at://bnewbold.net/sh.tangled.feed.comment/3mx5g4temlt22
Back to Collection
Record JSON
{
"$type": "sh.tangled.feed.comment",
"body": {
"$type": "sh.tangled.markup.markdown",
"original": "I thing STAR is a good addition to DASL. I think CID-removal has pros and cons (redundant integrity can be good), and I don't expect this to completely replace CAR files, but it is great for a lot of use-cases. I like the existence of a clear magic header, and that the metadata header object is separate from other data blocks and is extensible.\n\nThe name is good. I'd lean towards calling the DASL version just \"STAR\" (or \"DASL-STAR\" in long form), and drop the \"-lite\". And I guess rename any previous STAR variants \"legacy STAR\" or something like that.\n\nTo confirm, dholms and I both expect that STAR will be the space repo transport format, and for that to ship in the next couple weeks. I expect to do a basic golang implementation and to collaborate on test vectors (not sure if it will be a generic STAR implementation or spaces-specific).\n\nI hope that we can have some wiggle room for back and forth on spec details in the next couple weeks, though I don't have any asks today. In other words I hope this can stay at \"Hot New Stuff\" maturity through October 2026.",
"text": "I thing STAR is a good addition to DASL. I think CID-removal has pros and cons (redundant integrity can be good), and I don't expect this to completely replace CAR files, but it is great for a lot of use-cases. I like the existence of a clear magic header, and that the metadata header object is separate from other data blocks and is extensible.\n\nThe name is good. I'd lean towards calling the DASL version just \"STAR\" (or \"DASL-STAR\" in long form), and drop the \"-lite\". And I guess rename any previous STAR variants \"legacy STAR\" or something like that.\n\nTo confirm, dholms and I both expect that STAR will be the space repo transport format, and for that to ship in the next couple weeks. I expect to do a basic golang implementation and to collaborate on test vectors (not sure if it will be a generic STAR implementation or spaces-specific).\n\nI hope that we can have some wiggle room for back and forth on spec details in the next couple weeks, though I don't have any asks today. In other words I hope this can stay at \"Hot New Stuff\" maturity through October 2026."
},
"createdAt": "2026-10-05T20:14:59+03:00",
"subject": {
"cid": "bafyreia33nzg2zsl4n5e7gnjhljbm7ja6meixivwkazt44l7yevm6mujo4",
"uri": "at://did:plc:hdhoaan3xa3jiuq4fg4mefid/sh.tangled.repo.issue/3mwqteb3vkv22"
}
}