Just wonder: does it include downloads of 1.5 Alphas and Betas? :-S

minghong, no, it does not.

- A

nice :) well done all

Finally a post where I see no Opera trolls.

Congratulations! Well Done. :P

Wait for it...

Here I am :D

Heh. There you are! :-D

David Naylor is now an Internet Sock Puppet!.

Poor FF groupies...

It'll be nice to see 100,000,000 before 1.5. Will the counter be reset for Firefox 1.5 or carry on? If it's not being reset, how will upgrades get discounted?

I can personally account for about 100 of those downloads if not more and I still don't think Firefox is all that great. After reimaging my computer about that many times (Due to testing other software). I'm glad I help make Asa happy looking over his inflated downloads counter.

@cybermagellan ... at least you helped recover the lost downloads count for the 450 workstations i installed with just one download ... thanks

...and the 160.000 workstations I installed with just 1 download.

...and the 1.750.000 workstations I installed with just one download.

90 000 000 downloads are impressive, even though some users download it multiple times. That's the case with every download figure (quicktime, opera etc.).

Hope it reaches 100 million before the 1st anniversary ;)

Asa, you need to come up with a standard disclaimer anytime you post Firefox download stats. Something like the following:

"Yes, I realize downloads do not translate directly to usage or market share. It is possible for someone to download multiple times and use it once. It is possible for someone to download it once and install multiple times. It is possible for someone to download it and not use it at all. It is possilbe for someone to download it from some other source and use it. It is possible for someone to download the source code and compile it themselves.

"Regardless, download numbers and the change in the download rate over time can still be interesting to look at. They don't paint an entire picture of anything, but are merely another data point in looking at the general trend of browser popularity. Since it is impossible to know with certainty what the usage percentages or market shares are of any given browser, statistics such as number of downloads can still provide another angle into the nebulous topic of general browser market trends."

This is great, really it is, but these stats are getting increasingly difficult to decipher what this means in terms of success. Lets get this into context now, most of the following are a rough idea and mostly correct now I think:

Not all mirrors are counted
1) 1.5 beta downloads and more aren't counted
2) 1 download for 5+ users on 1 computer isnt counted
3) 1 download for network deployment on 10-10,000 computers is not counted
4) 1 download for all the users derived from installs from CDs, portable devices and more aren't included
5) Users on OS's with Firefox preloaded such as Linux's arent included

The user numbers is getting higher and higher everyday which is market share, whilst the download count is getting further and further from that figure, thus more and more irrelevent.

Users is now well over 115 million, roughly 10%+ usage/market share of the web. So the 2 figures are getting further and further apart, one a true reflection of success and adoption, the other not.

Come next year the stats are going to be so far apart its completely irrelevent, misleading, and merely celebrating how much Mozilla bandwidth has been used. We could have 1 trillion downloads and only one user.

I for one am far more pleased by accurate figures that reflect how much Firefox is being adopted and used overall in the grand scheme of things. Thats what its all about. That is more accurate in terms of usage and success, and needs to become more accurate in terms of the figure itself. Approx 115 million users, 10+% of the web, wow, now were talking, in in perspective.

@Toby Johnson

You cannot seriously expect that from Asa, do you?. Professional behaviour is not well received here.

Well, download counts are a lot easier to collect. An accurate count of actual users? That's tough, and it has its own set of problems: You need a representative collection of sites for collecting the stats. You need to account for target audience*. You need to account for UA cloaking. You need to be able to identify individual users who might be sharing an IP address. You need to avoid counting the same user twice when they get a new dynamic IP address. You need catch those even when people refuse cookies, or delete cookies between sessions.

It's a lot easier to just add together the downloads from official mirrors and filter out likely upgrades (IIRC downloads using using Firefox aren't counted) and just present that number for what it is, without claiming to be more.

*Target audience: My personal website, which includes a fairly busy comic book fan site, is running about 76% IE and 16% Firefox for September, with other browsers at 3% or below. The numbers are different, but the order matches most of the marketshare stats that get published. My alternative browser site is currently running 57% Firefox, 23% Opera, and 10% IE.

Is that the alternative browse alliance site Kelson, if so we've discussed that before on Sfx, great work and congrats on that, a real testiment to recent events too ;)

It is easier to get reasonably accurate usage/market share stats than made out above, this has been discussed here before and specific ways were given by people, but I can't remember off hand what they were, be greatful if someone could mention a few. Sites and more are getting more and more accurate with market share statistics, its clearly possible and fairly accurate, and more accurate then simply downloads, of which some downloads arent users, and 1 download is quite likely to be 1000 users, so what does it mean after all of that, and with that getting more common. Again, I think Mofo are looking at this more now and hopefully taking it seriously, but really, download stats are getting more and more in-accurate and irrelevent to success and overall usage of a product. We have to acknowledge that, and the fact thats becoming more and more true everyday.

hey minhong, here's another Opera troll, a proud one. But sadly you're too a stupid fanboy

90.443.698 / 315 = 287.122 Downloads/day.
(100.000.000 - 90.443.698) / 287.122 = 33 days.
ETA 100 mio must be around 24. oct 2005.

Nice to see lots of localisations updated so quickly ( http://www.mozilla.org/products/firefox/all )

@Opera troll, you're one funny troll who doesn't even show up your name.

Back to the topic, if you think I was really trolling there, then you're really stupid. IMO those were constructive replies.

Kerz has got TypePad auth available for these blogs, right?

- Chris

Does this include all versions 1.0 through 1.0.7?

Not trying to troll, just curious. 90,000,000 downloads either way is pretty big.

Congrats.

Even if the download statistics are inaccurate, 100,000,000 would still be pretty darn cool. :)

Thumper: yes, TypeKey auth is available on mz hosted blogs - I have it enabled on the Folding Blog. Its upto the authors to enable it on individual blogs.

I'm pretty sure you're not talking about version 1.07, right? So the 90,000,000 must include all versions? Let's see, I've used every version since 0.7. That makes 11 downloads just for one of my computers, so you can divide that number by 11 (55 would be more accurate if you count all my computers).

Another update ought to put us over the top for 100,000,000. It's nice to celebrate download numbers, but they don't mean a darned thing.

New stats: 97224150/334~291088 downloads per day
In about about 9.5 more days from now, the counter will be at 100000000!
Even thought those numbers aren't really accurate, it still would be cool.