W3C home > Mailing lists > Public > public-html@w3.org > April 2016

Re: Remove table sorting - issue 56

From: Sangwhan Moon <sangwhan@iki.fi>
Date: Mon, 11 Apr 2016 19:34:32 +0900
Cc: "Schnabel, Stefan" <stefan.schnabel@sap.com>, Steve Faulkner <faulkner.steve@gmail.com>, Chaals McCathie Nevile <chaals@yandex-team.ru>
Message-Id: <167C7A6E-D1F7-4D91-B654-90D1E9DF448D@iki.fi>
To: "HTML WG (public-html@w3.org)" <public-html@w3.org>

> On Apr 11, 2016, at 7:06 PM, Chaals McCathie Nevile <chaals@yandex-team.ru> wrote:
> On Mon, 11 Apr 2016 09:44:41 +0200, Steve Faulkner <faulkner.steve@gmail.com> wrote:
>> On 11 April 2016 at 08:32, Schnabel, Stefan <stefan.schnabel@sap.com>
>>> Instead, proposal for respective API sort mappings should be actively
>>> communicated to the browser vendors and the group should be open for
>>> discussion here.
>> It has been in the spec for years, and browser vendors have declined to
>> implement. moving the spec to recommendation involves trimming
>> unimplemented features ,as they don't meet the requirements for a rec.
>> I don't think its Chaals intention to bin the sorting alogorithm, but to
>> move it out of HTML5.1 where it is not appropriately included.
> Correct. I think the situation is exactly as Mikeā„¢ described elsewhere, a failure in this case largely of browsers to take the "Priority of constituencies" seriously.
> I don't know the reason, but probably no browser engineer ever felt this was a priority given that there are any number of scripts that do it. Although that is generally a sign that this feature is a priority for developers.
> One of the obvious questions is whether the scripts shipped around are sufficiently performant, and I don't know the answer.
> I'll probably put something into the WICG group, which is meant to be a place where browsers consider things which are not in the HTML spec, but which people would like to have there.

Native table sorting only solves a small subset of the problem; sorting what is already on a document for a reasonably small dataset, which in most cases should be performant enough with userland scripts.

For any application with a large set of data (e.g. any "Enterprise" application), where the table sort mechanism's intent is to "sort from the entire data set" - this wouldn't work without some sort of data table-server data binding standard, which probably is not something that is going to happen in the foreseeable future, if not never. 

(There is the point of "but it would let people sort tables with Javascript off", but that's a separate macro discussion which I prefer not to bring up here.)

Received on Monday, 11 April 2016 10:35:00 UTC

This archive was generated by hypermail 2.3.1 : Monday, 11 April 2016 10:35:01 UTC