Wikipedia:Reference desk/Archives/Computing/2024 March 15

= March 15 =

Aspect Ratio Problem (ShotCut)
I have an mp4 file of something that was originally widescreen but has been reformatted to fit an old-fashioned TV. That is to say, it has gone from (I think) 16:9 to 4:3. The effect is that all the characters in the clip appear taller and thinner than they really were in the original.

Does anyone know if there is a filter (in ShotCut, ideally) which could correct that problem and return the file to its original aspect ratio? Or another way to address the problem? AndyJones (talk) 12:31, 15 March 2024 (UTC)
 * VLC-player for example offers under its tab "Video" options to format-per-view. --46.114.5.149 (talk) 15:06, 15 March 2024 (UTC)
 * Thank you. ShotCut doesn't have a option called that, or indeed called anything that sounds to me like it would be the place to look! However FWIW I have found a workaround, which is that if you open the clip in isolation in ShotCut (as distinct from adding it to the playlist), then you go into its properties and amend the aspect ratio there to 1600 x 900, then you export the result, you get what I was hoping for. AndyJones (talk) 17:22, 15 March 2024 (UTC)

When I am typing the words stop appearing
This happens on several sites when I am typing in a box like this. It doesn't happen on Wikipedia because there are no ads. I don't know if there is a fix even if I can get through to the people who can fix the software.

I'm not sure how I figured it out but it seems to happen when an ad appears on the screen. Not all ads, though. And I don't notice right away that what I typed stopped appearing because I look at the keyboard, something my typing teacher told me not to do, then I don't realize the words didn't show up.— Vchimpanzee  •  talk  •  contributions  •  23:18, 15 March 2024 (UTC)


 * A totally non-expert reply from someone who experiences the same phenomenon – it may be that your device is either quite old or quite small in memory terms by current ever-advancing standards, and is unable to allocate sufficient resources to running everything in complete parallel. In the long term, applications get more numerous and more resource hungry (which in part drives the need to update hardware). I also seem to notice more frequent updates of OS and security systems these days, which in combination slow my ageing PC for a significant proportion of my usage. {The poster formerly known as 87.81.230.195} 51.198.186.221 (talk) 00:12, 16 March 2024 (UTC)


 * I do get high memory usage warnings but the computer is quite new. Well, two years old. Wow, times passes quickly. Anyway, I made a note of that on my user page.— Vchimpanzee  •  talk  •  contributions  •  15:10, 16 March 2024 (UTC)
 * The fact that it's exacerbated by ads probably means it's a graphical rendering issue, where your browser neglects updating the text box because it has to focus on rendering the ad, which is much more resource intensive. This definitely shouldn't be happening on a computer made in 2022, even if it is low end, so it sounds like your browser is not taking full advantage of your hardware. As a start, make sure hardware/graphic acceleration is enabled on your browser. If it isn't, all the work of rendering has to go to the CPU, which can bog down and possibly cause graphical errors like what you're experiencing.  Pinguinn     🐧   07:50, 22 March 2024 (UTC)

When did electronic computers first allow input in base 8+ instead of just manual binary?
With a button to load your machine instruction or tell it you want it in this address, after you've entered it by physical binary toggle switches. Sagittarian Milky Way (talk) 23:51, 15 March 2024 (UTC)


 * I think you are asking about compilers which turn human-readable source code into machine code. You may find answers at History of compiler construction. The first implemented proto-compiler was written by Grace Hopper of blessed memory, with her A-0 System of 1951/52. The first compiler in the modern sense was for the Manchester Mark 1 in 1952, written by Alick Glennie who worked at Manchester with St. Alan Turing. MinorProphet (talk) 13:52, 16 March 2024 (UTC)
 * Wasn't there a time when you could only input raw machine code but had a keyboard with all the octal or maybe hexadecimal or decimal numerals? And before that you had to flick switches one way or the other (one switch per bit) then press something? Sagittarian Milky Way (talk) 14:23, 16 March 2024 (UTC)


 * From the BINAC manual, 1949:
 * "KEYBOARD: A device for translating manual key strokes into "computer language". There are eight keys, representing the octal numbers zero thru seven, each of which when depressed, produces a unique set of binary pulse codes (3 pulse combination). Keyboard is used to introduce either the "program" or quantitative data into the computer and memory."
 * AndyTheGrump (talk) 15:25, 16 March 2024 (UTC)
 * By "8+", do you mean binary-coded decimal? Typically, numerical input was by means of peripheral input devices reading prepared punch cards or punched tape in which numbers were represented in the form of a sequence of codes for the decimal digits and possibly the sign and the decimal point (or comma). Even if the internal arithmetic was binary, as for the Z3, the input and output used decimal representation. Assuming you mean a form of inputting data, one word at a time, through a row of toggle switches on the computer's console that are manually set by the operator, another question is, when were electronic computers first equipped with such switches? --Lambiam 16:01, 16 March 2024 (UTC)
 * Fascinating stuff about BINAC. Bubba73 You talkin' to me? 01:04, 17 March 2024 (UTC)


 * Looks good. But I think that BINAC's approach of interfacing with the hardware directly in this way was already being superseded. See Manchester Mark 1: "In 1948 Alan Turing devised an encoding scheme based on the standard ITA2 5-bit teleprinter code, which allowed programs and data to be written to and read from punched tape." This included a keyboard which controlled the punching mechanism. The concept of using a keyboard goes back at least to Émile Baudot's printing telegraph of 1874, on which ITA2 was based. MinorProphet (talk) 16:40, 16 March 2024 (UTC)


 * The 1943 Colossus computer used punched tape for data input, though it was 'programmed' through hardware settings rather than any sort of input device. AndyTheGrump (talk) 17:27, 16 March 2024 (UTC)


 * The Z4 (computer) was like a sophisticated programmable calculator. It had a special program construction unit for programming it and used decimal floating point for the external world though it worked in binary. No need to work in binary or even octal or to toggle switches. But then again it was mechanical not electonic. NadVolum (talk) 18:43, 16 March 2024 (UTC)
 * The PDP-11 I worked on around 1979 had toggle switches that could be used to enter machine code in binary. You toggle in a value, then press a LOAD button and it would store that value in memory. Then toggle in the next value, press LOAD again and it stores it in the next memory location, etc. But this was usually only used for entering the bootstrap program; there was no non-volatile memory so you had to key in the bootstrap program every time the computer was powered on. After it was booted, it ran Unix and you could write programs in a high-level language, run a compiler, etc. So this machine did allow only binary input for the bootstrap program, but acted more like a modern computer after it was booted. CodeTalker (talk) 22:13, 16 March 2024 (UTC)


 * Lambiam asked, "when were electronic computers first equipped with such switches? You mean, after [[File:Glen Beck and Betty Snyder program the ENIAC in building 328 at the Ballistic Research Laboratory.jpg|center|thumb|this type of thing? (ENIAC)]] According to Z3 (computer) "The ENIAC computer, completed after the war, used vacuum tubes to implement switches and used decimal representation for numbers. Until 1948 programming was, as with Colossus [1943], by patch leads and switches.""The Manchester Baby of 1948 along with the Manchester Mark 1 and EDSAC (both of 1949) were the world's earliest working computers that stored program instructions and data in the same space. In this they implemented the stored-program concept which is frequently (but erroneously) attributed to a 1945 paper by John von Neumann and colleagues." MinorProphet (talk) 22:39, 16 March 2024 (UTC)
 * I assume the expression "physical binary toggle switches" in the OP refers to manually set switches, as seen in the fourth image here for the console of an EL-X8, and here for an IBM System/360 Model 40. --Lambiam 00:25, 18 March 2024 (UTC)
 * So, Edsger Dijkstra of the eponymous algorithm is the guilty party. Fab pix of the System/360, beautiful in their own way... The Propellorheads knew all about knobs and switches. MinorProphet (talk) 15:14, 19 March 2024 (UTC)
 * Teletypes pre-dated computers, so the concept of a single key that transmitted a whole letter (or any value more complex than 0 or 1) already existed. -- Verbarson talkedits 22:59, 16 March 2024 (UTC)