April 13, 2026: Update on the progress of our Flash preservation work: We have completed preserving 178 Flash works with Ruffle. Left to finish are 502 more. We will finish all but the 200 works held in “The Museum of the Essential and Beyond That Collection” in time for Wednesday’s anniversary celebration. This means James (featured here) and Ruth will complete 302 works @ 5 minutes/record, divided by 60 min/hr = 25.16 hours for 2 people over 2 days = about 6.5 hours per person per day).
680 Versions of Flash works held in The NEXT
179 Versions of works already preserved with Ruffle
460 Unique copies of works preserved with Conifer
225 Unique copies moved from Conifer to Ruffle, ready to be added to The NEXT
235 Unique copies undergoing conversion from Conifer to Ruffle
We have grown accustomed to media formats and software used for born-digital art, literature, and games becoming obsolete, but when preservation methods and systems, themselves, become outmoded, it is difficult not to see it as highly ironic. Such is the case of Conifer, the Web archiving system that the lab had been using alongside Ruffle for keeping Flash works accessible to the public.
The announcement from Rhizome.org about the “twilight[ing]” of Conifer came on December 29, 2025. Since that date, we have busily been testing the 282 Flash works that we had preserved with Conifer to see if they could function effectively with Ruffle.
Some back story: The lab had been using both Ruffle and Conifer for preserving Flash works held at The NEXT. Our method was simple: We would test a work with Ruffle and compare it to the output on a legacy computer that still supported Flash or on a PC running Pale Moon. If the Ruffled version was not close enough to the original version, we would test the work with Conifer. Over time, we found that Conifer handled the more complex works better (or at all). We actually preferred it over Ruffle since it did not require us to intervene in the work’s code by adding the Ruffle code to it, which was not needed with Conifer. Our preference for Conifer led us to create our own instance of Conifer on a AWS server so that we could manage our own preservation activities with Conifer rather than lease space from Rhizome.org. At the same time, we bore a large cost in paying for the additional server space to host Conifer. We begrudgingly agreed that the plan one day would be to shift our process completely to Ruffle once Ruffle could grow to support all of our works.
The loss of Conifer has, indeed, pushed us to Ruffle entirely sooner than we expected.
Since Rhizome.org’s announcement, James Lesperance, our digital preservationist, has revisited the 282 Flash works that we preserved with Conifer to see if any of them can shift to Ruffle. As of today, 85 works can be. The rest can be partially preserved, need further testing, or not at all. So, less than a third.
As Ruffle continues to develop, we may be able to preserve other Flash works held in The NEXT that are not able to be saved at this juncture. In the meantime, we will add our scholarly note to each one no longer functioning. The truth is, the lab never imagined we were preserving works so that they would last “forever.” We have imagined, instead, that we were part of a relay race and that we are running with the baton for this leg of the event, long enough for the next runner can take the baton from us for another several laps or so.
Special Note: We will have the works that were originally preserved with Conifer moved to Ruffle before April 15th. Below is our Workflow Process:
Ruffling Process (without full testing)
- Determine whether the Conifered work was in the archive or external
- If external, then the work is marked as unavailable on the spreadsheet
- Collect the external URL
- If in the archive, locate where in the archive the original files are sitting
- Download the original files and copy it to a local directory
- Recreate the file structure
- Determine where inside the work Flash is embedded
- Add the Ruffle script to each page that has Flash
- If work has <3 pages, input the script manually
- If work has >3, using AI tools to input the script on each page
- If Ruffle is displaying a result, a sticky note is added to the landing page of the work
- If there is a Flash detection script in the original work, that must be deactivated in order for the embed to work.
- The Flash detection script is sometimes simple and sometimes complex.
- If Ruffle is not displaying a result, then the work is marked unavailable
- If there is a Flash detection script in the original work, that must be deactivated in order for the embed to work.
- Full functionality testing has been on pause in order to get through the list
This process takes ~15 minutes per work. Some can take 5 minutes; some can take 30 minutes. For 225 works, this took ~56.25 hours without full testing or metadata.
In Museum of the Essential, because it was originally preserved as a single .WARC, each work inside has not been gone through individually. If automated, Ruffling the Museum could be as fast as 2-3 hours. One sticky note will be used for the entire Museum. “This unique Flash-based Museum contains x number for works created with Flash. These have all been saved using Ruffle.”
Metadata Process
- Uploading the now-preserved works in the preserved directory in the archive
- Access the work entry in the dashboard
- Update the following metadata:
- Preservation notes and boilerplate 1.0 notice
- Preservation method
- Hosted URL
- Version update to 1.1 (we will leave version 1.0 alone for now)
This process takes ~5 minutes per work. For the 481 Conifer links in The NEXT, this will take ~40 hours.