RE: [Jersey] Goals for hypertext constraint support

From: Markus Karg <>
Date: Thu, 18 Feb 2010 21:20:26 +0100

> My concerns have to do with what someone (maybe Paul) commented as
> moving the goal posts -- that is, changes in what was assumed. This is
> often tactics used by people who care more about appearing to win
> discussions. Based on your comments, you are not such a person, I am
> just bit sensitive to appearance of such traits, having had to work
> with people that have had the problem.

Actually it was not tactics that I changed position but I in fact got
convinced by others whilst the discussion that I was wrong *in part* so I
adapted my opinion about some technical details.

> > I do not see what is not honest in the discussion, and the idea of
> invite
> > Roy was to shorten the discussion, not to "win" in any sense. In
> fact, I
> Inviting Roy was fair, I did not mean to imply that was not.
> I was feeling follow-up comments were not necessarily appropriate at
> least wording (regarding "silly RPC approach" [or something to that
> effect]). This felt like framing discussion less objective.

Maybe my wording was too hard, cannot actually remember what I wrote in
detail. For me the question was not precise enough and it looked like Marc
tried to ask in a way that only his answer would be possible (what in part
happened), so for me *his* question was unfair since Roy was not presented
the actual context. But do we really have to roll-up again this all? Can't
we just drop that thread?

> I do not feel that others made more interpretations and you less.

Well, it's a feeling then, because I did not do any interpretations but just
cited, what is a huge difference (maybe you missed my blog entry about that
-- it's intention is that people shall not do interpretations at all, so it
would be strange if I would interprete on my own).

> And
> I did read Roy's responses, and consider them to be among good useful
> part of discussion. I did notice he did consider headers/body
> distinction as less of a semantic issue, which is more in line with
> what others were suggesting.

This is true, but take care: He explicitly mentioned some headers with
different "validity"! This is what I mean with "unfair kind of question". I
never doubted that it is possible to have links in headers. But I said that
is invalid to put them in ANY (in the meaning of FREE CHOICE), and it will
lead to such strong incompatibility that it is just useless. As you see, his
answer was more or less diplomatic and salomonic, but not precise -- since
he couldn't answer precisely without the actual context of the question.
That I understood to be unfair, and it was the trigger why I wrote back
-possibly- too hard.

> But as to being external, as far as I know, Roy has not been
> participant of this mailing list; and so although has was obviously
> part of that particular thread, in larger context he is external.

Right! And that is why I said that the type of question was unfair, since it
doesn't contain the context.

But again, let's drop the threat and concentrate on the API. :-)