Keeping free culture alive at the source
Reading David Revoy's post on trying to keep around high resolution PNG files has made me think again about the various difficulties in making a "complete and corresponding source" type free culture future possible. These seem like issues to me:
- Free software's sources are usually comparatively slim as plaintext. Free culture's sources are usually large and heavy binaries. So, storage space, on your local computer and especially for long term archiving, seems to be an issue. We could say "just use archive.org!" though I'm not sure that's satisfactory.
- Free culture does not usually have "build reproducibility", even for the assets that could be reproduced. Certain parts of the process could be makefile-reproducible, for example rendering XCF->PNG, or rendering a .blend file, etc. (To some extent this is accomplished in open movie projects like Tube, but not totally.) This is something that could be tackled, but is cultural.
- Many sources are hand-altered after being rendered, ie "post processing". This cannot be mechanically reproduced. Our best bet here is to keep the original source around, and the file used to post-process around.
- As for version control, it's possible, but nothing is ideal, to the point where @Bassam Kurdali told me that a number of frustrated artists have expressed their desire to create their own version control system just because of how underserved they feel by contemporary VCS'es. This is true especially for git, most of the command of which generally expected that you have the full history around (and even if you pulled down just the recent head, future commands may pull the rest down anyway). git-annex may help, but does not really provide generally a workflow for large files you wish to keep around. Subversion is still the most popular option for hosting large free culture project sources.
- There is little to no culture of sharing source in non-software contexts. Who aside from the Blender Institute does it on anything major?
What can we do to improve on this?
(Even saying all this, I still have an angry task on my orgmode agenda to get the MediaGoblin campaign video sources uploaded to archive.org... blah!)
der.hans, AJ Jordan, Jason Self, Claes Wallin (韋嘉誠) and 5 others likes this.
AJ Jordan, Jason Self, Blaise Alleyne, Claes Wallin (韋嘉誠) and 2 others shared this.
@Blaise Alleyne It seems weird when we try to use software terminology to talk about music. I fear substantive meaning is lost, because everything turns into a strained metaphor and no solid conclusions can be reached.
I imagine that relatively little of the intermediate information that could be saved and shared is valuable. Most of it should be discarded, because it would only confuse and distract us.
As a thought experiment, think about how much stuff you don't share when you're coding. Tons of stuff.
Think about how much stuff you don't share when you're releasing a binary, even as freeware. Think how much even free software didn't share back when we relied solely on tarball releases. Why did we go to publishing the complete repos?
I don't think it's strained at all to talk about preferred forms of modification for things other than software. Most of most git repos are more or less discarded or ignored. The fact that the full history is there doesn't confuse or distract anyone. And then that day comes when someone needs to go back and trace why something was done the way it was done.
Why was there a glitch in that tree rendering in that hypothetical movie I saw? Oh, actually it was because they were working around some limitation in the software they were using, and that limitation no longer exists because the movie was made ten years ago. And then I can fix it and rerender that part of the movie. I don't get what's difficult about seeing the value of this.
Longevity of cultural artifacts is very much increased if you have the precursor forms and the instructions for turning those into the finished form.
I have the sheets for Queen's Greatest Hits and Queen's Greatest Hits II. One of the greatest gifts I ever received. They have tremendous value to me, in addition to the value of being able to hear the finished rendering in the speakers. Just seeing that wow, for this song, John Deacon actually tunes one of his strings differently from the usual EADG, that teaches you something.
And seeing "Synth 2: Multiple laser beams" in the Flash Gordon sheets is hilarious.
Claes Wallin (韋嘉誠) at 2015-11-11T18:31:48Z
Daniel Koć, der.hans, Jason Self, lfam and 6 others likes this.
Yeah, I think it's a strained metaphor at some points, but I don't think the intermediate information is all like the stuff you don't share when you're coding. I'd love to see the process, love to see alternative imagined possibilities for a piece of music/art, love to hear recorded sound clips before they were compiled/edited/effected/processed/mixed, etc.
Sure, there's some information that many people wouldn't use. But tons of valuable intermediate steps. And it wouldn't have to be any more confusing that a git log is, right? Don't have to look backwards if you don't want to. Or you don't have to use all the source files if you don't want to.
But there's still lots more information that would be the "source."
Blaise Alleyne at 2015-11-13T23:29:22Z
Claes Wallin (韋嘉誠), Christopher Allan Webber likes this.
sorry I'm late to the party! for 'source' files you don't (often) want jpg, though it can work - we often like to use lossless compression and even higher bitdepth for image sources, increasing storage requirements even further! though it depends a lot on the nature of the particular project, I can see 'optimizing' projects for ease of sharing / hosting rather than quality as such. Other source files include blend (3d models/animation), raw data from cameras, svg, etc.
Git-LFS might be part of the solution, but it's not the solution on it's own. There are some specific problems related to work on animation/film projects that might be good to target before looking for a specific technology to use - another lil' issue is that artists tend to be much less tolerant of technical/command based interfaces, and we want to have a lower barrier of entry for them.
Bassam Kurdali at 2016-01-03T07:54:45Z
Claes Wallin (韋嘉誠), AJ Jordan likes this.