Re: [Devel] FT 2.1.9: Minor fixits/Consolidated make procedure for OpenVMS

[Werner LEMBERG]

> I checked out 2.1.9 recently and compiled it on OpenVMS. There are
> only very few minor compile messages.
> 
>     if ( family && --family->num_nodes <= 0 )
> ...................^
> %CC-I-QUESTCOMPARE, In this statement, the unsigned expression
> "--family->num_no des" is being compared with a relational operator
> to a constant whose value is n ot greater than zero.  This might not
> be what you intended.
> 
> --> Explicit cast to int

Thanks.  Interestingly, the compiler apparently hasn't found the
second place (at the end of ftcglyph.c) where exactly the same problem
exists.  BTW, the proper solution is to make an explicit cast to
`long' -- I wonder whether it would be even more consistent to change
the type of `num_nodes' to FT_Int.  David?

>     if ( ( error = FT_NEW_ARRAY( delta_xy, n_points ) ) )

This has already been fixed in the CVS.

> I did further tweaks to the vms_make.com toplevel make procedure.
> With this version all of the individual descrip.mms files in the
> subdirectories are obsolete and can be removed from the source tree,
> they will be created on the fly by vms_make.com if building on a VMS
> system.

Excellent!  A minor issue: Do you also handle the `otvalid' directory?
Otherwise it might be a good idea to note in INSTALL.VMS which modules
aren't handled automatically.

I've put your stuff into the CVS.  Please check.


    Werner