Re: [Tools-discuss] A major restructure to the datatracker database will be deployed soon

[Robert Sparks <[e-mail]>]

On 11/20/23 9:41 PM, John C Klensin wrote:
> Robert,
>
> Tiny apparent glitch with a link that leads to itself.
>
> Example:
>
> Look up RFC5321.  Everything looks fine except there is a "Was"
> entry listing draft-klensin-rfc2821bis.  That is correct, except
> it is a link leading to
> <https://dt-rfc.dev.ietf.org/doc/draft-klensin-rfc2821bis/11/>.
> And that immediately redirect to
> <https://dt-rfc.dev.ietf.org/doc/rfc5321/>, which is where I
> started.   Confusing and time-wasting at best.

Captured at #6699

This is actually an issue with older documents not having correctly 
populated DocHistory. While we'll make an immediate change to more 
closely follow what the datatracker main branch currently does,  we will 
be addressing better using the data we _do_ have with the cluster of 
issues similar to #6059

>
> Based on a small sample, the problem is not unique to that
> document.
No, but it isn't _all_ documents.
>
> Comment (I don't know if this is a Tools issue or an RSWG one,
> but let's start here): Since before there was a datatracker,
> before there were streams, and, indeed, before there was an
> IETF, the authoritative source of information about RFCs has
> been the RFC Editor, including their web site since that web
> site was established.  This revision and interface seems to take
> us much closer to the datatracker being the authoritative source
> or, perhaps, a redundant authoritative source (almost
> universally considered a bad idea among those with database
> management expertise).  If that is not your intention, it seems
> to me that the display pages could use some disclaimers.  If it
> is, it seems to be to be a strategic change, not just a tooling
> one.

I am aware of the concern you point to. I disagree that this increases 
the potential confusion for what's authoritative beyond what the 
datatracker currently displays. The model changes will make it _easier_ 
to build UI that steers people to the right places as the datatracker 
and the rfc-editor website evolve. The RPC and the RSCE are steering 
that evolution already.

>
>     john
>
>
> --On Monday, November 20, 2023 17:04 -0600 Robert Sparks
> <[e-mail]> wrote:
>
>> We've been working for several months on making RFCs
>> first-class objects in the datatracker instead of modeling
>> them as an artificial version of a draft.
>>
>> This change lets us simplify some code and query constructs
>> (*), resulting in quicker datatracker response, and makes it
>> possible to start modeling the subseries, such as BCP and STD.
>>
>> Most views won't change, but a few change significantly. The
>> history view for documents splits the history of the RFC and
>> the history of the draft into separate tabs. A few things
>> (like internal-to-the-ietf metadata for the processing of the
>> draft) won't be visible when looking at the RFC - you'll have
>> to browse to the linked draft to see that.
>>
>> We're in the last stages of preparing to release this, and
>> expect to have it in production two to three weeks from now.
>>
>> Please take some time to look at the development instance at
>>
>> https://dt-rfc.dev.ietf.org
>>
>> and watch for things that don't seem right. Raise any issues
>> you find by mail to tool-discuss or opening a github issue -
>> please mention that you're looking at dt-rfc in the issue
>> description so we know it's a issue with that branch and not
>> main.
>>
>> RjS
>>
>> * For those using the v1 API - this change removes the
>> DocAlias model, and adds a RelatedDocument object of type
>> "became_rfc" between the Document of type "draft" and the
>> related document of type "rfc".  Note in particular that the
>> RelatedDocument model changes such that target points to
>> Document objects and not DocAlias objects. See the changes to
>> the model files at
>> https://github.com/ietf-tools/datatracker/compare/main...feat/
>> rfc for details.
>>
>>
>> ___________________________________________________________
>> Tools-discuss mailing list - [e-mail] -
>> https://www.ietf.org/mailman/listinfo/tools-discuss
>
> ___________________________________________________________
> Tools-discuss mailing list - [e-mail] - https://www.ietf.org/mailman/listinfo/tools-discuss