> if you put too many icons in a row and they are all different, they won’t work.
uhh what? says who? where's the source of this? This is only the case if your icons don't look different enough. In the example those icons aren't great because they've overly complex and easy to mix up.
>you won’t come up with a great icon for a commit. Or a release. Or an issue. Or a license.
You can, and you also don't necessarily need to. Icons aren't only for the illiterate or first-time users... they create a signpost for repeat users to scan for without reading a word. As long as you consistently use the same icons and they make at least a little sense, they can be useful (I agree that some of Github's don't make much sense at all though).
>I mean, Icon + Label + Counter make for a symmetric and weak composition:
It's best to refrain from insults during critique, but this is flat out dumb. Completely unfounded. No evidence. The composition is "what is it" and "how many" — it makes perfect sense.
>The thing with vanity metrics is that there should be just one. One metrics is simple to understand and focus.
Again, no. This is how _you_ use it. You've done no research. I want to see forks because I want to know how many people are potentially engaged in _doing_ something with a repo. Stars matter less to me because there are some repos that just have a lot of stars because someone wrote an interesting article about it once. (I do think stars and watching is a bit redundant, I personally have no need for both but my personal experience does not solely dictate what the product needs to be)
> Watch button
Again, no. `Watch` is an action. If I click the dropdown I expect to have options for "watch". "Not watching" simply adds words and makes me put forth a little cognitive effort (minimal but it's there) to parse what the opposite of "not watching" is. You are adding the need to think more, not less.
--
Honestly I'm too exhausted to continue. This redesign isn't well thought out at all. At this point I skipped ahead to the end, and honestly it's just a garbled mess. You don't understand the product or its users. If you were my student (I'm a design professor) I'd give you credit for the effort, and the visual design is fine (though are you trying to trick me into thinking that making corners more round is meaningful?)... but there's an endless mountain of straight-up bullshit to call out here.