[reportlab-users] Widgets as Flowables...

[Robin Becker]

In article <LKENLBBMDHMKBECHIAIAEENCCLAA.andy@reportlab.com>, Andy
Robinson <andy@reportlab.com> writes
.......
>If people want to click on bars in a bar chart or data point
>(e.g. HTML image map, or GUI), then they need access to the
>bounding box info AFTER the draw method; and they may need
>extra info like tooltips and so on which had to "bubble up"
>from the innermost objects.  So I think it has to happen
>at draw time or, optionall, after draw time.
>
>OTOH, if we're doing 50 detailed plots inside a Platypus
>paper for someone's thesis; and it has a table of contents
>and may need 3 passes rendering, all Platypus needs to
>know is the size of the enclosing Drawing.  It COULD hold
>off drawing the things until the last pass when everything
>is laid out.  That's one reason I like Drawings continuin
>to exist in some fashion.
>
>- Andy
.......
I don't disagree, but explaining this is hard especially for the more
dynamic things. I tried really hard to make a drawing that based itself
on its top and found it virtually impossible; it seems the simple
concept of a drawing size which depends on its contents is hard in our
scheme. 
-- 
Robin Becker