unclear on the concept, or major report printing bugs?

[Graham]

On Thu, 24 Jul 2003 02:04:06 -0700 (PDT), Jan Schenkel 
<janschenkel at yahoo.com> wrote:


>First of all, I couldn't agree more : a good demo
>would go a long way towards opening it up ; the good
>people at Revolution HQ insist that all the parts are
>there to build wonderful reports ; but the absence of
>a sample stack leaves us all scratching our heads.
[...]

This makes reports, IMHO, pretty much the same as tables - quite possibly 
great **if** you could understand them! I think a key point about these 
additions to the system (Richard G has a term for them but I've forgotten 
what it is - basically features halfway between real new language elements 
and built-in procedures) is that on first encountering them, it's hard to 
work out what the intention was in creating them, and therefore it's hard 
to manage one's expectation of the feature.

For instance, when I read that tables were 'spreadsheet-like', I imagined 
that the user of a RR-developed app would be able to key in various types 
of data into individual cells, and that the script of the app would be able 
to know when this happened and then do spreadsheet-like things, such as 
data checking, evaluation, formula adjustment etc. But it seems it doesn't 
really work that way, and perhaps **was never intended to**. Similarly with 
reports, I imagined functionality which would give complete control over 
page layout (including multiple pages) including stuff like margins, 
headers, footers, page breaks, widowed lines... but again maybe this was 
not really the intention. I guess we now await a collection of examples 
from RunRev which will somehow incorporate the philosophy of these 
additions as well as the actual nuts and bolts. Then we can all adjust our 
expectations. Meanwhile I have to say I'm steering clear of both tables and 
reports.

My two Euro-cents

Graham

---------------------------------------------------
Graham Samuel / The Living Fossil Co. / UK & France