Hacker News Reader

Comment by JoshGlazebrook | Hacker News Reader

Say what you want about the ethics of piracy, but this truly is a sad day. No where else was there such a vast collection of music nor any other community as passionate, knowledgeable, and collective about all things music and audio as what.cd. Not to mention all of the friendships made and lost on the site, it truly was unique.

There are literally versions of albums and other musical releases that you cannot buy or find anywhere else that are now lost to the sands of time.

View on HN

I've seen commenters on reddit calling it the 'burning down of the library of Alexandria of digital music'.

It's true; the members of sites like what and waffles were dedicated to quality. People would go to great lengths to ensure quality was upheld: from keeping detailed logs of the ripping and transcoding processes, to generate spectrograms to weed out sub-par webrips and fake encodes.

It's really a shame...

It's amazing what humans can accomplish if they're allowed to cooperate.

Just look at the Linux kernel.

Well, to be fair, the music never was on what.cd itself. So it wasn't destroyed. So it's still somewhere out there… Somewhere. On somebody's computer. The problem is to find them and to connect with them again.

What made what.cd unique was the way everything was organized with tags and correct information about every release. It's going to take quite a lot of time getting all that right again.

True, no arguing about that. I just wanted to point out that no music is really, really lost yet. Unlike with the Library of Alexandria, if the community wishes so, it is still possible to rebuild it. Even making it more, uh, fire-resistant.

I already have a personal archive that is probably larger than the library of Alexandria. If this kind of stuff matters to you get a hard drive and start your own today. Once it gets large enough, put it on a NAS with RAID5 and a hot spare.

And if you can afford it, you really should use RAID 6.

The amount of RAID 5 arrays I've seen die during the rebuild process (because one of the other drives, with an equal level of wear of the failed drive, failed due to the increased IO) is non-zero.

If you have the drives, of course. Also try to obtain your drives from different retailers if possible. They have bad lots from time to time. Getting the drives off different pallets is cheap insurance.

Or if you care about the data, make a proper backup, don't fall for the fallacy that RAID is backup.

RAID is for data availability, not backup.

(if still downvoted/gray) this man speaks truth

Agreed, two is one and one is none. I say some sort of RAID as I have read it might be somewhat preventive of bit rot. I hope for my archive to last generations, unfortunately it does not appear that any current storage technology will permit this yet. Write once media like CD's and DVD's and Blu-rays apparently won't last the decade from what I've heard.

You seem to be serious about this, but rather vague on the details, so I wanted to reply, just in case. If you don't already know about ZFS, please look into it. A properly provisioned server with ECC RAM and a ZFS RAID6 will provide the best possible defense for your archive in the years (or decades?) to come. Current filesystems don't come close.

If this is all new to you, look at FreeNAS for a "turnkey" solution. As well as being a solid piece of software, their guides can help you configure the necessary hardware.

I started a long time ago. Still don't have many awesome stuff that was on what.cd. And there's much more I don't even know about yet, which probably also was on what.cd already. To continue using our metaphor, a vast personal book collection is still less than a real public library. It's still a lot about sharing. I'm proud of my collection, but in terms of significance it's not comparable to that of entire what.cd, not to say the Library of Alexandria.

That said, we still should think about fire-resistant public libraries, not just personal collections.

>RAID5

Not good enough. RAID6 is the absolute minimum, as with current disk sizes it's too easy to lose the whole array due to an additional disk failure while one disk is being replaced.

How is that different from last.fm ?

I cannot comment about the content, but for music metadata there is nothing as good a http://musicbrainz.org .

Incredible amount of data, versioned and CC0 licensed, great and diverse community of music lovers, strong ties with archive.org.

I've been a MB fan for a long time, recently found these scritps, https://github.com/murdos/musicbrainz-userscripts, to help import some other albums they were missing from https://www.discogs.com

You might also want to look at beets (http://beets.io/)

I've just got started with beets. It takes a lot of work to organize a music landfill that has accumulated over 15 years, but it may be the best tool for the job - even if the job will consume man-days.

MusicBrainz Picard never cut it for me.

That's simply not true. musicbrainz tagging is quite poor, to be honest. Properly moderated music torrents (not only what.cd) in general have much better and more relevant metadata for each release. Unfortunately, it's not that available and usually requires a human to be interpreted.

Discogs is vaster, but still quite disorganized. For rather obscure music you still don't have anything better than the metadata that comes with the torrent (and is required on all quality music torrent trackers).

You have missed last.fm probably.

last.fm brought me to MusicBrainz. In 2007. ;)

How?

I'm pretty sure I've read exact same comments back when oink shut down. Very sad indeed.

Reply on news.ycombinator.com