bout as far from a balanced load as there is. No neutral was required as neither of these applications needed a phase to neutral voltage. --Wtshymanski (talk) 01:33, 3 December 2020 (UTC)

Hi,

Just wondering – since you've commented in support of my proposal (I think?) would you mind adding something to the Survey section as well about it? If you don't want to give a formal response, that's fine, but so far the majority of text is in the Discussion section. I'd like it to be obvious to whomever (or should that be whoever?) is going to close it which option each person who participated was for :) Regards, DesertPipeline (talk) 05:05, 20 February 2021 (UTC)

Hello, Gah4. This is a bot-delivered message letting you know that Draft:Leakage current detector interrupter, a page you created, has not been edited in at least 5 months. Draft space is not an indefinite storage location for content that is not appropriate for article space.

If your submission is not edited soon, it could be nominated for deletion under CSD G13. If you would like to attempt to save it, you will need to improve it. You may request userfication of the content if it meets requirements.

If the deletion has already occured, instructions on how you may be able to retrieve it are available here.

Thank you for your submission to Wikipedia. FireflyBot (talk) 08:01, 20 March 2021 (UTC)

Did you really mean to put And often enough the term is disk. Well, disk drive goes back to the days of mountable disk packs, where the storage was separate from the source of rotational power. In the earliest days of personal computers, including IBM, floppy disks and floppy disk drives were what people got to know. In any case, I suspect that disk is now the generic synonym for any non-sequential access storage device, and more specifically flash memory based devices. It might even be the WP:COMMONNAME by now. in Direct-access storage device rather than in Talk:Direct-access storage device?

BTW, disk goes back half a decade before mountable disk packs. Shmuel (Seymour J.) Metz Username:Chatul (talk) 00:46, 14 May 2021 (UTC)

From a recent message on the IBM-MAIN listserv

IBM currently provides access to all IBM announcement letters starting from January 1, 1976, from this Web site:

https://ibm.com/common/ssi

As a random example, at that site you can find IBM U.S. Announcement #285-045 dated February 12, 1985, and entitled "VM/XA Migration Aid Release 2 Enhancements." You can search by product name, program number, or just about any keyword. There's also a link to a beta version of the site with a new user interface. The new UI seems to be a little easier to use for these purposes since it's easier to restrict a search to a particular date range.

Shmuel (Seymour J.) Metz Username:Chatul (talk) 10:04, 14 May 2021 (UTC)

Please just give me a few minutes to finish up.DGerman (talk) 19:53, 25 June 2021 (UTC) Done ( for now anyway) DGerman (talk) 20:05, 25 June 2021 (UTC)

Looks good to me. I have been using od -x since before all the others existed. On this OS X system, od and hexdump are exactly the same program, with different names. It checks the name it was called and acts appropriately. On my FreeBSD system, they are hard links to the same file. (On OS X, two different files.) On linux, it seems that they are very different size. I don't know why that is. One thing, though. The default for any of them on unix-like systems is to print the values of 16 bit words (octal, hex, or other base). That makes the bytes backwards on little-endian machines. Oh well. Gah4 (talk) 22:13, 25 June 2021 (UTC)

FYI, since you commented on it a while back: denormal number has been renamed to subnormal number Martin Kealey (talk) 01:35, 11 September 2021 (UTC)

 A small plip for forgetting to sign talk page messages (here and here). — LauritzT (talk) 11:35, 22 September 2021 (UTC)

A tag has been placed on Talk:List of Nobel laureates by university affiliation requesting that it be speedily deleted from Wikipedia. This has been done for the following reason:

WP:G8. You don't need the page to exist to see what links to it.

Under the criteria for speedy deletion, pages that meet certain criteria may be deleted at any time.

If you think this page should not be deleted for this reason, you may contest the nomination by visiting the page and clicking the button labelled "Contest this speedy deletion". This will give you the opportunity to explain why you believe the page should not be deleted. However, be aware that once a page is tagged for speedy deletion, it may be deleted without delay. Please do not remove the speedy deletion tag from the page yourself, but do not hesitate to add information in line with Wikipedia's policies and guidelines. If the page is deleted, and you wish to retrieve the deleted material for future reference or improvement, then please contact the deleting administrator, or if you have already done so, you can place a request here. RandomCanadian (talk / contribs) 00:07, 21 October 2021 (UTC)

• If you want to coordinate removal of these, you can leave a message at a relevant Wikiproject, or maybe even make an AWB request (WP:AWBREQ). Or you can put the list in your userspace. As you can see, the WhatLinksHere is accessible even if the page or its talk page don't exist: works without problem... RandomCanadian (talk / contribs) 00:11, 21 October 2021 (UTC)
  • I only made one edit to this page, and am not especially attached to it. I am also not especially interested in going through the 252 pages to change them. It would be nice if there was an automated way to change the articles, though I suspect that isn't so easy, depending on how well they integrate into the page. It would be nice of those who do decide to delete a page, to help change all the links. Gah4 (talk) 01:13, 21 October 2021 (UTC)

Since the Help desk topic you replied to is closed, let me answer you here. In your reply you repeated an aversion to IPA. (Even though it wasn't part of the question, as can be seen by the fact that the text in question included the “respell” “SHEE ...”).

So let me reply to your statement “note that IPA only works if you know how to pronounce the sample words.”. That is patently not true. We can even stay with the example: The sound of [ɕ] is one which I learned without any previously known sample word. Generally, if your statement were true, one could never learn any new sounds.

Of course, one can always try to use crutches like “respell” in an attempt to make one's life easier by only sticking to what one already knows and forcing anything else into the existing pattern. (My cousin did that as an infant once: For animals, they already knew “wauwau” (dog) and "gaga" (chicken, or maybe bird in general). So when there was a hedgehog in the garden, what did they call it? “Gaga-wauwau”!) We could cater to that mindset. But that, IMHO, would diametrically counter the purpose of an encyclopedia. ◅ Sebastian 14:18, 27 March 2023 (UTC)

OK, I mostly agree. The latter reminds me when our son was first learning words, he learned colors and shapes. Then would name things after their color and shape, so trash cans were blue rectangle, and diaper cream was blue circle.

If someone has a Boston accent, or southern accent, and then pronounces the sample words with that accent, it could be pretty far off.

In many cases, we learn the pronounced word first, and then later the written word, but not all of them. Especially for technical words, which we might not be in the place where they are needed.

Reminds me of the story of Michelle Yeoh, who learned the lines for Crouching Tiger, Hidden Dragon phonetically, before she learned any Mandarin. In any case, I suspect that SHEE isnt exactly right for Xi, but maybe close enough. Gah4 (talk) 21:35, 27 March 2023 (UTC)

Thank you for your nice reply. Sorry that I missed it for half a year, but luckily watchlists are patient. Did you mention Xi because I brought up the sound [ɕ], or is that just a funny coincidence? As for your son: Was ‘rectangle’ really one of the first words he learned? I would be curious how that influenced his further development, but I completely understand if you don't want to write that here for the whole world to see. ◅ Sebastian 16:46, 17 October 2023 (UTC)

I think it was because you mentioned it, but as you say, half a year ago. We did have much fun with our sun as he was growing up, and the things he would learn. And yes he learned rectangle very early. Not so much later, he learned letters, and was very interested in them. The only book we could read him was Dr. Seuss' ABC book. He didn't know about words, but just letters. He would read letters anyplace. His daycare had to have a rule that teachers couldn't wear shirts with letters, as he would keep reading the letters. And after not so long, he figured out words and reading. So, back to the subject. If the only one you learn is the written word, without anyone giving a pronunciation, you say it the way that seems right. I do remember this for hysteresis, which I pronounced wrong for a long time. Thanks! Gah4 (talk) 17:15, 17 October 2023 (UTC)

That's funny about the daycare rule! (Although I don't quite understand what the problem is with a kid reading the letters off a shirt. And I would contend that it's like the old adage with the hammer and the nail: Once you know how to read letters, you see letters everywhere.)

Thank you for explaining your original intention behind what I questioned in my first paragraph. Now I see what you meant, and I agree that there probably are many readers who just want a a rough pronunciation, for whom learning IPA requires a higher initial effort. (I'm avoiding the common misnomer “steep learning curve” here.) That said, I still doubt whether respelling does any good. It replaces one mistake that a naïve reader would make (“Ksee”) with another one that is about as bad. In addition, it is frustrating for readers if they have to clamber over three more words of gibberish before they get to the actual text about the man. (My move to a footnote hopefully improved that situation, but unfortunately introduced two other problems: Some readers who want the pronunciation will miss the footnote, while, conversely, others who don't want it will waste time looking up the footnote. BTW, IMHO, the text in parentheses still constitutes an obstacle, and I don't see why it has to be duplicated from the text in the box, but that's another question, and maybe there are accessibility reasons for that.) A better solution - which could unobtrusively remain in the main text - would be to include an OGG file with the pronunciation, which allows a learning experience we all got used to as toddlers. ◅ Sebastian 20:16, 17 October 2023 (UTC)

Thanks for your contributions to List of semiconductor materials/programs. Unfortunately, it is not ready for publishing because it has no sources. Your article is now a draft where you can improve it undisturbed for a while.

Please see more information at Help:Unreviewed new page. When the article is ready for publication, please click on the "Submit your draft for review!" button at the top of the page. BoyTheKingCanDance (talk) 05:23, 9 May 2023 (UTC)

Someone asked about making the table sort right on bandgap. That is an awk program that fixes the table so it sorts right. It doesn't have any source, as I wrote it. If someone wants to change the table sorting, they might find it useful. Is there a place where someone should put programs useful for editing a page, such as table making?

I also have awk programs to make Wikitables from IAEA tables, such as for fission fragment tables. I thought about posting them, but never did. Gah4 (talk) 05:37, 9 May 2023 (UTC)

Hi, never did this before, so not sure I am doing this right. I played around with floating point formats and found that some of the entries in the tables must be slightly off and tried to change them. You changed them back. I'm happy to discuss. I have a libre office calc sheet that I could share if it helps? Nsmeds (talk) 14:54, 10 September 2023 (UTC)

Discuss in the talk page for the article. I suspect, though, that you aren't considering that the significand is between 1 and 2. Gah4 (talk) 15:58, 10 September 2023 (UTC)

I found the talk page and made my point there in more detail. Nsmeds (talk) 19:00, 10 September 2023 (UTC)

Hello, Gah4. This is a bot-delivered message letting you know that Draft:List of semiconductor materials/programs, a page you created, has not been edited in at least 5 months. Drafts that have not been edited for six months may be deleted, so if you wish to retain the page, please edit it again or request that it be moved to your userspace.

If the page has already been deleted, you can request it be undeleted so you can continue working on it.

Thank you for your submission to Wikipedia. FireflyBot (talk) 06:01, 9 October 2023 (UTC)

Hello, Gah4. It has been over six months since you last edited the Articles for Creation submission or Draft page you started, "programs".

In accordance with our policy that Wikipedia is not for the indefinite hosting of material deemed unsuitable for the encyclopedia mainspace, the draft has been deleted. When you plan on working on it further and you wish to retrieve it, you can request its undeletion. An administrator will, in most cases, restore the submission so you can continue to work on it.

Thanks for your submission to Wikipedia, and happy editing. Liz ^{Read! Talk!} 05:42, 9 November 2023 (UTC)

I overhauled the articles on Otto Hahn and Lise Meitner after I wrote the Discovery of nuclear fission article back in 2020, but I need a co-nominator to send them to Featured. Let me know if you would be interested. Hawkeye7 (discuss) 01:00, 2 July 2024 (UTC)

Sounds fine to me. What do I need to do? Gah4 (talk) 08:05, 2 July 2024 (UTC)

Nothing, hopefully. I will respond to the reviewers but will be away in August so you might have to hold the fort for a while. It usually takes months for an article to get through FAC. The big decision was Hahn or Meitner. I've nominated Lise Meitner, as her article attracts more page views. According to Google ngrams, Hahn was four times as popular as Meitner in 1970, but Meitner is slightly ahead today. I just hope she is not too controversial. Hawkeye7 (discuss) 20:08, 2 July 2024 (UTC)

This is to let you know that the above article has been scheduled as today's featured article for 8 February 2025. Please check that the article needs no amendments. Feel free to amend the draft blurb, which can be found at Wikipedia:Today's featured article/February 2025, or to make comments on other matters concerning the scheduling of this article at Wikipedia talk:Today's featured article/February 2025. Please keep an eye on that page, as notifications of copy edits to or queries about the draft blurb may be left there by user:JennyOz, who assists the coordinators by reviewing the blurbs, or by others. I also suggest that you watchlist Wikipedia:Main Page/Errors from two days before it appears on the Main Page. Thanks, and congratulations on your work! Gog the Mild (talk) 12:42, 29 December 2024 (UTC)

I added a paragraph to the discussion of storage class in PL/I. I’d like to do something toward a “storage class” article. Maybe I could hack a draft and get comments. Peter Flass (talk) 18:46, 23 March 2025 (UTC)

Does seem that an article on storage class would be good. PL/I makes its storage class more obvious, and some of the names have later been used by other languages. C has the (rarely used) auto keyword, and the confusingly used static keyword. Fortran is somewhat strange on the ambiguity in its storage class. Gah4 (talk) 21:02, 23 March 2025 (UTC)

OK, there is now Storage class as a redirect to the page about storage class in C. So, we need to undo that. Probably the discussion should go into Talk:Storage class. I will write some there. Gah4 (talk) 21:08, 23 March 2025 (UTC)

Sounds good Peter Flass (talk) 04:14, 24 March 2025 (UTC)

Social media 101.99.247.90 (talk) 15:37, 15 September 2025 (UTC)

The issue is the general correctness of the statement that energy flow is from outside the conductor. Constant314 (talk) 15:07, 17 December 2025 (UTC)