HN Explorer

GitHub Discussion about the recent feed changes becomes 3rd most upvoted ever

40 points by beeman ↗ HN
The discussion [0] about the recent changes has been upvoted by so many developers, that it's currently the 3rd most upvoted discussion on the platform [1].

I've been talking about it on Twitter [2] and tried to ping a few people I know who work at GitHub, but to no avail - there has been total silence since they made the change 5 days ago.

Is there someone here on HN who can bring this to the attention of the right people? Even though I understand it's a corporation, I can't imagine nobody cares about so many developers being unhappy.

The ask is pretty simple: leave the old chronological feed [3] as an option.

0: https://github.com/orgs/community/discussions/66188

1: https://github.com/orgs/community/discussions?discussions_q=is%3Aopen+sort%3Atop

2: https://twitter.com/beeman_nl/status/1701278872858178005

3: https://github.com/dashboard-feed

40 comments

[ 3.3 ms ] story [ 129 ms ] thread
verdverm
Has GitHub responded? Are they staying the course or listening to the feedback they asked for?
beeman
They didn't respond yet.
verdverm
Did you see that recent Google open source that uses git-notes for all the extras something like GitHub adds on top of git?
justinclift
Didn't that get cancelled shortly after their announcement?
justinclift
They've finally responded. They've doubled down on their original stance, but now also claiming the change is for performance reasons.

Which makes no real sense when you read the details of their response.

Their previous approach was to provide feed events based on people's watched repos. Now they're doing it from people's starred repos instead, which for most people is a couple orders of magnitude more data items. And also happens to be the data items with no value in providing events for. :/

xonix
I've just turned off all the Feed filters to show empty feed. This way it's more useful, than with it's "new content". I've bookmarked the direct link to https://github.com/dashboard-feed, unfortunatelly on mobile the font is too small.
gered
It's really sad, but yes, this is what I ultimately ended up doing too. There was no way to configure the filters to very closely approximate what the feed was showing previously that I could find. The new algorithm always seemed to want to throw stuff at you that was slightly outside the bounds of what you were specifically following (clearly done to help you "discover" something else that you may be interested in, treating it as if this was a social media platform instead of a coding platform ...).
marcinzm
I wonder if Microsoft hired a few too many laid off Meta and Twitter product people.
politelemon
I wonder if there's a UX team somewhere that got bored and needed to justify their existence with unasked for changes and user research deliberately misinterpreted to support their stance. I'm only bitter about this having experienced it a few too many times.
Animats
Github has a social media feed?

I'm not seeing any of that. I look at files on Github. I search code. I push commits to Github. I create issues. I look at issues. I look at project pages. I look at code. I search code. Git and Cargo talk to Github on my behalf.

None of this shows me any social media stuff. If I start seeing "For You" on project pages, I'll switch to a new provider. I really hope Github doesn't pull a Facebook.

chrsig
Given the text of that first post, it's no wonder why there's been silence. What do you even do with that?

It gives no details about why whatever the new feature is bad, or what was good about the prior behavior. It's presumptive, antagonistic, and doesn't leave any room for discussion.

It really seems like OP is hoping that dogpiling on the issue will bring about positive change.

troupo
The first post is exactly what people see and feel. It is the enshittification of GitHub and needs no lengthy explanations [1]

[1] https://twitter.com/dmitriid/status/1700908879486345622?s=20

WirelessGigabit
Lol, finally someone else who hates relative timestamps.

One year ago... What does that mean? September 11th, 2022? Or a wider range?

troupo
It's especially bad for the history/commit view. "We had an incident in a large repo with multiple committers, let's narrow it down to a specific commit..."
chrsig
The departure from precise timestamps in UIs is unpleasant. It's not just github, gitlab does it as well.
nomilk
I also prefer precise time stamps but am okay with time_ago_in_words as long as the full time stamp is available on hoverover
marcinzm
>It gives no details about why whatever the new feature is bad, or what was good about the prior behavior.

Isn't that the job of the PMs at Github and not the users saying they don't like the new experience?

chrsig
Sure, but if you're looking to have a dialog with the corporation about it, bringing specific actionable feedback is going to be met better than petulant foot stomping.

Honey vs vinegar, and all.

beeman
The actionable feedback is that the old option should still stay available next to the new one.
troupo
If the corporation cared, it wouldn't have introduced the algorithmic feed with little to no usable info in the first place.
chrsig
If the corporation doesn't care, why are you spending energy complaining about it?

Are you hoping to enact change, or are you just complaining?

troupo
Sometimes corporations budge due to outcry (or regulations [1]).

GitHub's "designers" and PMs knew exactly what they were doing and they knew exactly what would happen. They still went ahead with it. No idea why you feel the need to defend them.

[1] The algorithmic feed runs afoul of upcoming EU regulations regarding this crap

chrsig
I'm not defending them, I'm being critical of the approach taken by the OP, because of the clear attempt to rile people into picking up their pitchforks.
beeman
I'd rather not be bothered by this at all and just move on, but it's something I care about - and hundreds of others with me.

It's ok that you don't care about this, so feel free to put your pitchfork down.

marcinzm
In my experience corporations respond more often to bad publicity than they do to actionable feedback in terms of larger changes like this. One big reason is that often they already got that actionable feedback during the planning of this project but decided the business upside was worth it. Bad publicity cuts into the business upside which can change their value calculation.
matt3210
Use hub here via CLI and forget the gui https://hub.github.com/
beeman
I tend to interact with GitHub mostly through the CLI.

However, having a chronological timeline is pretty useful if you follow other people with similar interests, and it also helps you keep up to date with repos you are interested in.

So for a lot of people, it's a useful addition to the main features.

matt3210
Microsoft has always told people what they want.
beeman
They also brought back the Start menu in Windows after removing it, so there is hope
jmclnx
I no longer care about GitHub anymore. Since I am being forced to use 2FA starting Oct 12, I moved my repositories elsewhere. The methods they are forcing us to use are ones I am unable to use or will not use (in the case of Cell Phones).

For me, this was the straw that "broke the Camels back".

shhsshs
There are several methods of 2FA available that do not require a cell phone. I use KeePassXC for all of my 2FA codes, including GitHub.
politelemon
I saw a blog post a few days ago about this a few days ago, http://ratfactor.com/leaving-github

Which might have been you or someone with a similar sentiment. Anyway Github's 2FA does not require a phone or SMS. I do not use a phone or SMS for Github 2FA myself, or for most sites.

Might be unrelated - I also notice the blog post tries to clarify that it isn't about phone numbers, but they fail to clarify what it's actually about, and aren't interested in other methods... which is confusing.

jmclnx
No, that is not me. If email was allowed I would have stayed with github.

But I need to access it from NetBSD, OpenBSD and Linux. No one option would work for all 3 unless I purchase some hardware or give M/S my Cell Number.

If Microsoft would buy the hardware for me, the maybe my opinion would change :) But I just finished my move, so no looking back.

dewey
> I can't imagine nobody cares about so many developers being unhappy.

I use GitHub every day and I don't have a problem with this change, I also didn't really notice it.

As usual with design changes, for the sub-set of loud people there's a massive chunk of people who don't care about it and might even like it.

officialchicken
The one feature I care about - github outages - has bitten me too many times this year to care about any UI/UX changes. These kinds of changes are infuriating b/c adding a feature (almost) never improve availability. These kinds of UI/UX changes are in the best case a misallocation of engineering resources IMO.