Congratulations, very cool. BUT... I can't help but feel this is an... impolite?... use of archive.org?
I see archive.org as being a repository, not a streaming service. I would rather they spent their money on content ingestion and resilient content storage, not on having to scale up their outbound bandwidth capacity.
What would polite look like? Download the content off archive.org, then stream it from your own server.
Now, I know the IA are perfectly happy to let you do this, and of course part of their mandate is to make content accessible, which necessitates having low-friction ways to download it. But still, I would hate to see IA become a CDN-first, archive-second.
Maybe a compromise would be to suggest that users make a donation?:
I agree, I don't think this is a impolite usage of Internet Archive's resources. I think they would love it. Though I also agree that putting a donation link to the source is also a good idea.
A couple of years ago, I created a simple “Donate to the Internet Archive” logo for a site I made using IA resources. It links to the IA’s donation page. Anyone is welcome to use the logo on their own sites:
Well, you could certainly just download from archive.org for the first-time usage of any particular song and then cache them from there. Like being your own CDN.
This is a good idea. I did this with about 38,000 books in the public domain and hosted and indexed the text files on the page so I could do quick full text searches of all of them: https://www.locserendipity.com/Google.html. Interestingly, it is not nearly as good as a RegEx search of a directory of the same content, revealing some inconsistencies in their search algorithm (at least for custom search engines).
donations are a fine thing, so is discovery, anything that helps to surface quality will long term perhaps help the archive prove its worth and ensure a positive attitude towards their endeavours.
Blindboy Boatclub, Irish artist formerly of The Rubberbandits, credits Joe Frank as an influence on his podcast style, tracing it back to Samuel Beckett as a potential influence on Joe Frank, who taught Ira Glass of This American Life, and may also have been inspiration for Chris Morris of Blue Jam and Brass Eye, programs that Blindboy experienced as a youth.
(Shoutout to spaced repetition with Anki for helping me remember all this. It was one of my cloze cards to help me get back into using Anki and I didn’t expect to actually reference it)
I will take a look at that artist too. Anki is a great tool. I have tried to do something like that with languages, like Japanese: https://www.locserendipity.com/jp/Japanese.html and am working on voice simulators based on phonemes not dependent on APIs. It is not that easy to do. This English one based on my voice doesn't sound that great: https://www.locserendipity.com/sim/voice.html
As a compromise to this, I added a features where you can add a search string to the URL and the player will only play songs that that search string in the URL. For example, this only plays Jazz: https://locserendipity.com/PushPlay.html?q=jazz-
Some people really prefer novelty. Try listening to say WFMU for a while, you might hate it but it’s likely to hold your attention in ways personalized streaming doesn’t.
This. The like button all too often acts like the "echo chamber": you're going to go around in circles. Writing a good recommendation system is very hard, and it gets harder as the number of users grows (because primitive algorithms will get more and more certain of their predictions, and diverge less and less from the favorites).
You realize how much you're a product of your time, when you hear a piece of music being played[1] and your first thought is that it must be satire. I'm so used to music from that era being satirized (e.g. in the Fallout games) that my mind had a hard to accepting that something like that could be sung unironically.
Turns out, my spider-sense wasn't off. Here is an interesting comment from YT by "stephen wallin":
As so many query whether the Lyrics were understood in the 1920's I got a friend to check BBC archive and the record was never broadcast on the radio, indicting deep problems, and a note still exists that it was banned for sexual innuendo. So the double entendre was fully noticed as soon as the disk came out. It however sold well!! There was a market for such disks, many dance bands did them, notably Bill Cotton.
Harry Roy was noted as having a wicked sense of fun, he loved playing prank versions of songs. You have to appreciate that the English audiences loved such lyrics and jokes, Max Miller made a whole career on such comic material.
So far I didn't hit skip once, I'm surprised by the overall quality! I've had nice Cuban music, nice classical music, songs from various countries (Sweden, etc), some jazz and rock 'n roll, etc. And sound quality is generally pretty good or even excellent.
Hmm...though I'm wondering were there always greater than 365 full days of audio recorded every year? I'm not sure that is true for the late 1920's-1930's, though by today (and maybe by the 1960's or so) that has become true.
So I guess it will eventually become infinite with the metric of being able to listen to them all, though at the moment I'm not sure it is quite infinite yet.
There must have been a moment when the total recorded output of the world surpassed the total time a person, a city, a country (have we reached that point yet?) could listen or consume all of them. I suspect we will reach a point where songs are played once only, never to be heard again.
Combine public domain growth with Sturgeon’s Law, and you need 10x lifetimes of material in order to have a full lifetime’s worth of good material. Increase that to 100x to have a full lifetime’s worth of great material.
Working against this is the “sliding window” effect of older materials becoming less relevant to modern audiences, and hence “less good”.
Counter this with longer cultural continuity that turns the sliding window into a stretching window.
Is that even possible though? Copyright expires after 100+ years these days. We'll probably be dead before we see new works enter the public domain. Assuming the monopolists don't extend it even more.
I love this! I previously used Parker Higgens' 78 sampler twitter bot to discover new entries in the collection, but I really like your autoplay/jukebox feature.
Thank you so much! Great suggestion--will see about adding that feature in coming days. I hope you enjoy the music! You might also like the reduced static version (a little different interface--it is searchable): https://www.locserendipity.com/Listen.html
That is my fault--nonintutive interface. With this one, you need to select the songs you like and they are added to the list in the box. When you press play, that list plays. The idea was to create some customization without a login. Please see if that works for you.
Next up an infinite infinite record playlist of out-of-copyright works that dynamically remixes mashups of said works in the catalog with an auto populating infinite scroll page of song titles…
It works a bit differently. You press "Random" or do a search, then click the songs that appeal to you. They are added to the (text) playlist. The idea is that you can change and save the playlist--no login required. You can even download the whole thing and use it offline, completely privately. It still works. The only bug is that you cannot skip.
I’m not getting a play button on the low-static version. I did manually copy one of the mp3 URLs [1] and it sounds great though. I’m on iOS (Safari on iPad) if that makes a difference.
Hey thanks a lot! It's soft launched. I'd love to hear what you think. If you try iOS, the TestFlight build [0] is most current. I deploy the web app all the time (just deployed, actually, so if you reload, maybe it'll be a bit different :). Android soon. I'm hoping to do a Show HN in a few weeks.
I see archive.org as being a repository, not a streaming service. I would rather they spent their money on content ingestion and resilient content storage, not on having to scale up their outbound bandwidth capacity.
What would polite look like? Download the content off archive.org, then stream it from your own server.
Now, I know the IA are perfectly happy to let you do this, and of course part of their mandate is to make content accessible, which necessitates having low-friction ways to download it. But still, I would hate to see IA become a CDN-first, archive-second.
Maybe a compromise would be to suggest that users make a donation?:
https://archive.org/donate/