jakarta-taglibs-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "A. Keyton Weissinger" <keyt...@bellsouth.net>
Subject RE: Discussion Initiation: Incorporation of custom tag libraries into authoring tools.
Date Sun, 03 Dec 2000 22:49:07 GMT
Excellent points, Shawn. I am coming into this topic from a slightly
different viewpoint. Your explanations help a great deal.

On the standardization issue and whether Sun sees this as part of its goal,
I have no disagreement whatsoever. In fact, I was not -- in any way --
trying to imply "Why is Sun not doing X?" I'm sorry I may have come off like
that. The main point I was trying (perhaps poorly) drive home was the
importance of this issue. You may see that as self-evident, but I'd seen no
mention of it to date in this (dev) or the users list (although admittedly
I've not been involved for very long).

Also too, I am (unfortunately) not as well seated in "the right way" as many
of the participants of the group are. I'm scrappin' with too few developers
trying to do too much. Now before anyone hammers me about bringin' out the
violin, I know that I am in no way unique in that situation.

Like probably some of you, I'm the manager of the software development
group. We are part of a small startup working on an ASP for hotels. I
have -- count 'em -- 5 developers with very widely varying skill sets, from
a seasoned EJB guru on one end to two folks with extensive hotel industry
and very little java experience on the other. Tag libraries will go a long
way in making the one end productive using the labor fruits of the other
end, if you see what I mean, but right now, they are struggling to make
heads and tails of the java language, web development, the tools, etc.
Perhaps my frustration at not being able to incorporate tag libraries into
their tools and send them on their way was showing a bit. I apologize. Like
many of you, I see the power that taglibs present...and also their
immaturity.

As to your questions, I am very excited about discussing them:

"How can we make tag libraries easier to incorporate into editing tools?" or
"How can we design a standard tag library such that we can approach the goal
of complete, complex JSP pages universally written without scriptlet code
and relying solely on tags?"

These are the types of questions I wanted to lead to. I have some ideas and
will follow up. Any from anyone else in the group?

Finally, as for the JavaDoc statement, this may be better positioned as "How
can we make JavaDocs for tag libraries easier to incorporate into editing
tools?"

Again, thank you, Shawn, for your thoughtful (and not flameful) reply. My
(perhaps misguided) goal was only to stimulate conversation about this, not
to step atop a pedestal and say "This is how it should be." ;-) Hell, I
don't know near enough about anything to do that.

Keyton



-----Original Message-----
From: Shawn Bayern [mailto:shawn.bayern@yale.edu]
Sent: Sunday, December 03, 2000 5:00 PM
To: taglibs-dev@jakarta.apache.org
Subject: Re: Discussion Initiation: Incorporation of custom tag
libraries into authoring tools.


On Sun, 3 Dec 2000, A. Keyton Weissinger wrote:

> Few want to address it because it's sticky and pretty non-glamorous, but
it
> is bigtime important. It's my understanding that Macromedia has made some
> mention of putting this feature (the ability to "hook" in tag libraries)
> into Dreamweaver. Anyone have any details? What about other design/layout
> tools (Homesite, FrontPage (doubt it), etc)? This seems like a hugely
> important part of getting tag libraries into the hands of those people
from
> which we all talk about "hiding complexity."

Although I can't speak for Sun, I believe this has been their intent for
some time; I think this point is already generally accepted.  For specific
details as to which individual vendors have plans to incorporate tag
libraries into their editing tools, I'd suggest asking the vendors
directly.  But the idea of an association between tag libraries and
editing tools, as well as an awareness of its importance, seems evident
from Sun for a while.  For instance, JSR 052 refers explicitly to
simplification of editing-tool use as a benefit of tag standardization.
(That's one answer for "Anyone have any details?" -- that through JSR 052,
the community will converge on a standard that, among other things, will
make incorporation of tag-based functionality into editing tools simpler.)

> As we all know, approachability is the key to this topic's growth in
> importance. Incorporation into authoring tools is a biggie. We (and I
> will definitely help on this) need to "lobby" to make this "cake" for
> authoring tools vendors.

It might be the case -- although I couldn't swear to it -- that vendors
are already convinced and are merely waiting for standardization.  Asking
the group of vendors that are part of the expert group for JSR 052 might
be particularly worthwhile.

> If these taglibs do not become easier to incorporate into the
> day-to-day development tasks of the average web page developer,
> taglibs' success going forward will be more difficult, relying solely
> on its use by junior java developers that have been relegated to the
> JSP end of the project because the EJBs are, at least currently, not
> within their skillset.

I'm no expert in the social dynamics of programming shops, so I suppose I
should accept this on face value, but it seems simplistic.  My personal
inclination is to deal with the issue as a technological, not social, one;
I'm personally more interested in questions like, "How can we make tag
libraries easier to incorporate into editing tools?" or "How can we design
a standard tag library such that we can approach the goal of complete,
complex JSP pages universally written without scriptlet code and relying
solely on tags?"

Also, I'm still not sure I follow the specific point you're making.  What
do you mean by "use by junior java developers"?  'Use' as in development
of the taglibs themselves or incorporation of them into pages?  I'm afraid
I'm just not getting why the average skillsets of JSP programmers is
relevant. :)

Taglibs essentially provide an interface across which different groups of
programmers and users can communicate.  This boundary may or may not be
useful to a particular organization, but it can appropriate programmers
and page designers on either side of the boundary as it sees fit, no?

> Along these same lines, help systems is a whole 'nother topic along these
> lines. Yes, we have this incredible JavaDoc system that programmers can
use
> to see how tags are put together, but joe-schmoe web page designer doesn't
> care. He wants easy to use (preferably pop-up) help on how to USE the
> things. How best to do this? It's up to us as tag library developers to
> provide this help. How to make the incorporation of this help and the tag
> libraries themselves easy?

That's a good question, although I wonder if JavaDoc isn't actually
sufficient already.  Each tag is a class; the JavaDoc header for that
class could provide general information suitable to its usage, and
individual accessor methods could (as a first thought) describe what
various parameters mean semantically.  Are you suggesting that a markup
with greater semantic structure be proposed on top of (or independently
from) JavaDoc to describe to a tag user how the tag represented by a
particular class is supposed to function?  If so, what specifically would
it add to the current system that can't already be done?

(That's not a rhetorical question :) -- I'd actually like to know more
about what you're proposing.)

Shawn



Mime
View raw message