Re: Re: Re: Re: Have you looked at UIML?

I should have read all the posts before responding to your last post. My 
apologies. I'll make a few additional comments.
On Saturday 31 August 2002 02:47 pm, Gerald Bauer wrote:
>    Anygui is a fantastic project and by its very
> nature is more pythonic than XUL. Both can exist
> side-by side peacefully.

We agree here.

>    As mentioned before I want to stress the importance
> of XUL as it has the same goal as Anygui, that is,
> creating rich UIs independent of the GUI toolkit.

And we agree here.

>    You don't need to rush into supporting XUL, but
> putting it on the backburner isn't the way to go
> either.

I don't believe there is a desire to put it on a back burner. I think the main 
concern is jumping into XUL before the relevant subject matter is studied. 
I'm using libglade and .glade files in a small application at the moment and 
it has been instructive in understanding how GUI development can be enhanced 
by an XML declaration of the UI. I'll reimplement it using XUL once a 
reasonable XUL library is available for Python (Jython's not an option 
because of other constrains on the project.) An XUL builder application would 
also be an added incentive to move to a toolkit-neutral UI description 
language.

>    XUL somehow seems to be held back by the myth that
> it is a Mozilla only affair: No, XUL (that is, XML
> User Interface Language) is like HTML independent of
> the browser or toolkit (although XUL is still evolving
> and not sanctioned by the W3C it will eventually get
> standardized like HTML in the early days; after all a
> shake-out in the real-world is better than
> recommending a spec that is dead on arrival).

No, I get the connection and I understand the history. I watched in horror as 
the wartish features, such as the <blink> tag, were added to Mosaic. Those 
are the kind of impulsive decisions that need to be avoided. That requires 
carrying on a dialog with co-developers. Simply accepting XUL as it currently 
exists is short sighted.

>    XUL and XML UI alternatives: For now XForms is the
> only viable alternative. However, XForms is just a
> building block (that is, a subset) of XUL. Conclusio:
> XUL and XForms don't compete. XForms is just a
> stripped down, minimal XUL ultra-lite version. As an
> example: XUL includes tags like toolbar, menubar,
> tree, datagrid, popup, and so on all of which are
> missing in XForms.

Agreed, the XForms specification shot too low, but they had a different 
target. They wanted to replace the minimally featured <form> tag, not define 
a toolkit independent UI definition language. That was probably considered 
too ambitious. Coordinating the good intensions of people willing to help 
define standards is a cat-hurding exercise. It's a messy business, every once 
in a while the stars line up and everyone faces the same direction for a 
brief period of time.

> > Given my current understanding of the situation
> > (which I am more than
> > happy to revise, mind you) I think adding a XUL
> > component/parser/thingy to Anygui is fine. Other
> > than that, I'm not
> > sure what you would like us to do. (Changing
> > component names etc.
> > might be fine too, but then I'd like concrete naming
> > suggestions
> > really soon -- from anyone who't interested in the
> > matter.)
>
>   Anygui is your project and all I wanted to achieve
> is to stress the importance of XUL. It's up to you to
> decide how to add XUL to anygui and make anygui more
> accessible and useful.

Please make Anygui your project as well. Send in suggestions, and code, on how 
XUL would best be integrated.

>   As I wrote earlier, I will get back with more
> specific thoughts once I read your Practial Python.
> Expect some specifics in mid-October.

Excellent. I look forward to reading what you have to say. Please take a look 
at Andrew's comments on Python. His analysis is wonderful. 

> > Whether he is an expert, has studied the issue for
> > long time and all the trade-offs or he is just in an
> >
> > honey-moon phase with a technology  or a zealot,
> > it does not matter, talking down to people and
> > brochureware arguments sound simply incompetent and
> > arrogant to technical ears.
>
>   Unfortunately, I don't have the time to reproduce my
> XUL talk slides and tech docs in every single email
> and in addition who is going to read dozens of pages
> or more.

I've read the XUL talk slides. They are very helpful, but referencing them 
instead of declaring UIML junk would have been more constructive.  
Regardless, I still don't see why you want us to ignore the UIML document. If 
the design is bad we'll see for ourselves. It's possible to learn from bad 
designs as well as good ones.

>   I invite you to browse over the Luxor XUL docs at
> http://luxor-xul.sourceforge.net and see if your
> brochureware argument is justified.

I've looked at it. It sounds great on paper, but I was not able to get the 
software to run on my Linux system. I didn't have time to track down what 
would be required. (I was looking at the version that used the IBM widget set 
instead of Swing.)  All I found was a Windows executable. I'd like to tinker 
with it, but unfortunately the Java base for the implementation is a bit of a 
disincentive. My project is using the C version of Python.

>   The goal of XUL like HTML is freedom (that is,
> freedom to choose your GUI toolkit, your scripting
> language, your runtime, and so on).

Yes, I got that from your documents.

>  Brand me as zealot or whatever, but I don't accept
> the single-vendor milk machine toolkits that are
> prelevent today if you build desktop apps .

This is the Anygui mailing list. I suspect there are few on this list that are 
content with a single-vendor solution:-)

>  Hoping that Micropoly, Sun, IBM or any other
> multinational will liberate you is misplaced.
> Tim-Berners Lee lead the HTML campaign, now it's time
> to champion the XUL campaign to build on the succes of
> the HTML story.

And more power to you in the quest to make XUL ubiquitous, but being a 
champion requires articulating a well reasoned case for why those who are not 
"true believers" should care to join your cause. It's not your goal that are 
grating, it is the tactic for selling the elixir.

I look forward to trying Luxor in a couple weeks. I've got a deadline and I've 
been avoiding it somewhat by indulging myself in this thread.



-------------------------------------------------------
This sf.net email is sponsored by: OSDN - Tired of that same old
cell phone?  Get a new here for FREE!
https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390