W3C home > Mailing lists > Public > spec-prod@w3.org > July to September 2013

Re: Editor's drafts on /TR/… ftw, was Re: new TR tools and editor's drafts?

From: fantasai <fantasai.lists@inkedblade.net>
Date: Wed, 03 Jul 2013 12:41:29 -0700
Message-ID: <51D47E69.3030009@inkedblade.net>
To: spec-prod@w3.org
On 07/03/2013 06:55 AM, Robin Berjon wrote:
> Aside from converting from ED to WD, which ought to be trivial [...]
> If as an editor you want to be allowed to make publication requests
> on a regular basis in order to keep the TR copy fresh

Sorry, it's not that simple. CSS specs for example, include
a lot of images. To send a spec to the webmaster, I have to
create a zipfile. And then he has to link check it, because
I don't have write access and can't link-check in place, so
sometimes there's some back-and-forth because some file is
missing or something else is broken. This is definitely not
a 5-minute process, *especially* for someone who only edits
one spec as a side-project to their implementation and doesn't
have all the steps to publish and relevant commands memorized.

Then oftentimes I'll notice a typo or something else stupid
three days later, and can't fix it because it's not worth
the overhead of generating a new WD and engaging the Webmaster
to retro-actively fix an old publication.

No, the current publication process is NOT trivial. Totally fails
the 5-minute requirement, even if we allowed for "it shows up on
/TR" to have a 24-hour automatic delay (which may not take up any
of my time, but still takes up mental space). There is no way I'm
going to spend that much time and memory publishing specs *every
single week*.

Then there's the signal:noise problem for announcements and

If W3C announced a new publication every time Marcos and I fixed
a minor glitch in the specs, the newsfeed would be mostly noise.
You really only want front-page announcement *occasionally*. In
my style of working, it would be after I've collected together a
series of changes and am ready to ask the wider public, that
doesn't have time to follow CSS commit microfeeds, to take the
time to look at what we've done in the last X months/weeks and
send comments.

(Also, currently the WD publications in the CSSWG serve as
"checkpoints", and it is sometimes useful to walk back through
these rather than dive through the "continuous save" hg log.
It's definitely nice to have a high-level history sometimes.
If I republish every spec every Tuesday and Thursday to keep
it up-to-date, we lose that.)

The current system is designed for snapshots. It does them well.
I'm happy for it to continue to do them well. But let's stop
pretending that the same system does "live" publications well.
It is not worth it to anybody to waste my time on that.

Received on Wednesday, 3 July 2013 19:42:07 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:55:19 UTC