Topic: Emulation Location: Great Hall B Time: 1-2PM Participants: Julia Kim, Peter Chan, Hillel Arnold, Euan Cochrane, Rebecca Fairmow, Laurie Duke, Vicky Steeves, Karl Blumenthal
Sheepshaver documentation is lacking. This is another project that needs to happen in order for all of us to exploit it better. Another grant is probably needed. Should a grant be pursued for something as discrete as an Emulation Conference that brings together vendors, researchers, technicians, etc.? Just bringing the vendors together could be its own useful separate track. Regardless, we need to step up the training and education (especially among funding agencies! They will only fund us insofar as they understand us).
What is the most useful thing is that we could ask for from a grant? Answer: A turnkey systemlike BitCurator. What other projects (other than the ones discussants bring to the table) are out there?: CMU Olive Project. CMU's staff working on different emulated environments. The advantage there is the distribution and depth of the computer science capability!
What about the concerns regarding proprietary software running from the cloud? Is there a Spotify-like subscription service model would help? What about hiring a team of lawyers to find us the loopholes we'd need! CMU might have already explored this...and CMU does in fact take down a lot of software upon request. Internet Archive presents the alternative model as someone who puts a lot online and waits for someone to stop them.
What is a tolerable risk for different institutions? Conversation turns to possible partnerships. Non-profits, especially small ones, are great partners for the above because they take risks--they're not such attractive lawauit targets and don't have the complex bureaucracies that slow these processes down. E's work with Microsoft to enable access to their software on-site and remotely for Yale staff. Are there others at places like Apple we can reach out to? Can we build a group to push this forward?
While we discussed emulating older environments, one discussant pushed the need to begin better practices now. Can the repositories that store all of our digital files get ahead of the game? Instead of collecting just the files, collect the whole environment, so you have some way to run them when the time comes. This is not difficult and will save people in 5-10 years lot's of work. It will make more systematic emulated access much more possible. Keep moving images to new machines, because they will be much easier to work with even virtually. And virtual machines are the only sustainable long-term solution we now of today. Certainly you can't rely on running any of the physical machines for any meaningful length of time. (E has documented to process of moving a database from Windows 2000 to VMware in an OPF blog post).
Another problem: Decision-makers can still be skeptical of emulation. They think migration, for example, is cheaper and easier. Our challenge remains to articulate the benefits to them. Recommend citing the Theresa Duncan CD-ROMs (and bwFLA) to make that case. Jason Scott's work on the Internet Arcade is another accessible example.
What cases do we have of researchers asking for emulations of original forms/context? (We can cite these in advocacy) J is looking into this a lot herself, and is noticing how difficult it is for researchers to use older systems. Euan recommends migrating everything early and often so you at least have a back-up.