Lyle, Vijay,

I read the draft quickly, and it sure is an interesting piece of work.

Let me start by understanding the context. To map into ALTO, we need to map
it into a client server architecture, and here is what I have:

- The (ALTO or another name) server has a set of ground facts. In a more
general setting, I see no need to be limited to the specific 3 (Sec. 5 BNF,
sub, node, link). Such a schema can be announced in ALTO IRD in
capabilities. The semantics should be defined in a document.

- I assume that the server can also predefine a set of predicates (aka
derived factors, database views) as well.

- The interface is that a client sends a query as a datalog program (btw,
the grammar in Sec. 5 appears to be only a subset of the examples in Sec.
7, which include ~, ==) to the server, who executes the program and returns
the result. I assume that the designers will argue that such a query
language is more general than SQL/XPATH.

This can be a major extension to ALTO, which uses essentially only simple
select (e.g., filtered services in ALTO) of an underlying (abstract)
network information database. Routing State Abstraction is going the
direction of trading computation at the server for better query,
abstraction results. Such a query language, if we build on top of ALTO, can
be quite interesting but will be a major extension.

Richard


On Mon, Jul 18, 2016 at 12:34 PM, Vijay K. Gurbani <
[email protected]> wrote:

> On 07/18/2016 03:56 AM, Lyle Bertz wrote:
>
>> All,
>>
>> Am I off track here or can ALTO pick this up rather easily?
>>
>> Link:
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__tools.ietf.org_html_draft-2Dcai-2Dnfvrg-2Drecursive-2Dmonitor-2D00&d=CwICAg&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=4G36iiEVb2m_v-0RnP2gx9KZJjYQgfvrOCE3789JGIA&m=46YuQ946RagfI5sboI-b-tEWqY6bJgdkCNnCA9oPJ7s&s=e_k_xhh8SEIIPxIbrdZH0RprASyhE-PwbbQuq09_RWs&e=
>>
>
> Hmmm ... speaking as 1/2 the chair of course, there is certainly no
> charter item against which ALTO can adopt this draft.  Process-wise, I
> don't know off hand whether a WG can add a new deliverable without
> re-chartering.
>
> But the broader question is what prompted you to correlate the draft
> with ALTO.  I could be missing something here, of course, but my quick
> scanning of the draft seems that it is a bit orthogonal to ALTO.  But
> as I said ... I may be missing something important.
>
> - vijay
> --
> Vijay K. Gurbani, Bell Laboratories, Nokia Networks
> 1960 Lucent Lane, Rm. 9C-533, Naperville, Illinois 60563 (USA)
> Email: [email protected] / [email protected]
> Web:
> https://urldefense.proofpoint.com/v2/url?u=http-3A__ect.bell-2Dlabs.com_who_vkg_&d=CwICAg&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=4G36iiEVb2m_v-0RnP2gx9KZJjYQgfvrOCE3789JGIA&m=46YuQ946RagfI5sboI-b-tEWqY6bJgdkCNnCA9oPJ7s&s=pIyGVFETcsy0HunWMxU7NzI_Ye2eoOkDfHpHbcktDI4&e=
>  | Calendar:
> https://urldefense.proofpoint.com/v2/url?u=http-3A__goo.gl_x3Ogq&d=CwICAg&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=4G36iiEVb2m_v-0RnP2gx9KZJjYQgfvrOCE3789JGIA&m=46YuQ946RagfI5sboI-b-tEWqY6bJgdkCNnCA9oPJ7s&s=dxs-VcSriMGSGg5_RDNp8AaEuj3mIJba-IbjWf0l6uk&e=
> _______________________________________________
> alto mailing list
> [email protected]
>
> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.ietf.org_mailman_listinfo_alto&d=CwICAg&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=4G36iiEVb2m_v-0RnP2gx9KZJjYQgfvrOCE3789JGIA&m=46YuQ946RagfI5sboI-b-tEWqY6bJgdkCNnCA9oPJ7s&s=B6C13fyjgDaTvAGUidK_6N88RsBJI5uioPBjkzzGFVk&e=




-- 
-- 
 =====================================
| Y. Richard Yang <[email protected]>   |
| Professor of Computer Science       |
| http://www.cs.yale.edu/~yry/        |
 =====================================
_______________________________________________
alto mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/alto

Reply via email to