Mark talks that can be used audio-only
NightMachinery opened this issue · comments
Sometimes we can’t afford to see the video, so knowing which talks don’t lose value in audio form is very beneficial.
That’s an interesting proposal!
I’d be open for that, but I don’t think I have the bandwidth to mark all of the talks that are already on the list. Would you be willing to do that?
Also, do you have a format in mind already that should be used to mark those talks?
I think a trailing [A] to mark talks looks good and won't screw up the years script (maybe use a different brace kind), and can be parsed out by a machine, maybe:
* [Audio talk](url) [duration] by **presenter** (year) [A]
* [Needs video](url) [duration] by **presenter** (year)- Audio talk [duration] by presenter (year) [A]
- Needs video [duration] by presenter (year)
@hellerve @catb0t I unfortunately don’t have enough time to do this systematically, but I don’t think we need to do this for all the talks at once. We can ask people to mark any talk they watch. I think this format would serve such a crowdsourced data better: [A0/V0], in which the number after A is the number of people who think it works with just the audio, and the number after V shows the people who deem the video essential. Over time, this scheme should mark virtually all the talks. Though opening PRs for such small changes seems a bit daunting ... Can anything be done about that?
I actually prefer the version @catb0t suggests.
The idea by @NightMachinary seems nice, but a little overengineered for me, and would require some tooling for people to make it simple—and I don’t really want to introduce tooling to this repository. Does that make sense?
@NightMachinary
I don't think we need to do this for all the talks at once... [somehow suggest] people to mark any talk they watch
Exactly! -- in the case of the format I suggested, I'd watch / listen to a talk and afterward make a PR (or edit my own audio branch so that I can make one big PR with a build-up of [A]'s I've added)
[An/Vn]where n is the number of people voting for audio-only or needs-video
I think this kind of data collection would work well on a website or something with interactive voting interfaces that don't require committing and making a PR just to increment a number by 1, git/GitHub is not exactly ideal though (especially since people's forks become out-of-date just for incrementing a number, not a real change)
I don't think crowdsourcing the data is necessary for such a small piece of metadata anyway -- someone submits a PR adding some [A]s and maybe it would be nice for them to include a brief general explanation of why these talks are audio only, the generic "the slides are not really key explanatory visuals" or whatever.
If someone else listens to it / watches it and strongly thinks the visuals are important, they might be bothered to make a PR / issue to change it back.
Hopefully this doesn't lead to Wikipedia edit warring, but a note at the top of the README like
[A] denotes a talk that someone thought could be listened to as audio, without needing the video. This may not reflect your experience with the talk, but if you feel strongly you can make a pull request to change it.
might help people to think of the [A] as just a suggestion.
I think those suggestions all sound pretty reasonable to me!
I don’t think people usually feel super strongly about the metadata in this repository, so I don’t anticipate an "edit warring" situation; I might be mistaken, though, and I think it’s great that you’re thinking about that!
I’m also all for the suggested disclaimer at the top.
I'm finishing up my audio branch's first commit just now :)
Where does the [V] get used? On every talk, or just ones that really need video? If we add one of these [X] tags to every talk, the file is going to get quite ugly, much more than scattering [A]s...