On 28.08.1998 15:02:26 goccebbs@dimtel.nti.it (Giuseppe Chillemi) wrote:

>  I have these lines in one script
> 
> DBC_Cliente.CODPROD.count = x
> DBC_Cliente.PROD_NAME.count = x
> DBC_Cliente.PROD_PZ.count = x
> DBC_Cliente.PROD_IVA.count = x
> DBC_Cliente.PROD_UM.count = x
> DBC_Cliente.PROD_SC.count = x
> 
> ADDRESS 'FIASCO.1'
> 
> SETFIELD 'CreateListEntries' 'STEM' '"DBC_Cliente"'
> 
>  Where DBC_Cliente. are connected SELECTONLY listviews.
> 
>  x = 0 and DBC_Cliente. are all uninitialized becouse I want no lines
> in the listviews.
> 
>  When the SETFIELD command is executed Fiasco does not recognize that the
> database has changed thus if I select LOAD in the menu no warn requester
> appears.

Okay, that bug is simple: When no new listview entries are created,
the SetField command misses to mark the field and database changed.

>  I could say the problem is all here but becouse I like to let you be
> desperate in front of my bug reports I have experimented with the script a
> lot !  :-)
>  There are many problems in the LV implementation..
> 
> 
>  Using the previos script, the command > '"DBC_Cliente"'> where DBC_Cliente..COUNT is 0, a LINE in then
> empty LVs is always created.
>  (Verified executing a Getfield '""' 'listentrycount' before and
> after the SETFIELD command)
> 
>  Obviusly I have the same result when x (.COUNT) is 1 instead of 0.

Okay, that's also a bug: SetField Stem does not use COUNT at all,
but only relies on the initialized ARexx variables. I.e. if you
have a stem.1 initialized, SetField will currently always possibly
create an entry and set it to stem.1

>  Then I have experienced these other workings:
> 
>  If I set .count of all the LV to 1 and I initialize the respective stems
> line to something and the SETFIELD command is executed using these data,
>  returns 7 entries for the first LV and 1 for
> the others.  There are 7 selectable lines in the LV (but empty).
> 
>  If I set .count of all the LV to 1 but I set only the stem number 1 of the
> first LV and do the SETFIELD,  returns 7 on the
> first LV and 0 in all the others (and 7 lines are selectable).
>
>  In the same situation, If I set only the first line of the second stem,
> its stem returns 2 lines and 1 in the others.

These problems all seem to come from the bug described above.

>  When I manually select an item in the LV, it is then updated and fiasco
> reveals to have set the second line of the LV but not the first as
> requested.
> 
> 
>  A little confusing, isn't it ?
> 
> 
>  However, I have reread the documentation.  It explains that
> 'createlistentries' replaces all the elements of the LV with new ones, so I
> assume the correct working should be this:
> 
>  SetField 'CreateListEntries' 'stem' where 'STEM..count = 0
> should delete the entire LV.

Yes, it should. It does when stem..1 is not initialized.

>  SetField 'CreateListEntries' 'stem' where
> 
>  STEM..n (n = numentry) is an uninitialized element, fiasco should
> initialize it to the defaul value set for that field.

Currently, these are ignored, the old contents remain the same.

>  When we have connected LV, the number of entries of the other LVs should be
> adjusted to match the .COUNT value of the stem you are SETFIELDing so, if
> they have more entries, they should be deleted starting from the last, if
> they have less entries, new ones must be added and their content shoul be
> the default value the user has set for each field.

I will probably add an argument that handles grouped listviews.

> PS: A "DeleteAllLVEntries" command maybe good idea.
> PPS:  It took me the entire morning to discover this problem !!!  You
>    should pay me something to be one of your betatester :-)

In another mail, you just wrote that you want to pay me when
you have a new feature, summed up, this becomes zero. ;-)

Bye

-- 
Nils Bandener          ///   Now completely updated:
nils@dinoex.sub.org  \XX/    http://www.amigaworld.com/support/fiasco/
nilsb@amigaworld.com