Re: properties

["Peter B. West" <pb...@powerup.com.au>]

Hello devs.

Does the near-silence on this one signify consent?

Peter

Keiron Liddle wrote:

> Hi Peter and others,
>
> What is the status of property handling?
>
> Are we going to go with Peter's suggestion and commit the java source 
> to cvs and work on handling properties fully from there?
>



---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org

Re: properties

["J.Pietschmann" <j3...@yahoo.de>]

Bertrand Delacretaz wrote:
> On Wednesday 01 May 2002 18:19, Peter B. West wrote:
> 
>>Does the near-silence on this one signify consent?
> 
> 
> I don't know enough about this to give meaningful advice, so in my case yes, 
> silence means consent.
> 
Yes, it ought to be a reasonable idea.

J.Pietschmann


---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org

Re: properties

[Bertrand Delacretaz <bd...@codeconsult.ch>]

On Wednesday 01 May 2002 18:19, Peter B. West wrote:
> Does the near-silence on this one signify consent?

I don't know enough about this to give meaningful advice, so in my case yes, 
silence means consent.

- Bertrand

---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org

Re: properties

["Peter B. West" <pb...@powerup.com.au>]

Karen Lease wrote:

>Peter,
>
>I see you've committed all your stuff. Great! I hope to be able to take
>a look at it, in particular the properties-related stuff, but I'm not
>setting any dates, as May is going to be rather hectic, due to some
>non-FOP and non-work related activities I'm involved in.
>  
>

When you do look, you may be immediately interested in the changes to 
the parser and tokenizer, and the handling of complex property 
expressions. In Properties.java some of the properties define a 
complex() method for interpreting the results from the parser. These 
methods are accessed through the pseudo-array complexMethods in 
PropertyConsts.java. I hae made quite a few extensions to the tokenizer 
and the parser after a rethink of exactly what types were being returned.

With complex properties, I eventually gave up on trying to define a 
common set of elements which would deal with all of the cases, and 
particularised them using the complex() method. I plan to take a similar 
approach with the Auto datatype (yes, datatype.)

>Although I'm sure I'll have an opinion, I don't have any more vested
>interest in this area than any other FOP developer, so you certainly
>don't need to wait on my reaction!
>
>My feeling is that the current scheme does a lot of stuff fine, but
>doesn't do everything. Of the things it doesn't do, some are just due to
>incomplete implementation; others are perhaps due to fundamental flaws
>(perhaps layout-related value evaluations, but this is a general
>problem).
>
>As for the build system for the property maker classes, which I recall
>you discussing with Arved a few weeks ago, I agree that it's rather
>complex. However aside from having giving me some XSL practice, it did
>make it possible to quickly introduce changes into the whole "Maker"
>scheme such as handling of corresponding and (to a limited degree)
>shorthand forms. I don't at all agree that generating code with XML/XSL
>is a bad practice; in general languages are always evolving to a higher
>level of abstraction, and this is just another step in that ladder, much
>like IDL for defining interfaces.
>

I'm all for high levels of abstraction, as long as the result is also 
complete; that I have all of the expressiveness I need to solve the 
problems I face. IDL is, I think, a bit different from the case of code 
generated for purely internal consumption.

I don't want to sound too negative about, in particular, your efforts, 
but I see that there is an increasing amount of XSL-generated code out 
there in the wider world, and I think it's a maintenance disaster in the 
making. If your language won't do it, either get a new language or live 
with the limitations. (I confess that I have never been one for living 
with the limitations - e.g. statics - but it's a case of do what I say, 
not what I do.)

Peter


---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org

Re: properties

[Karen Lease <kl...@club-internet.fr>]

Peter,

I see you've committed all your stuff. Great! I hope to be able to take
a look at it, in particular the properties-related stuff, but I'm not
setting any dates, as May is going to be rather hectic, due to some
non-FOP and non-work related activities I'm involved in.

Although I'm sure I'll have an opinion, I don't have any more vested
interest in this area than any other FOP developer, so you certainly
don't need to wait on my reaction!

My feeling is that the current scheme does a lot of stuff fine, but
doesn't do everything. Of the things it doesn't do, some are just due to
incomplete implementation; others are perhaps due to fundamental flaws
(perhaps layout-related value evaluations, but this is a general
problem).

As for the build system for the property maker classes, which I recall
you discussing with Arved a few weeks ago, I agree that it's rather
complex. However aside from having giving me some XSL practice, it did
make it possible to quickly introduce changes into the whole "Maker"
scheme such as handling of corresponding and (to a limited degree)
shorthand forms. I don't at all agree that generating code with XML/XSL
is a bad practice; in general languages are always evolving to a higher
level of abstraction, and this is just another step in that ladder, much
like IDL for defining interfaces.

Regards,
Karen


"Peter B. West" wrote:
> 
> Karen,
> 
> I was your opinion that I was most concerned about on this one. I, for
> one, am quite happy to hold off to give you more time to make an
> argument. Part of my point was that it *was* in pretty good shape the
> first time around.
> 
> Peter
> 
> Karen Lease wrote:
> 
> >Not really. As a fairly interested party, I must admit that mentions of
> >throwing it out and starting over don't leave me totally indifferent,
> >but I've got little enough time as it is for FOP these days, and I'm
> >trying not to let myself be distracted.
> >In any case, we all know that it's never right the first time around.
> >
> >
> >
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
> For additional commands, email: fop-dev-help@xml.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org

Re: properties

["Peter B. West" <pb...@powerup.com.au>]

Karen,

I was your opinion that I was most concerned about on this one. I, for 
one, am quite happy to hold off to give you more time to make an 
argument. Part of my point was that it *was* in pretty good shape the 
first time around.

Peter

Karen Lease wrote:

>Not really. As a fairly interested party, I must admit that mentions of
>throwing it out and starting over don't leave me totally indifferent,
>but I've got little enough time as it is for FOP these days, and I'm
>trying not to let myself be distracted.
>In any case, we all know that it's never right the first time around.
>
>  
>



---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org

Re: properties

[Karen Lease <kl...@club-internet.fr>]

Not really. As a fairly interested party, I must admit that mentions of
throwing it out and starting over don't leave me totally indifferent,
but I've got little enough time as it is for FOP these days, and I'm
trying not to let myself be distracted.
In any case, we all know that it's never right the first time around.

Regards,
Karen

"Peter B. West" wrote:
> 
> Hello devs.
> 
> Does the near-silence on this one signify consent?
> 
> Peter
> 
> Keiron Liddle wrote:
> 
> > Hi Peter and others,
> >
> > What is the status of property handling?
> >
> > Are we going to go with Peter's suggestion and commit the java source
> > to cvs and work on handling properties fully from there?
> >
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
> For additional commands, email: fop-dev-help@xml.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org