5 ms·

Microsoft: Firefox's billion claim is 'interesting math'

13 points by davecardwell 17y ago

jacquesm 17y ago
A billion downloads, not a billion users. If there are only 200 million users and they kept up with 1 point release in 5 then you'd already have a billion downloads.
- rottencupcakes 17y ago
  I download the latest trunk on all computers I use and all OSs on those computers. The latest trunk build automatically downloads and installs the newest nightly build every day. I've been using firefox since version 0.5 (2002) and I've probably been running the latest-trunk for half of that time. Conservatively, I've downloaded firefox > 1000 times. Just sayin'.
  - andreyf 17y ago
    Yeah, every time I see that stat, it just sounds like a great argument for diff-updates: http://blog.chromium.org/2009/07/smaller-is-faster-and-safer-too.html http://blog.chromium.org/2009/07/smaller-is-faster-and-safer...
    - sid0 17y ago
      Firefox does do diff updates.
    - kinetik 17y ago
      Firefox already uses bsdiff to generate binary diffs for update distribution. Courgette looks like a promising way to shrink those updates even further. (Details in the discussion linked from: https://bugzilla.mozilla.org/show_bug.cgi?id=504624 https://bugzilla.mozilla.org/show_bug.cgi?id=504624) I can't find a reference, but I think the "one billion" count only counts full downloads; updates installed via "Check for updates" aren't included in the count.
steiger 17y ago
I went to Mozilla presentations at the last FISL (Forum Internacional de Software Livre / Free Software International Forum) here in Brazil. They explained clearly that one billion downloads don't mean one billion users. And, honestly, I don't see nothing wrong in this kind of marketing. Even though the download counter isn't a reliable source for the exact number of users, it's still reasonable to estimate a software's growth based on the growth in number of downloads between versions. I can't think of a better, marketable way to promote Firefox based on it's growth.
- mike_organon 17y ago
  But saying a billion isn't the same as giving an estimation of growth. It just seems like a misleading (even if true) statistic.
  - Herring 17y ago
    It's probably meant the same way as McDonalds saying so many billion served. A lot of people come in more than once.
  - steiger 17y ago
    If a billion downloads today is much larger than what it used to be, then yes, I would see it as an estimation of growth. I see easily how can it mislead people, but, really, I don't see it as enough reason to not promote the software that way. 1 billion downloads is really something to celebrate, and I would celebrate it even if it has a potential to mislead some people.
thwarted 17y ago
Classic issue of "how do you measure the penetration of open source". I've been installing Firefox as part of my distribution's packaging system for the last few releases. Since this is a download from a distro mirror, Firefox can't really include these statistics, but merely estimate. I'm not sure this qualifies as "interesting math". On the other hand, the only place to legitimately get Internet Explorer is from Microsoft, so it's easy for Microsoft to count.
chrischen 17y ago
"If you're satisfied with what you're doing and you're not particularly curious about new technology and don't really care, upgrading sounds like a hassle..." That's why developers should make it worth the hassle by simply not supporting them. Twitter is a good example of a way to do it. At least with the old design their rounded corners didn't show up in IE and only showed in firefox, safari, and maybe Chrome.
Niten 17y ago
In other news, Microsoft considers addition 'interesting'.

13 points by davecardwell 17y ago

A billion downloads, not a billion users. If there are only 200 million users and they kept up with 1 point release in 5 then you'd already have a billion downloads.

rottencupcakes 17y ago
I download the latest trunk on all computers I use and all OSs on those computers. The latest trunk build automatically downloads and installs the newest nightly build every day. I've been using firefox since version 0.5 (2002) and I've probably been running the latest-trunk for half of that time. Conservatively, I've downloaded firefox > 1000 times. Just sayin'.
- andreyf 17y ago
  Yeah, every time I see that stat, it just sounds like a great argument for diff-updates: http://blog.chromium.org/2009/07/smaller-is-faster-and-safer-too.html http://blog.chromium.org/2009/07/smaller-is-faster-and-safer...
  - sid0 17y ago
    Firefox does do diff updates.
  - kinetik 17y ago
    Firefox already uses bsdiff to generate binary diffs for update distribution. Courgette looks like a promising way to shrink those updates even further. (Details in the discussion linked from: https://bugzilla.mozilla.org/show_bug.cgi?id=504624 https://bugzilla.mozilla.org/show_bug.cgi?id=504624) I can't find a reference, but I think the "one billion" count only counts full downloads; updates installed via "Check for updates" aren't included in the count.

steiger 17y ago

I went to Mozilla presentations at the last FISL (Forum Internacional de Software Livre / Free Software International Forum) here in Brazil. They explained clearly that one billion downloads don't mean one billion users. And, honestly, I don't see nothing wrong in this kind of marketing. Even though the download counter isn't a reliable source for the exact number of users, it's still reasonable to estimate a software's growth based on the growth in number of downloads between versions. I can't think of a better, marketable way to promote Firefox based on it's growth.

mike_organon 17y ago
But saying a billion isn't the same as giving an estimation of growth. It just seems like a misleading (even if true) statistic.
- Herring 17y ago
  It's probably meant the same way as McDonalds saying so many billion served. A lot of people come in more than once.
- steiger 17y ago
  If a billion downloads today is much larger than what it used to be, then yes, I would see it as an estimation of growth. I see easily how can it mislead people, but, really, I don't see it as enough reason to not promote the software that way. 1 billion downloads is really something to celebrate, and I would celebrate it even if it has a potential to mislead some people.

thwarted 17y ago

Classic issue of "how do you measure the penetration of open source". I've been installing Firefox as part of my distribution's packaging system for the last few releases. Since this is a download from a distro mirror, Firefox can't really include these statistics, but merely estimate. I'm not sure this qualifies as "interesting math". On the other hand, the only place to legitimately get Internet Explorer is from Microsoft, so it's easy for Microsoft to count.

chrischen 17y ago

"If you're satisfied with what you're doing and you're not particularly curious about new technology and don't really care, upgrading sounds like a hassle..." That's why developers should make it worth the hassle by simply not supporting them. Twitter is a good example of a way to do it. At least with the old design their rounded corners didn't show up in IE and only showed in firefox, safari, and maybe Chrome.

Niten 17y ago

In other news, Microsoft considers addition 'interesting'.