Re: The first five minutes ... a thought experiment (long)

On 18 February 2014 11:35, Romain Deltour <rdeltour@gmail.com> wrote:
> It seems we all agree that one big issue with XProc is it's learning curve, which is steep and rathe long tailed. The 5 minutes problem is *one* aspect of it, but as others pointed out it practically affects far more than 5 mins of an XProc developer's life.
>

oh, I completely agree, but taking the approach that XProc big problem
is just a composition of smaller problems, we will also talk about 1st
day, 1st month and 1st year as well ... just breaking them up into
reasonable sized definitions as a first step.

> To be honest I was expecting this understandable kind of reaction to my proposal ;) The proposed change is indeed significant (hence the "v4" joke).

yes and pls do not interpret my comments as 'not for v2' ... the WG
decides that as a whole, with input from the community.

at a minimum I think the WG should a capture all proposed requirements
and spend some time discussing them.

> However, I would change first statement to "XProc v1 got the hairy things right **given the agreed-on constrained context**".
>
> In v2, there are two significant changes: (a) allowing any XDM in options -and ports?- and (b) allowing non-XML documents. I assume these changes would impact how you'd figure out the hairy things nowadays.
>

there has been some significant time (years) thinking through the
implications of XDM everywhere and non xml documents flowing through
pipelines; their inclusion in requirements doc is a direct function of
WG higher confidence that we understand the problem domain they
address, would serve to dramatically increase usability and
applicability, and have a good grasp on exact techniques of how to
implement them.

It maybe that with time and discussion, recent proposals will also get
included in requirements doc ... my reaction is based purely on the
fact that the WG has not spent the same amount of time rationalizing
inputs with options as outlined by your proposal; of course its a good
to have a minimum number of concepts and primitives and that impacts
usability ... and please do understand we will seriously discuss all
proposals.

> Changes like this make some concepts peculiarities even more subtle, which might mean even more difficult to fully understand for a newcomer.

Agreed, its like 'skating on thin ice, juggling chainsaws with people
shooting at you' to get this right.

but I have confidence that we have a good opportunity to achieve
something quite special with XProc vnext.

Keep the conversation rolling!

J

Received on Tuesday, 18 February 2014 11:23:30 UTC