Re: XS: Guidelines for the discussion

Subject: Re: XS: Guidelines for the discussion
From: James Clark <jjc@xxxxxxxxxx>
Date: Sat, 24 May 1997 11:11:36 +0700
At 17:09 23/05/97 -0500, Alex Milowski wrote:
>> 2. Bearing in mind that the point of dsssl-o (and hence XS) is to give
>> implementors something reasonable to aim for, I am going to be guided
>> by the principle that nothing in DSSSL that is not in Jade is going to
>> be required by the specification.  If James can't implement it,
>> there's no point in talking about it.
>
>I think this should be rephrased as "all DSSSL implementors".  Although
>James has been doing very good work with Jade, his priorities for what goes
>into Jade should not affect what goes into XS. 
>
>James, Norbert, myself and others, as DSSSL implementors, should be able to
>tell this group that the features being asked for are reasonable and
>appropriate regardless of how fast they can or *want* to be implement such
>features.
>
>Remember, there is more than James out here working on DSSSL.

I agree that Jade (and myself) shouldn't have any special status in the
process.  But I also don't think there should be anything in the XS spec
that isn't implemented in at least one implementation (preferably several)
that is:

- publicly available (preferably with source)

- reasonably close to production quality in terms of performance (ie an
implementation that  is unacceptably slow shouldn't count).

While it's under development, I think the XS spec can move ahead of
implementation, but I would be very uncomfortable if, when it is finalized,
there is anything unimplemented in it.

I can make a good guess about how hard something will be to implement, but
without actually implementing it I can't be sure.  The only way to be really
sure of ironing out all the bugs in a spec is to have it be implemented and
used.

James




 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist


Current Thread