filipf added a comment.

  In D14869#311891 <https://phabricator.kde.org/D14869#311891>, @davidedmundson 
wrote:
  
  > > what do you mean by facts though, isn't telemetry just a bunch of 
personal preferences heaped up?
  >
  > Yes.
  >  That is /considerably/ better than having just a  few people's personal 
preferences heaped up, which is the  current state of this thread.
  
  
  Everyone unfortunately thinks numbers = hard facts and this can be pretty 
erroneous and lead to manipulation. If you are going to resort to telemetry for 
making decision such as this one (which you shouldn't, but potentially only to 
support claims):
  
  1. please make sure to know what kind of sample you are operating with ie. 
who has turned ON telemetry (since I see it's going to be opt-in) -> the 
expectation here is that you'll have developers overrepresentened because 
regular users tend to dislike even the mention of telemetry
  2. make sure to record the screen resolution used AND the screen size -> of 
course people with small high-res screens are going to be more likely to 
increase font sizes
  3. make sure to take into account that not all fonts have the same metrics-> 
I don't believe you are going to build a database that would somehow categorize 
fonts in relation to how big they are on their own. Ever turned on one of those 
niche fonts out of fun, only to have to increase them to 14pt because they're 
just so small? Cantarell is smaller than Noto Sans e.g. You must take these 
things into account.
  4. above all, please take into account that you will have zero data on what 
someone's eyesight is like, which is a major variable here
  
  You are never going to get a full picture with numbers, not even if you solve 
the first 3 points suitably. Qualitative insights are much needed here, so it's 
good to have look if someone's been saying anything about fontsizes and also to 
ask e.g on Reddit to see what the users' input is.
  
  Let's not forgot the method of experiment either, if still feeling about this 
strongly, as a more sensible change to test this I'd suggest increasing the 
font DPI to 102 and then try it out in KDE Neon for a little while. There are 
too many steps and arguments being skipped here.

REPOSITORY
  R119 Plasma Desktop

REVISION DETAIL
  https://phabricator.kde.org/D14869

To: ngraham, #plasma, #kde_applications, valorie, #vdg
Cc: mart, oysteins, rooty, filipf, safaalfulaij, rikmills, harmathy, rkflx, 
abetts, davidedmundson, rizzitello, plasma-devel, ragreen, Pitel, ZrenBot, 
lesliezhai, ali-mohamed, jensreuterberg, sebas, apol

Reply via email to