[DBWG] Fwd: Re: --list-versions query on deleted resources?

[Ben Maddison]

Hi all,

The list got left off of Michel's initial response to my email, and we
only noticed once several further replies had been exchanged.

See the reconstructed exchange below.
TL;DR: Should prior versions of deleted objects remain query-able?

Cheers,

Ben

----- Forwarded message from Ben Maddison <benm at workonline.africa> -----


> Date: Thu, 3 Dec 2020 20:48:32 +0200

> From: Ben Maddison <benm at workonline.africa>

> To: Michel ODOU <michel.odou at afrinic.net>

> Subject: Re: [DBWG] --list-versions query on deleted resources?

> 

> Hi Michel,

> 

> On 12/03, Michel ODOU wrote:

> > Hi Ben,

> > 

> > On 03/12/2020 18:53, Ben Maddison wrote:

> > > Hi Michel,

> > > 

> > > On 12/03, Michel ODOU wrote:

> > > > Hi Ben,

> > > > 

> > > > --list-versions does not show the versions of a deleted object.

> > > > If the object has been deleted several times, then you will see the

> > > > versions since the last deletion.

> > > > 

> > > This seems undesirable. Why not provide the ability to see the versions

> > > that existed prior to the eventual deletion?

> > > Was this a conscious design decision, or a limitation of the db code?

> > > 

> > > > Example, versions of object X:

> > > > 

> > > > #1: created

> > > > #2: updated

> > > > #3: updated

> > > > #4: deleted

> > > > #5: created

> > > > #6: updated

> > > > 

> > > > Then --list-versions will display versions #5 and #6.

> > > > 

> > > > In case there is an additional deletion (#7), then --list-version will

> > > > simply inform you the last time the object was deleted.

> > > > 

> > > > And of course, it does not work on role and person objects.

> > > > 

> > This decision was taken by RIPE - our current WHOIS derives from their

> > 2013 code base, even if it has strongly diverged since. Maybe there were

> > some privacy concerns (at least, I know that it is why you don't see the

> > history for person and role objects). On our side, we have never changed

> > this, simply because no one in the community ever complained about the

> > current behavior.

> > 

> Ack, makes sense.

> 

> > If we had to change this, it would be easy. There is no specific

> > limitation in the WHOIS code, it is possible to get all the versions

> > that existed for any object, even with deletions in the middle. It is

> > just a matter a getting a consensus around this.

> > 

> I would personally be in favor of this.

> 

> Everyone else? Thoughts?

> 

> Cheers,

> 

> Ben


----- End forwarded message -----
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.afrinic.net/pipermail/dbwg/attachments/20201204/3f3ca74d/attachment.sig>