Keyboard silliness (was Re: [MVLUG] meeting next Monday?)

g gordon at
Thu Mar 3 19:37:08 MST 2011

On 2011.03.03 1708, Samat K Jain wrote:
> On Thursday, March 03, 2011 08:47:14 AM g wrote:
>> Since when has Linux behavior eschewed the 90% rule, ignoring de facto
>> standards and "gracefully degrading" instead, which winds up being
>> user-hostile to a huge majority of the population? (I know, but for
>> those that don't: at times it seems that the approach to usability in
>> Linux follows a philosophy of "If that guy can't have<user friendly
>> feature>, nobody can!")
> I don't follow you… what is the "90% rule" in this case? Over 50% of personal computers sold since 2005 have been laptops, the vast majority of which don't have dedicated numpads. Without a way to detect whether a dedicated numpad exists, the 90% rule is Numlock should be OFF.

Eh, I was suspending the sales data of the home PC market temporarily 
for the sake of facetiousness...

That said, the correct implementation of the 90% rule here would be to 
infer that the presence of a battery, "mobile" video chipset, or some 
such similar hardware likely indicates a laptop (which usually lack 
dedicated numpads), so disable NumLock. Otherwise enable it for the rest 
of the users. End result is that 90% of us have the correct setting.

This detective work, of course, really only applies to systems with 
legacy hardware and needn't be used on those with USB or similar 
keyboards that can quite happily announce their layouts when asked.

I should perhaps add that I'm generally less concerned with solving the 
problems than I am with ranting about them, since complaining about the 
quirks of our chosen OS is more fun and noncommital than building and 
submitting diffs.


More information about the MVLUG mailing list