I did cut another RC, but subsequently, another PR, 2829, came in to the mix. This has reduced the time for obtaining quotes on the production network, from ~4.5 minutes to ~30 seconds. Like I said, this was tested directly on the production network, so that’s good.
So now the hope would be we will see quicker and significantly more reliable uploads. Maybe not fast yet, because we still need some optimisations for outdated peer information, but I think something that is reliable would be a good first step.
This PR needs to go through a testing run tonight, then I will cut another RC tomorrow and put that through a set of standard tests. So we should hopefully be looking at Wednesday for a new release.
Yeah, I think you are on a good track and please don’t feel any pressure to deviate from it. I just tagged you and Mark because you seem to have the best understanding on this kind of subject.
What I am after here is to have a truly public space, where data can be published without interference from anyone else than the one wishing to publish it. Something that is similar to Autonomi itself: Once set up, no one is in charge.
I think that public data is not really removable?
But making in difficult to find with flooding would be like 100% certain to happen with anything that’s even a bit objectionable to enough people. But once up, it could be linked to forever from different similar public spaces, and flooding them all is at least going to cost something. But yeah, it would be a battle ground for sure.
@happybeing How hard would it be to get a text file, or csv, with the datamap network address from your system? I ask because the download-exercise.bash script could have those addresses added to its list of files to download to exercise the network. And have the nodes actually do some work and use some of the bandwidth people are “keeping aside” for node use.
The address of files, archives and history are all output to the terminal as part of the unused process so they can be captured there.
You can also get details retrospectively from the network using the dweb inspect-??? subcommands by passing them the address of an archive or history too, if the terminal output is not available from the upload.
Also, if you don’t have the address of the history, the uploader can get this information with just the name of the directory they published.
When dweb is published take a look at the help for the inspect commands and let me know if you have questions.
