cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nicola Ken Barozzi <>
Subject Re: [RT] Flowmaps
Date Tue, 18 Jun 2002 07:33:24 GMT

Stefano Mazzocchi wrote:
> Vadim Gritsenko wrote:
>>I also strongly hope that we can eliminate this extra matcher and won't
>>trip your alarms :)
> Let's see :)
>>I hear your fears. Let me make second try on this. This snippet below
>>will allow sitemap designer to choose what method of continuation
>>passing he wants to use. You can use either URI matching, or Cookie
>>matching, or parameter matching. Here, continuation obtained from the
>><map:match pattern="calc/*">
>>  <map:flow method="calculator" continuation="{1}" />
>>This means: start flow if no continuation is provided or continue flow
>>if continuation is present.
>>We can change wording (map:flow -> map:whatever), can change number of
>>arguments, etc, but my feelings is that it will be easier to use and
>>understand if flow requires only one sitemap "operator", but not two as
>>of now: map:continue and map:call.
> I hear no alarm ringing, that's a good thing :)

I like it too :-)

One more thing, last night I was thinking of what you said, that 
flowmaps should be a new type of Actions, not a component on the same 
level of a sitemap.

Though conceptually the sitemap-flowmap duality is appealing and seems 
more clean, I do see the problem: sitemaps that mount flowmaps that call 
sitemaps that mount flowmaps that get it all in an *undebuggable* mess.

I think that we could even change the name "flowmap", that is too 
similar to "sitemap". It's not a "map".

It could be quite easy to explain that a "flow" is a new type of action 
that keeps "special session=continuations" between requests.

>>I agree with Nicola's argument: users might get confused if call is
>>loaded with one more meaning.
>>Or might be not...
> What do others think about this?

Nicola Ken Barozzi         
             - verba volant, scripta manent -
    (discussions get forgotten, just code remains)

To unsubscribe, e-mail:
For additional commands, email:

View raw message