For what it's worth, Firefox has been my daily driver for years and I'm very happy with it. The new tab design doesn't bother me at all.
Thank you for this reply, and trip down memory lane; I hope it is well received. I'm not trying to be a contrarian, I just wanted to interrupt this glass of Metamucil to share an alternative viewpoint.
[0] https://addons.mozilla.org/en-US/firefox/addon/sidebery/
Does anyone have an up-to-date user CSS that hides the top tabs and moves the toolbar up into the dead space? I used to have a snippet that worked, but it broke a few dozen versions ago.
https://gist.github.com/eproxus/f95c2d2bd0f9b8f5d2b6f464e27c...
On the other hand, Firefox has been steadily losing tons of users and market share for the past decade so maybe people really do hate it?
Yeah, even with your Bayes mention I'm gonna go ahead and get a second opinion on your statistical analysis there. I know, there's always someone in every crowd who quibbles with a poster's perspective, but you're hanging your entire sentiment on this one.
To rephrase what I was saying, I think that when we read 10 comments, of which nine are negative, we think "90% of people must dislike this". I don't think most people would actually think 90%, but I do think we "get a feeling" it's a very high percentage. We get this impression that the 10 comments are by 10 random people, so it's close to a random sample. But it's not, it's a very biased selection. The very post itself acts as a selection filter, only selecting people with certain properties. In this case, people with strong opinions about Firefox's UI.
In short, the denominator isn't 10, but something much much greater. And this isn't a stastical analysis (not yet anyways!), just a statistics/psychology hunch :)
I don't. When I see something like that with such a small sample size, I think that the downvoters just got there first. That could be because the YouTube algorithm put the video in front of people who happen to have higher standards, or who are bitchier, or who are familiar with the poster and their disruptive complaints over on the Slack for $X. Could be anything...anything other than generalizable overall quality.
And your statement does show this bias in a different way. You comment on the 10 being too small a sample size. That implies that if there were more posts, say 90 vs 10, you would draw some conclusions about the question "Do people like the new UI?". In the same way you're saying that the sample would be biased because someone posted on YouTube, I'm saying it will _always_ be biased, regardless of size, because the post itself (and all posts) attracts a biased sample. A larger sample is good for a random sample, but I think posts attract a sufficiently biased sample by their very nature, so increasing the sample size won't help.
We just get used to it later with the flatness and then they make it more flat. The cycle continues.
And even more exactly, it's "9 people out of all Firefox users who saw this on Hacker News and cared enough to leave a comment dislike this UI"
But yeah it seems like every release of ff there's a vocal group saying how it's going to pot and it's all doom and gloom.
I have been using ff since it was called Phoenix and it's fine. I fire up Chrome and Edge out of necessity for work and have never been compelled to switch.
Maybe I just don't need whatever it is that Chrome has that makes it so special but all the browsers are pretty decent these days.
At some point around IE11 I realized I wasn't writing W3C standards HTML/JS and then having to hack it for noncompliant browsers. I wrote code once and it just worked.
Exactly this. All posts on the internet act as a selection filter; the people most likely to comment are those with strong feelings on the subject. When we read comments, we tend to think "wow, I just read 10 comments and all but one hated the UI! People must really dislike this UI". But our sample isn't random, it's heavily biased! And we develop a biased understanding of the general population -- the vast majority of whom won't have left a comment at all. Instead of thinking "9/10 people dislike this UI", it's probably closer to "9 people out of all Firefox users dislike this UI". And more exactly, it's "9 people out of all Firefox users who saw this on HackerNews dislike this UI." Actually, I wonder if you can use Bayes' theorem to get the actual population's distribution somehow...