Hey all!
I’ve been doing some noodling about the “completeness” of Mirlo
from a technical point of view, which I think will be very useful to draw parameters around what it is we’re doing and where we’re going, and what the lines are around the project.
Wanted to share the draft post we’re working on to get y’alls feedback and see what you think and what questions you might have around it.
Thank you! Roadmaps with checklists are always interesting. 
What data are you sharing or planning to share, to artists/labels and/or publicly? Your doc only mentions “Keep track of your sales and income”. Are you considering statistics about how many times tracks are played? Logged in users playing them?
Also, are you considering listing the users who have purchased an album, like Bandcamp does?
And something sorely missing in fair platforms (unless I missed it somewhere) are automatic recommendations based on [something reasonable and basic, no complex algorithms]. Whenever I land in a Mirlo track / album / artist, I listen, the song ends, and… that would be the perfect time to queue up another track, somewhat related. But no, music stops and most of times I leave, because I’m not in the mood of starting a new search.
And no Playlists?
Just curious about your thoughts on these features. I’m not pushing for them, and I’m aware feature creep is a very real problem.
We’ve honestly approached this with a little bit of wariness, we currently log plays but we don’t reveal that information anywhere. The main intent for logging plays is to prevent people from infinitely playing a song (and so we’d pop up something like Basecamp that’s "you’ve played this song X times, you should buy it). But mainly my hesitation here is because a while ago a couple of people expressed that actually they prefer not having these metrics.
You mean publicly? We could!
Yeah, that’s an experience that would be nice. What I think would be nice to have is artists get to recommend other artists on Mirlo, and then you could navigate there. But we’re not an infinite jukebox like Spotify. There are legal reasons for this (if you are an infinite jukebox then your plays aren’t promotional plays anymore, which means that you’re really a radio station, which gets a little hairy and we’re not ready to support (and I’m not sure we’ll ever be ready to support) mainly because we’re just not sure about the ground we’re walking on here. This is the reason that Bandcamp doesn’t do infinite play either.
Likely the closest we’ll get to playlists in Mirlo is our blog posts. We have the infrastructure in place that tracks embedded in a blog post can be played sequentially, but we haven’t implemented the UI for it yet. This is largely for the same reason as shared above. Again, same reason that Bandcamp doesn’t have public playlists.
Thanks, yeah! Questions help us sharpen the edges.
This is a topic with many different opinions. As someone who doesn’t sell music but publishes for free, number of plays is a number I care. It doesn’t obsess me, but it does motivate me somehow. But Mirlo is designed to satisfy music makers who sell music, and in that context I understand that the numbers that really count are sales (and subscriptions/followers).
Yes, just like Bandcamp. It’s a quick visual hint about the popularity of an album, maybe a little extra motivation to buy, and also a way to browse other people’s profiles (if they are browsable) to discover peer and artists.
Mmmaybe, but this may open other distorting cans of worms like artists recommending other artists with an actual motivation to get the reciprocal recommendation, popular artists feeling pressure to recommend less known artists that have recommended them for visibility… We have seen these types of behaviors in many places.
I didn’t know about “the infinite jukebox problem”. Interesting. And sad, but ok. Maybe Mirlo could recommend, say, three artists automatically, based on people’s purchases and subscriptions. Users still need to click the artist to listen more music, but at least they are offered some recommendations.
EDIT: another Bandcamp feature that would help discover music and music makers that Mirlo users might like: a music feed including not only artists they follow but also fans they follow (for their purchases) and also tags they follow (new releases using these tags).
bandcamp has this but successfully avoided follow4follow kind of mentality. not 100%sure how, but maybe it has to do with how this is implemented: after someone purchases your album they are offered three albums from other artists that you recommend.
have you noticed there is no number? it’s only a grid of avatars, and its not even everyone. you have to click to reveal the next page of avatars. I’m always tryin to lean into quality over quantity in the sense that what matters to me as a listener that is looking for new music are reviews - i would love to see bandcamp style reviews under the releases. people actually bothering writing a sentence or two or three about why they like the album they purchased means more than hundred plays. mirlo creating a friction-free and intelligently incentivized way for people to write something that they bought would be what i would really like to see.
Unfortunately cannot see the document since i don’t have a Google account 
(Analyzing myself) I actually haven’t paid close attention to these avatars but I always look at them when I land in an album that sounds interesting, and I like the little variation they provide to pages that otherwise would look more like the same. I also like when from time to time I see a little avatar I recognize. I think they help build a light sense of community and companionship.
Thanks for sharing @simon. I’m in the same boat as Crapahute, maybe a PDF version would help?
Without seeing the document, I had some lightbulbs go off while reading this thread. So at the risk of going slightly off-topic, here’s what clicked in regards to what a “platform” could be. I would like to think it’s relevant in defining a scope for Mirlo or any other similar platform My goal in doing this is to: distribute worker effort and empower artist creation by untangling the various use cases addressed by Mirlo and similar platforms.
Footnote: If I’m not mistaken, Mirlo was designed to be a “self-hostable WordPress.com”—note the .com. I think the subtle shift here is that a website should only serve one artist OR many artists working together in some way (e.g. label, collective)