Hacker News Reader

Hi, author here.

> Bit confused on one point -- so SoundCloud revised their ToS to deny downloads, but they kept the download URL in the API response? Amazing that after two different illegal take-downs it just kept getting back on top of the listings.

I think they eat their own dogfood here. As far as I can tell, they have no special privileges for their owned and operated clients. If you for example look into the network panel when using the web client, you'll see that they directly access `api.soundcloud.com`.

> Random Android question: When the app was taken down, I guess you lose app storing listing and IAP. But existing users would still be able to run it, and the developer would still get ad impressions and ad revenue?

Yes indeed. I'm surprised how much of a long tail the app has. Even though it's now been almost two months since I've taken down the app and SoundCloud has also revoked the API key, I'm still getting a sizeable amount of ad impressions in a completely dysfunctional app. :/

> I don't quite understand SoundCloud's rationale... Why is it OK for me to do an HTTP GET from one user agent but not another? How is that something they have legal standing to dictate?

It's their platform, they can do whatever they want. As operator of a platform like this, it's natural wanting to control the experience. I think it's remarkable how open SoundCloud remained over the years whereby most other platforms leverage their APIs as a way to find initial adoption and then quickly close them down as soon as they reached a certain point.

> Hartig seems perfectly happy to let his obviously well liked software be stomped into oblivion? SoundCloud could perfectly well provide the equivalent functionality, but instead they snuff out an honest developer?

I'm certainly not happy about this, quite the contrary. But I see and understand their rationale, learned a lot along the way and SoundCloud has always treated me fair.

> How does SoundCloud get Google to nuke an app for properly using this interface?

I'm quite interested in the exact process myself. I presume that this all works under the ominous "copyright" umbrella.

>> Bit confused on one point -- so SoundCloud revised their ToS to deny downloads, but they kept the download URL in the API response? Amazing that after two different illegal take-downs it just kept getting back on top of the listings.

> I think they eat their own dogfood here. As far as I can tell, they have no special privileges for their owned and operated clients. If you for example look into the network panel when using the web client, you'll see that they directly access `api.soundcloud.com`.

But it is strange that they in their TOS for the website state:

"You must not copy, rip or capture, or attempt to copy, rip or capture, any audio Content from the Platform or any part of the Platform, other than by means of download in circumstances where the relevant Uploader has elected to permit downloads of the relevant item of Content."

In other words, they seem to allow users to download content that is available for download. So if I write my own web browser, I'm free to download downloadable content -- but not if I write an app for Soundcloud...

Or perhaps, it's this clause:

"(x) You must not rent, sell or lease access to the Platform, or any Content on the Platform, although this shall not prevent you from including links from Your Content to any legitimate online download store from where any item of Your Content may be purchased."

It's unclear to me how Soundcloud cannot get Firefox for Android banned under this clause as well... (well that would be insane, but you can ask for the desktop version of the site, and download from there?).

Note that this is the user TOS[1], not the API TOS[2] -- but that line is a little more fuzzy than they would like it to be IMNHO.

[1] https://soundcloud.com/terms-of-use

[2] https://developers.soundcloud.com/docs/api/terms-of-use

[edit: entirely forgot to thank the author for being open about this, and publishing the app for everyone to access. On another note: did you already provide source to some of the users on request? Or did you just recently add cert pinning?]

So how much $ did you make from this?

>It's their platform, they can do whatever they want.

Surely that's proper legal standing, next to the user can cram it, arguably. There are basic rights and to forbid the use of any can't be broken by contract. IMHO access to arbitrary data on the net should be included, because otherwise it's restricting the market, obviously to artificially drive up the prices. That's against the free market principle, but maybe it's necassary to get the userbase to generate revenue. Soundcloud is not profitable right now. I'd contest that it is necessary, though there is responsibility on the user not to abuse Soundcloud's offering. I'd argue that download rates would be enough to meassure popularity or for whatever they need the data, which is listening behaviour. I'm not sure the law intends to protect Big Data collection on privacy grounds, if data is cleaned up. used for all but what comes down to accounting.

I used soundcloud-dl.py in the past, because the page didnt work for me. Don't know if it still works. I found it very convenient since I'm not interested in the networking features either. I don't understand why people depublish pieces and very much like to keep them. Copyright-Law grands the right to rewoke copyright or perhaps doesn't grant it to the public in case of transmissions to begin with. I'd argue though that with the event of computers and the net, where you select the transmission yourself, the lines are blured. If program-code needs to be specifically excempt from the copyright, in order to legalize loading into RAM any data you bought, then receiving a transmission is a copy as well. I don't know if there are exceptions to the exception, or the transmissions - it's not a persistent copy. Still, you aquire it and the ability to record it is a functional requirement, it's not like a performance that could be given without recording it.

They aren't snuffing out any developers, they just don't want people to download their content directly to their phones per their TOS.

I'm sure their rationale is reasonable and probably has to do with some overall terms probably with major content providers or studios.

So go to the link on your phone's web browser and tap the download button?

The only answer is lawyer garbage.

Yes, you only have to go around the mobile version.

It sounds like they are claiming their ToS can dictate how you can use the data they are putting in your computer's clipboard. They then used that claim to remove the equivalent of a 'wget' program from the app store.

I don't like what SoundCloud has done here. I'm pretty sure it was line noise for Google, but I don't like that they went along with it either.

Note these are published files that are generally available for download. The download is an HTTP GET on a public URL, which they publish freely and directly on their own site. It's in no way "ripping" the file from the site or defeating any DRM.

> It sounds like they are claiming their ToS can dictate how you can use the data they are putting in your computer's clipboard.

Yup. When you are using their api to power your app you have to abide by their TOS.

>It sounds like they are claiming their ToS can dictate how you can use the data they are putting in your computer's clipboard.

They key word here is "they". They "are putting" that data. If you put the data yourself, with your own API and your own service, you can do whatever you want.

Because lawyers don't get what you just explained.

Or they get it just fine, since top lawyers of the kind hired by SoundCloud are as smart or even smarter than most programmers, and some even have technical training, but they have their reasons.

Like, they know that the fact that you can "get around" something doesn't matter as much as there being a slight inconvenience.

If, for example, at some point the official SoundCloud app allows direct download, people will use that, and not the "get around" method -- whereas if they allowed third party clients to have that feature it wouldn't be so clear what the users would choose.