4 ms·

Those graphs are misleading. The first one visually indicates that Chrome currently has something like 100x the share of Firefox -- while the numbers indicate i

by chasing 13y ago

Those graphs are misleading. The first one visually indicates that Chrome currently has something like 100x the share of Firefox -- while the numbers indicate it's more like 3x. You should at least have the base of the graphs at 0%. Probably even better to have to top line of the graphs at 100%. That'd give a truer visual representation of the data.

cleaver 13y ago
Whenever I see one of these graphs that is not zero-based, I'm forced to expend the unnecessary mental effort to decide whether the author is intentionally deceptive or merely stupid. Usually, to save the expenditure of this increasingly rare resource, I stop reading.
zobzu 13y ago
they also take prefetching as unique hits. only chrome does prefetching by default. makes results extremely misleading
dspillett 13y ago
That and the use of "Windoze" without a hint of irony makes me label this an unreliably biased source. As much as I'd love to se IE8- vanish and don't care what happens to IE9+, if this article said the summery sky were usually blue I'd want to verify the facts with other sources before taking them as stated.
- putlake 13y ago
  The blog post links to several other sources of browser market share. You are free to look them up. I have nothing against IE and there is no cause for bias in something like this. "Windoze" is general humor, it's not irony. Check http://www.isitironic.com http://www.isitironic.com
  - dspillett 13y ago
    You seemed to have missed the words "without a hint of irony" in my post. The use of "windoze" did not give me the impression is was used purely in jest (intentional irony or some other form of humour). It, like other such respellings such as "M$ Widnows", tends to scream BIAS HERE, ENGAGE LEVEL 2 SALT PINCH BEFORE READING FURTHER in most cases.
putlake 13y ago
No it won't. I hate non-zero based graphs as much as the next guy. Except when necessary. In this case, a customized Y-axis was necessary to really show the magnitude of the drop. 4 percentage points dropping out of a scale of 100 won't even register. But it is a HUGE drop and you need a graph that amplifies the Y-axis to show the difference.

by chasing 13y ago

cleaver 13y ago

Whenever I see one of these graphs that is not zero-based, I'm forced to expend the unnecessary mental effort to decide whether the author is intentionally deceptive or merely stupid. Usually, to save the expenditure of this increasingly rare resource, I stop reading.

zobzu 13y ago

they also take prefetching as unique hits. only chrome does prefetching by default. makes results extremely misleading

dspillett 13y ago

That and the use of "Windoze" without a hint of irony makes me label this an unreliably biased source. As much as I'd love to se IE8- vanish and don't care what happens to IE9+, if this article said the summery sky were usually blue I'd want to verify the facts with other sources before taking them as stated.

putlake 13y ago
The blog post links to several other sources of browser market share. You are free to look them up. I have nothing against IE and there is no cause for bias in something like this. "Windoze" is general humor, it's not irony. Check http://www.isitironic.com http://www.isitironic.com
- dspillett 13y ago
  You seemed to have missed the words "without a hint of irony" in my post. The use of "windoze" did not give me the impression is was used purely in jest (intentional irony or some other form of humour). It, like other such respellings such as "M$ Widnows", tends to scream BIAS HERE, ENGAGE LEVEL 2 SALT PINCH BEFORE READING FURTHER in most cases.

putlake 13y ago

No it won't. I hate non-zero based graphs as much as the next guy. Except when necessary. In this case, a customized Y-axis was necessary to really show the magnitude of the drop. 4 percentage points dropping out of a scale of 100 won't even register. But it is a HUGE drop and you need a graph that amplifies the Y-axis to show the difference.