forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ross Gardler <rgard...@apache.org>
Subject Re: Refactor sitemaps for XHTML2 and the LM (Re: Forrest Tuesday)
Date Mon, 29 Aug 2005 19:33:07 GMT
Tim Williams wrote:
> On 8/29/05, Ross Gardler <rgardler@apache.org> wrote:
> 
>>Ferdinand Soethe wrote:
>>
>>>Ross Gardler wrote:
>>>
>>>
>>>
>>>>The *implementation* of the move to XHTML2 would be eased considerably
>>>>by refactoring our sitemaps to use the locationmap.
>>>
>>>
>>>Why is that? In my limited understanding lm allows to specify the
>>>location of a resource in more sophisticated ways, where is the
>>>connection?
>>
>>Right now the sitemaps are a complex web of possible locations for many
>>resources, with extensive tests for the existence of a file. These are
>>duplicated across many sitemaps. The LM will remove all of that by
>>having a central file describing the location of resources.
>>
>>That is, there will be only one place to edit as we replace chunks of
>>the sitemap functionality.
>>
>>
>>>>Tim reports that the
>>>>project locationmap mount doesn't work properly yet, but unless he tells
>>>>us otherwise I think we can safely move forward with this refactoring.
>>>
>>>
>>>Not sure this kind of lazy consensus is the right approach here. I
>>>understand is that Tim has pointed out a problem with the lm
>>>implementation. Should it not be up to us now to demo that this
>>>problem has been solved (us of course including Tim :-) rather than
>>>expecting Tim to object?
>>
>>No, you misunderstand the issue Tim has identified.
>>
>>As Locationmaps currently stand there is only one locationmap file. Tim
>>is working on allowing that file to import a project locationmap as
>>well. The project locationmap will be able to override the default
>>locationmap, thus users can customise Forrests directory layout without
>>having to redefine the whole thing.
>>
>>Tim is having a problem with this *extension* to the locationmap not
>>with the locationmap itself, which works perfectly.
> 
> 
> This is correct, it does *not* effect the refactoring work.  Anyway, I
> hope to have this resolved within the next day or so anyway so that we
> can mount project-level locationmaps *if* they exist.

Thanks to Ferdinand for highlighting this potential oversight and Tim 
for confirming my thoughts were correct. If I'd been incorrect this 
could have made a mess ;-)

Ross

Mime
View raw message