On 9/10/25 1:14 AM, Peter Eisentraut wrote:
> On 06.09.25 00:05, Magnus Hagander wrote:
>> Given the amount of time that has passed since back then, I think it
>> would be fair to discuss the concept again. But a good way to do that
>> is to not start over from the beginning, but to go through the old
>> thread, and then propose a new solution that to at least some level
>> considers the arguments made back then.
>
> Ideally, both the algorithm and the discussion that led to it would be
> documented, like in a footnote or something.
>
> This kind of algorithmic highlighting of certain things over other
> things without further explanation is the kind of thing that people
> constantly complain about in the realm of social media. We have an
> opportunity to be more transparent here.
Well, it was previously documented in the code[1] the original commit
message[2][3], though I can't find the original email thread (likely due
to not including a ref to it in the commit messages).
But what you're getting at is that we could include a small description
for how the algorithm works. I'm OK with that, but I'd prefer to put the
full explanation on the events page and not the homepage, so we can save
home page real estate for things like listing more events.
Jonathan
[1] https://github.com/postgres/pgweb/blob/master/pgweb/core/views.py#L57
[2]
https://github.com/postgres/pgweb/commit/d252bdedf30c235433fed9137aba8db255d5cfb3
[3]
https://github.com/postgres/pgweb/commit/5c3f229e44dbca4ecf3950f50a3bcb3d00987d71
