Re: [Oorexx-devel] Problem With SHVBLOCK Definition

2008-10-09 Thread Rick McGuire
David, I need to rethink my position on this. The thought behind making the name a CONSTRXSTRING was that it was a read-only part of the interface. In other words, by declaring this as CONST, the API was declaring a contract that it was not going to be altering the storage referenced by that RXS

Re: [Oorexx-devel] Problem With SHVBLOCK Definition

2008-10-08 Thread Rick McGuire
No, not true. Just a simple case to a (void *) is sufficient. See the equivalent code in the rexxutil SysDumpVariables code. I'm not sure I want to give up the advantages that are gained from making that a const value. Rick On Wed, Oct 8, 2008 at 2:25 PM, David Ashley <[EMAIL PROTECTED]> wrote

[Oorexx-devel] Problem With SHVBLOCK Definition

2008-10-08 Thread David Ashley
Rick - Before I fixed this I wanted to run this by you to see if you had ideas on a work around. Currently the SHVBLOCK defines the shvname RXSYTING struct as a const. This is fine for all operations except RXSHV_NEXTV. In this case ooRexx can allocate the memory for shvname.strptr. The proble