W3C home > Mailing lists > Public > semantic-web@w3.org > June 2020

Re: defining the semantics of lists

From: Hugh Glaser <hugh@glasers.org>
Date: Wed, 10 Jun 2020 18:35:31 +0100
Cc: thomas lörtsch <tl@rat.io>, Peter Patel-Schneider <pfpschneider@gmail.com>, semantic-web <semantic-web@w3.org>
Message-Id: <64E13E03-4EAE-4909-9835-56612D2355FC@glasers.org>
To: Thomas Passin <tpassin@tompassin.net>
Hi Thomas P,

I had a feeling you didn't want to :-)
Thanks for doing it.
But as you probably expected, I agree with Thomas L - why on earth would you want to use a list for your unordered favourite colours?
That would force you to decide on an ordering, and worse still the ordering might appear to have some meaning.
Surely three simply triples is a much better representation of this knowledge?

[And I know that hans.teijgeler@quicknet.nl said "For each of those you guys will undoubtedly invent some workaround, but that isn't good enough.". I don't think that is a workaround - it a correction of actually incorrect representation of the knowledge.]

I am reminded of trying to program in languages with poor qualities for the representation of complex data structures (such as Lisp), where you end up squeezing everything into the same single structure - a list or perhaps an array.

Massaging data structuring and deforming the data into your favourite construct is not a good way to go about programming, methinks.

Have you got any better examples, perhaps?


> On 8 Jun 2020, at 16:20, Thomas Passin <tpassin@tompassin.net> wrote:
> On 6/8/2020 10:26 AM, thomas lörtsch wrote:
>>> On 8. Jun 2020, at 14:07, Hugh Glaser <hugh@glasers.org> wrote:
>>> Hi Thomas,
> [snip]
>>> As a specific example of my problem, below you say: "I don’t see the pressing need for empty lists."
>>> I think that means you have ideas of applications in mind - if you can expose those, please, that would be great.
>> I didn’t for two reasons:
>> - lists are such a fundamental concept that IMO examples rather cloud the view
>> - my primary interest is to replace rdf:Lists with rdf:Collections and being able to close is what differentiates them on the semantic level
>> But if you insist:
>> - a list of my three favorite colors, unordered but finite.
> Not a list, so why try to make a list out of it?
>> - tables made from lists where it is important that each list (table row) has the same length
> If they are not the same length in the original set of triples, then you have to truncate them somehow.  If they are the same length, then you don't need to do anything special.  Either way, I don't see this as anything that needs list characteristics.  The list-ness of the data has nothing to do with how you fill a table with it.
>> And the application that I have always in mind is the contextualizing integrate-all-the-things that will finally bring love, peace and understanding to the world, ahem. But lists don’t play a very prominent role in that.
>> Hope that helps
>> Thomas

023 8061 5652
Received on Wednesday, 10 June 2020 17:36:11 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:46:04 UTC