Re: Naive DEFSYSTEM replacement

On Mon, Jul 28, 2025 at 1:39 AM Adam Weaver <[email protected]> wrote:

> On Sun, 2025-07-27 at 15:13 +0200, Marco Antoniotti wrote:
> > >
> > > Well it generates what I think it should ...
> > >
> > > CL-USER> (riskmate::>box)
> > > "<div class=\"box\"></div>"
> > > CL-USER>
> > >
> > > ... what did *you* think it would generate?
> > >
> > >
> >
> >
> > A proper object with a proper and well controlled PRINT-OBJECT
> > method 😏 (controlled by the printer's parameters and others as
> > well).
> >
>
> Thanks Marco. That makes sense, as long as I'm right in guessing that
> you're using XHTML's output for later frobbing?
>

Yep.

Me, I'm only generating HTML just-in-time for returning via FastCGI
> responder to the web server to the end browser.
>
> Originally I wrote a HTTP/1.1 web server, but when I realised how much
> work it'd be to upgrade to HTTP2 and HTTP3, I decided to write a
> FastCGI responder instead, and use an established web server (nginx or
> caddy for example), and let that one handle SSL and the newer HTTP
> variants.
>
> Yet another reason to love LW, COMM:START-UP-SERVER makes writing
> internet services *just*so*easy* !
>
>
> > CL-USER 6 > (<:div (:class "box" :id 42) (<:p () "Foo")) ; You can
> > write whatever functions you want (*)
> > <div class="box" id="42"><p>Foo</p></div>
> >
> > CL-USER 7 > (describe *)
> >
> > <div class="box" id="42"><p>Foo</p></div> is an
> > IT.UNIMIB.DISCO.MA.XHTMLAMBDA::ELEMENT
> > TAG             IT.UNIMIB.DISCO.MA.XHTMLAMBDA:DIV
> > ATTRIBUTES      ("class" "box" "id" 42)
> > CONTENT         (<p>Foo</p>)
> >
> > CL-USER 8 >
> >
>
> So I'm guessing the value here to you, is that you can pass your
> XHTMLAMBDA::ELEMENT object to other functions for munging in some way?
>

Yep again.

I get that, and it's a cool idea, if that's a thing you do. Like if you
> wrote a cold-start documentation generator for example? Just an idea :)
>
> I chose differently, because I'm just using HTML only as the "final
> mile" - what's being delivered to the browser, and not as an
> intermediate storage of state.
>
> No point in using an intermediate object model, because I'd only be
> throwing it away instantly as I generated string HTML.
>
> *Functions* (procedures) are the engines of state, here, so my client
> code creates HTML output by calling functions to create the HTML
> interfaces.
>
> Like this sort of crap ...
>
> (defun >client-contacts-card-page (client)
>   (>column :gap t :content
>            (html
>              (:div (:id "contacts") (>client-info-card client))
>              (:div (:id "notes") (>client-notes-card client))
>              (:div (:id "policies") (>client-policies-card client)))))
>
> ...
>
> >
> >
> > (*) Now for the second shameless plug of the day: here is where
> > (X)HTM&Lambda; is used: https://helambdap.sf.net
>
> That's pretty cool. Reads like you wanted that code walker too, what
> with the READing of unloaded and uninterned symbols.


READing is a problem in fact; things that happen when reader errors
(unknown package) are not standardized.
I have been toying with the idea of using Eclector (
https://github.com/s-expressionists/Eclector), but did not have time.
And yes.  I wanted CLAST (https://clast.sf.net), but for a different reason
:)  It's because I am CLAZY (https://git.common-lisp.net/clazy/clazy) 😏

Cheers

MA








-- 
Marco Antoniotti, Professor, Director         tel. +39 - 02 64 48 79 01
DISCo, University of Milan-Bicocca U14 2043   http://dcb.disco.unimib.it
Viale Sarca 336
I-20126 Milan (MI) ITALY

REGAINS: https://regains.disco.unimib.it/