keep dat address when debugging via deleting the .dat folder

[jcolomb]

Hi,
When dat share is not working anymore, the docs tell to erase the .dat folder and run dat share again.

It works, but then the folder gets a different dat address. This is problematic, because

1. I have told people the address of the content
2. the content itself has internal links who are broken when the dat address changes

Is there a way to do the debug while keeping the dat address ? If yes, it would be nice to update the documentation.

[pfrazee]

You can't nuke the .dat and keep the site. Even if the keys were kept, the lost data would create issues.

When you say "dat share is not working anymore," what's happening?

[jcolomb]

note: I had to reinstall nodejs today, and reinstall dat as I was started already.
note 2: it is also supposed to be available on hashbase (dat://rdmpromotion.hashbase.io/)

the terminal looks good:

dat v13.11.4
dat://30241d3351ff12f514e22fead851cc7d047d30439b45c4b2ff1d75cf57f9342b
Sharing dat: 944 files (68 MB)

2 connections | Download 0 B/s Upload 0 B/s

Watching for file updates

on the beaker browser, It can't find anything, if I look at the network debugger, the peer is found but then dropped

[pfrazee]

Huh. Yeah I'm able to create a connection with hashbase to view that dat but I'm not receiving the data. Nothing obvious in the log either ☹️

[jcolomb]

strange, the files seem to be accessible here: beaker://library/dat://30241d3351ff12f514e22fead851cc7d047d30439b45c4b2ff1d75cf57f9342b+latest

probably because I was seeding it via the beakerbrowser, too.

[pfrazee]

Yeah this is strange. I'm able to see the file listing and dat.json but none of the other files. Any chance the original source went down before those files got uploaded?

…

On Tue, Oct 30, 2018 at 5:38 PM Julien Colomb ***@***.***> wrote: strange, the files seem to be accessible here: beaker://library/dat://30241d3351ff12f514e22fead851cc7d047d30439b45c4b2ff1d75cf57f9342b+latest ?? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#1047 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABNhU9genEWKhyZW4LCdHBKOq1YGfQ2Jks5uqNTZgaJpZM4YC7vb> .

[jcolomb]

The hashbase version used to work. I never checked whether the beaker browser seeding was working.