Bulk Downloading Option?

[zooker]

Ah, I'm sorry if I sounded a bit hostile.
(and no, I'm not the Korean site mirror guy, whomever he/she may be, i said i had access to various proxies)


I have been frequenting this site for a while- wondering when it will be taken down through DMCA or through a for-profit-venture..
Now the NYTimes article raises the risk considerably I believe.
The very fact that it was not first in results for "sheet music" on google was a blessing.


Anyhow, yes, I'm not interested in audio or video. [also, I can't imagine a non-ad revenue, free site being able to support their necessary bandwidth costs - but maybe it can, who knows - anyway - i obviously don't]


I would hope people move quickly to make this happen.
Legal fee's are expensive. Deep pockets can make uncomfortable things occur.


I will try late next week to set up spidering your site. It will run slow for a couple days to make sure things are working correctly. And then I can ramp things up at whatever times you'd like.

Best wishes to all. I'm not an adversary. I merely am taken off-gaurd by the nytimes article.
While I'm not a google-phile, I wish this were one time where google and their lawyers were backers.

[KGill]

I have been frequenting this site for a while- wondering when it will be taken down through DMCA or through a for-profit-venture..
Now the NYTimes article raises the risk considerably I believe.

The same article says pretty clearly that everyone's favorite Viennese publisher already tried to back in 2007, and succeeded only temporarily. Now, the company that owns the site is registered as an LLC, and is much more secure legally than it was three-and-a-half years ago. It's not going anywhere, don't worry.

[also, I can't imagine a non-ad revenue, free site being able to support their necessary bandwidth costs - but maybe it can, who knows - anyway - i obviously don't]

The site is paid for out of pocket, helped by donations, a deal with Amazon, the new merchandise, etc. Again, I don't see any reason to worry about that at this point.

I will try late next week to set up spidering your site. It will run slow for a couple days to make sure things are working correctly. And then I can ramp things up at whatever times you'd like.

I don't think it is advisable to go ahead with this - while I'm sure there are scripts in place to stop it anyway, 'ramp[ing] things up' sounds dangerously close to site ripping, which in turn can get dangerously close to DoS. If this is going to be implemented, it should be only after input from our server admin (Feldmahler) is heard.

[zooker]

Hey,

I feel like you are telling me to not worry.
But I can assure you I am worried. In some ways I am even more worried because you seem unworried.


I will make sure I do not dos you.
If your system administrator has a better idea than an outside entity spidering your site, please do it.

[imslp]

First, there are already several mirrors outside of IMSLP's control that contain the entire IMSLP archive.

Second, bulk downloading will likely put way too much traffic on the servers; the archive itself is approx. 400GB. Furthermore, the copyright controls are quite complex, and one misstep or bug could mean copyright infringement on a large scale, and especially in the case of torrents it would not be something we can really fix.

So it would seem like any bulk downloading option would only increase the possibility of suits, especially considering that there are already mirrors outside of IMSLP (and so the benefit is not as strong).

Also, if you start ripping IMSLP, I will have to ban you from the site altogether if your ripping degrades the performance of the site for other users.

[Melodia]

Well if it's a torrent, it doesn't have to be related to the servers at all (and torrents by nature delegate bandwidth across anyone seeding and leeching anyway)

[daphnis]

I agree with Feldmahler here. This opens us up to huge legal complaints and I don't see any feasible way to observe the various laws in place and the licensed pieces.

[klawrence]

Hi, I was wondering, could not a few guys unofficially get together seed in parts? Even a cheap webhost (something like $50/month can get you unlimited bandwidth), to separate it from the IMLSP entity? I don't understand the liability issue if dealing with public domain scores but certainly respect your concern about it.

Also, I wonder if you could let me know what would be considered too fast for site ripping in terms of MB/day. Would a few hundred MB/day be OK? Count me as one who is very worried about the state of the economy and how that might affect most aspects of our lives. Having a local archive and one's personal instrument would be a thing to cherish in difficult times.

In any case, thank you so much for providing this service.

[klawrence]

I had one other thought -- would it perhaps be possible to simply (for a talented coder of course) add an bulk downloading utility in the web interface, with bandwidth limit, on a per-composer scheme, that would be able to check regional/public domain tags and filter out the files that are public domain before creating a zip file for that downloader? Or perhaps a script that you could run on your end to create .zips for each composer and regional tag (that way you don't have to create an archive for every single user request). Then you could have a link at the top of each composer page for each region to download all files for the appropriate legal situation.

I know it is work, but it is certainly something I would donate to encourage.

[daphnis]

The concern is that we have files that are public domain in all regions, or in certain regions, and many files are licensed under a creative commons license. It would be impossible or highly problematic to make bulk files available based on the various laws in place that we observe, which then the question becomes: to whom will the rips be available?

[steltz]

Even the countries with the lowest copyright terms have some items they can't download from IMSLP -- this makes it highly likely that a bulk download would contain something that would be under copyright in the country of the bulk downloader. This leaves IMSLP open to the accusation that we encourage breaking the law. So, no bulk downloading, I'm afraid.

[haydenmuhl]

Perhaps explicit collections could be made available as bulk downloads via bittorrent?

These collections could be of scores likely in the public domain everywhere. For example, I would imagine there are scores for Beethoven's symphonies that are likely to be public domain everywhere. I would guess there are other similarly important collections (Mozart symphonies, Schubert songs, Donizetti operas, etc.) that might have a similar status.

[jujimufu]

I don't think it is an overall bad idea - aside from the problems of copyright and ripping by other websites.

It would be particularly handy if it would be possible (with a php script, perhaps?) to select which scores you want to download, and you could also bulk-select according to copyright (e.g. select options would include "All", "None", "Creative Commons", "Public domain EU" etc etc).

That would be extremely useful in orchestral scores with many parts, for someone who wants to download all parts of an orchestral work, or if someone wanted to download only the violin part of ten mozart string quartets, to compare them and see which one's easier or whatever. I would imagine a captcha would prevent most automatic downloads, and if it's a human doing the ripping then I don't see any problem with that as they are responsible for their actions - I wouldn't imagine why anyone on IMSLP would have any problems with someone downloading all the pubic domain scores on IMSLP and providing them on their website

Of course I do not know at all how works/files are organised on IMSLP (or any wiki site) in terms of what data is attached to them, so I don't know how easy it would be to do this, but I can see how it would be helpful.