Talk:Unicode input: Difference between revisions

Revision as of 18:53, 22 September 2020

Latest revision as of 01:05, 8 September 2024

(14 intermediate revisions by 9 users not shown)

Line 1:

{{WikiProject banner shell|class=C|

{{WikiProject Writing systems |importance=Low}}

}}

== Windows EnableNumKeypad clarification ==

Line 240 ⟶ 244:

Yes, I absolutely agree there is just misunderstanding here, not an argument. I believe Peter Brown has some fundemental error and I really am trying to be helpful in correcting it, though it is very hard to tell exactly what his error is. The basic question is why he started talking about 448, either implying that mod-256 can turn 980 into 448, or that for some reason 448 has fewer possible results of mod-256 than 980, when in fact both of them turn into the exact same number, 192.

I think your math expression is possibly messed up as you use letters in the last one that don't appear in any others, thus it's unrelated. But yes y=f(f'(f''"(x))) defines a function that applies f''" to x, f' to that result, and f to that result, and could be written as a new function h such that y=h(x).

You are wrong about what happens when a file is sent to Japan. All the software under consideration is storing the resulting unicode code points in the file, not the numbers the user typed, and the file will display the same there.

Line 254 ⟶ 258:

There is a further confusion in that 192 is not used as a Unicode code point, but instead it is used to index either the "ANSI" code page or the "OEM" code page.

If the ANSI code page is used, and it is set to [[CP1252]] (which it usually is), then 192 turns into U+00C0 or {{char|À}}. Thus 192, 448, and 960 all turn into the same character{{char|À}} in these programs. Most of CP1252 matches Unicode, including ___location 192, for these locations you can pretty much say the 192 is turned directly into Unicode.

If the "OEM" code page is used (which appears to be the case "in Notepad or in the Wiki edit box") it looks at ___location 192 in [[CP437]] (or some similar page), and gets U+2514, which is {{char|└}}. Thus 192, 448, and 960 all turn into the same character{{char|└}} in these programs.

I would be very interested in what happens if {{keypress|Alt|0}}{{keypress|9|6|0|chain=}}, ie with a zero prefix, is typed "in Notepad or in the Wiki edit box". This may cause 192 to be chosen from the ANSI code page and get {{char|À}}. Or it might cause Unicode to be used.

[[User:Spitzak|Spitzak]] ([[User talk:Spitzak|talk]]) 16:23, 22 September 2020 (UTC)

:"The basic question", according to Spitzak, "is why he started talking about 448". I have answered that. I switched my discussion from 192 to 448 because "with 448, I needed only to exhibit the one symbol {{char|ǀ}} rather than both {{char|└}} and {{char|À}}." I was trying to make things simpler, but clearly I failed.

:I am now retired, but I worked in computers 1972 –2003 and don't recall any talk of numbers turning into each other, which Spitzak seems unable either to avoid or to define. This is unnecessary for our purposes, however, as {{u|John Maynard Friedman}} has provided an excellent statement of the processes under consideration, not once using the troublesome phrase "turns into" or mentioning moduli as being "applied". I suggest that we leave it at that.

:In Notepad, {{keypress|Alt|0}}{{keypress|9|6|0|chain=}} yields {{char|À}}, from the Windows code page. As the section {{section link|Windows_code_page#ANSI_code_page}} notes, these are officially known as "Windows" not "ANSI" code pages. According to [https://web.archive.org/web/20181208141313/https://msdn.microsoft.com/en-us/goglobal/bb964658.aspx#a MSDN], the latter "is nowadays a misnomer that continues to persist in the Windows community." I suggest avoiding misnomers.

:[[User:Peter M. Brown|Peter Brown]] ([[User talk:Peter M. Brown|talk]]) 19:25, 22 September 2020 (UTC)

::"undergoes a transformation operation" is math jargon for 'turns into', if you must be purist about it, The question is made doubly difficult by Microsoft's track record of playing ducks and drakes with standards (aka "embrace and undermine") without us getting picky about choice of words when the meaning is obvious. That is why I'm suggesting that terminological exactitude is critical. --[[User:John Maynard Friedman|John Maynard Friedman]] ([[User talk:John Maynard Friedman|talk]]) 19:58, 22 September 2020 (UTC)

:::I'm torn between just letting the discussion die and objecting to being called "picky". After wrestling with the issue, I'm afraid that I come down on the side of protest. "960 turns into 192" cannot be parsed as "960 undergoes a transformation operation 192" unless "transformation" and "192" are in [[apposition]], which is clearly not the intent. My suggestion that it meant 192 = mod (960,256) was apparently incorrect, but it was an honest attempt at making sense of "turns into". To me, the meaning was and is not obvious. So long as I'm not called "picky", I'm content to let it remain obscure.  — [[User:Peter M. Brown|Peter Brown]] ([[User talk:Peter M. Brown|talk]]) 02:22, 23 September 2020 (UTC)

::::192 = mod (960,256) is exactly what I meant by "turns into". You take 960, apply the mod-256 operation to it, and you get 192. So the number 960 turns into the number 192. It is possible this terminology is popular in computer science because the input is often not needed after the operation and is discarded, or may even be replaced by writing something like {{code|1=x = mod(x, 256)}} which replaces x with it's value mod 256, actually turning x into a new number.[[User:Spitzak|Spitzak]] ([[User talk:Spitzak|talk]]) 02:36, 23 September 2020 (UTC)

== Direct drawing on a touch screen ==

CJK characters are routinely entered by drawing directly on a touch screen. Meaning that I have seen it being done more than once so it must be routine. {{smiley}}

Does anybody know enough to add a section to that effect? [[User:JMF|𝕁𝕄𝔽]] ([[User talk:JMF|talk]]) 15:56, 3 August 2023 (UTC)

== alt + x Windows ==

alt + x works also in Notepad 11.2112.32.0 on Windows 10, not sure about previous versions. [[Special:Contributions/213.184.17.126|213.184.17.126]] ([[User talk:213.184.17.126|talk]]) 15:48, 6 September 2023 (UTC)

alt + X does not work on my HP laptop running Windows 11. [[Special:Contributions/136.36.180.215|136.36.180.215]] ([[User talk:136.36.180.215|talk]]) 01:04, 8 September 2024 (UTC)

== Missing: UTF-8 hex input ==

The article doesn't currently mention it, but UTF-8 is the de facto standard for representing Unicode in computer systems. PHP, and perhaps other languages, have built-in ways to specify Unicode characters of any length (including exotic combinations of glyphs and properties) using hexadecimal number literals. Most online Unicode character descriptions include UTF-8 representations. So shouldn't the article reflect this reality, instead of keeping alive the mostly outmoded concept of code points? The most compact way to represent Unicode characters of any byte length greater than one is through UTF-8 hexadecimal. [[User:David spector|David Spector]] ([[User Talk:David spector|talk]]) 17:09, 29 January 2024 (UTC)

:I have not seen any input methods that use UTF-8 code units. Maybe you can use {{tt|\xNN}} for each byte in a string constant in some languages, but this is pretty uncommon.[[User:Spitzak|Spitzak]] ([[User talk:Spitzak|talk]]) 19:04, 29 January 2024 (UTC)

== Where? ==

"Microsoft Windows has provided a Unicode version of the Character Map program, appearing in the consumer edition since XP" Where is this found?[[Special:Contributions/136.36.180.215|136.36.180.215]] ([[User talk:136.36.180.215|talk]]) 01:05, 8 September 2024 (UTC)