Re: Web development

[MJ Ray]

> Well, before answering your questions, let me fill you in on the
> details of HB.

Hey, I do actually read the docs, you know ;-)  Seriously, examples are
always useful.

> There is another way to add objects such as "<+hello>", "<+tail>" and
> "<+counter>", which is defining them in the HB modules.  [...]

How does it know to load a module?

> Unfortunately, at this moment it is not possible to create modules in
> Scheme, only in C.  However, making it possible to create them in
> Guile should not be very hard.  Once this is done, you could have a
> .scm file with a function for each of the above methods (clean, load
> and unload are optional)...

And, presumably, a call or data structure of some sort to let you know which
functions are which, ie (define-hb-module '((init mymodule-init-fn)))

>     can we preload guile modules to be called from in the pages?  Can
>     we load code that runs once per server instance, rather than once
>     per request?
> Not at this moment.  At this moment all you can do is bind certain
> objects to Guile code.  [...]

At the moment, what does it have for guile over CGI other than a templating
mechanism, then?

Supplementary question: is the parsing recursive, ie can the guile output
include other modules eg (display "<+hello>")?

> Once it becomes possible to create modules in Scheme, the answer will,
> obviously, be yes.

Then things get very interesting.

>       (hb-argument hbargs "width")

What happens if there is no width argument?  Have you considered using 
<?hb...?> tags, which I think may be syntactically cleaner?

> By the way, most of the functionality in HB can be left out of your
> builds (gets #define'd out) passing appropriate parameters to the
> configure script... you could disable everything from your build
> except the Guile-specific functionality.

No, the interesting thing about HB is the ability to support a multitude of
languages at the same time in one framework.  What sort of performance
figures do you see?

>     I'd suggesting running squid in front of its httpd instead of
>     apache, unless you need other apache features.
> This is an interesting suggestion.  :)

It's an essential one, as apache doesn't seem to cache ProxyPass proxies,
which means you are still called for every page request.  Doing unnecessary
work is never a good idea ;-)

-- 
MJR