Compatibility of RCODE between OE11.6 and OE11.7

Posted by MBeynon on 26-Nov-2019 13:35

Hi,

Doing some research on the above and found this in the KB;

knowledgebase.progress.com/.../20922

3.  R-code Version Change

A recompile will always be required when moving to a different Major OpenEdge release as the r-code version always changes between major versions. 
Between Minor OpenEdge releases, forward-compatibility is guaranteed for r-code, but not backwards-compatibility. Code compiled in a lower minor release will run in a higher minor release; the other way around is not guaranteed.
Between Minor versions, the r-code version number may remain the same, but using new ABL features added in a later minor release will break compatibility with earlier minor versions.

Is this correct? Backwards compatibility is not guaranteed between minor RCODE versions?

Many Thanks,

Mark.

Posted by 2087 on 26-Nov-2019 13:47

Correct.

I have experienced this.

I accidentally compiled in 11.3 and some of the r-code would not run on 11.2.

All Replies

Posted by 2087 on 26-Nov-2019 13:47

Correct.

I have experienced this.

I accidentally compiled in 11.3 and some of the r-code would not run on 11.2.

Posted by frank.meulblok on 26-Nov-2019 15:29

Backwards compatibility means the newer minor release can run r-code compiled under the older minor release.

This does work for the most part, although class-based ABL code in libraries supplied with OpenEdge can cause hiccups. And I think there may have been some data-server related bugs that required breaking compatibiltiy to fix bigger issues, but my memory may be failing there. Never did do much with that particular bit of tech.

Compiling in a newer release and running on an older one implies forward compatibiltiy.

That is not always guaranteed, as the newer code can use AVM-level features that the old release simply did not have yet. And there may be more or less obvious issues as a result. Sometimes it outright errors (which IMHO is the better outcome), sometimes parts of the code may just be skipped and general weirdness ensues instead (knowledgebase.progress.com/.../P175091 was a good example of that).

Posted by gus bjorklund on 26-Nov-2019 16:04

This thread is closed