Export to Roam JSON (and EDN) does not work

Because Logseq is too slow with my large graph, I reluctantly want to go back to Roam for a while–until Logseq can handle larger graphs. However, the JSON (as well Roam JSON, and EDN) files that Logseq exports are not importing into Roam. The JSON import dies with a “block too large” message, and the EDN restore ends with a not a valid EDN file message (see screenshots). Does anyone know how to get around this? Thanks!

Update: I divided my graph into different folders, with no more than 500 markdown files in each, and Roam managed to upload the Json files created from each folder. There’s a bit of weirdness in the formatting, but everything seems intact. Still interested in anyone’s suggestions about best practices before I put more work into my graph in Roam, leaving me stuck with the graph as is. And I’ll check in on Logseq in a few months to see if performance has improved. Thanks.

How big is your graph? I want to know when to expect a bottleneck.

Curious to know how this works out for you. Looking to transition out of Logseq till the performance issues are sorted out. It’s not ready for mainstream yet :frowning:

It’s not that gigantic, but it was big enough to slow Logseq to a crawl on my computer. Around 1,500 pages. So far going smoothly in Roam, and I’ll try Logseq again in a while. I appreciate the work that folks are doing on it, but I can’t work in it if it’s too slow.

Damn. I am now trying to work in Roam and I see that my block embeds and references didn’t come through in the import. So, I’m back to square one.

Any luck? If so please share how you did it!

Unfortunately not. Have you had the same problem? I’ve been communicating with some folks at Roam about it, but no solution so far.

Just an FYI, I never fixed this Roam problem, but with Logseq version 0.6.1 (Windows) my Logseq sluggishness problems seem solved. It’s working great at this point.