On Tue, 2015-02-10 at 22:44 -0800, Russ Allbery wrote: > Whether that was intended or not, that's not what people actually did when > they made those keyboard layouts. They did not put the dot multiplication > sign on that key; they put the middle dot symbol on that key. Arguing like that, we could never every fix any bugs, or badly designed behaviour... "it is as it is and we won't change"... disturbing o.O I wouldn't now of any real typographical use of the MIDDLE DOT. It's not intended to be a bullet character and it's mostly misused as dot multiplication operator, which is why Unicode even denotes that. Most people that do an AltGr+., likely want the dot multiplication sign, just as most people wo press AltGr+e want the Euro-Currency-Sign and not the similarly looking ⋹ (U+22F9 ELEMENT OF WITH TWO HORIZONTAL STROKES) as most people wo press capital S on a latin keyboard layout want the character S (U+0053 LATIN CAPITAL LETTER S) and not Ѕ (U+0405 CYRILLIC CAPITAL LETTER DZE). > Those keyboard layouts now exist, and changing something like that is > almost never worth the trouble, Again, with that argumentation one could never fix bugs or improve anything. Apart from that, neither of these special character is probably used on a daily basis, most people likely even don't know which character is intended for what purpose and they just look for some Unicode character which looks like what they want to have. So I doubt that there would be any major troubles by correcting this bug. And it seems much more troublesome to me that (most) people will accidentally continue to use the wrong character forever. But hey, why does Debian support Unicode at all... things were so much better when we just hat 7bit ASCII, "*" or "x" are sooo much better characters for multiplications, aren't they? o.O > for exactly the same reason why we're not > going to remap the default keyboard layout to be Dvorak even if it's > better than QWERTY. I guess you compare apples with rocks to construct some "argument"... if you completely change the base keyboard layout from QUERT[Y|Z], which people have learned and are used to for years and which is used by most input devices... THIS would be a major problem. But changing a special alternativ keyboard mapping, for a character which is likely anyway not used by most people and for which it is most clearly that those who use it likely want the dot multiplication sign, is probably not the end of the world. > If you want a different keyboard layout, you pretty much need to make a > different keyboard layout, and then convince people to use that instead of > the existing one. Changing an existing one, even if you think it made a > logical error, seems like a really bad idea. Is there any special reason why you think it would be bad idea? Like a use case where people depend on having the MIDDLE DOT and which would break if it was changed? > That's even apart from the fact that diverging from upstream in an area > like this seems like an absolutely awful idea. That, I fully agree. But as I've said, I just wasn't sure what is actually the canonical place for keyboard mappings, and where I could ideally report this (which is why I had to use general). > I don't think this is an actionable bug report for Debian. It's an > interesting bit of speculation, and it's arguably a consistency flaw, but > it's not something that makes sense for us to do anything about. So many bugs reported in Debian are actually upstream bugs... > (Based on past history, I suspect the reply to this will be 200 lines > about why I'm wrong, so I'll mention in advance that I'm highly unlikely > to say anything further on this bug report and I'm happy for others to > have the last word.) Yeah, I'm out either... it seems more and more often that contribution to Debian is not really welcomed anymore. Cheers, Chris
Attachment:
smime.p7s
Description: S/MIME cryptographic signature