Gil Barmwater

If anyone is interested in creating/porting Rexx to this platform, I would suggest that they have a look at a project called "Compact Rexx" which was done by Brian Marks "in the years around 1999 to validate the ANSI Rexx Standard and its Errata." Details can be found at the Rexx Repository. "It is a 40000 byte CRX.EXE..." so a small footprint but a challenging port not for the faint-hearted.

Thanks for these changes. I had not forseen that the tools would be invoked from other programs so I had them issue a (hopefully) useful error message and exit when they encountered an error that prevented them from continuing. As for the Liberation Fonts - I'm not surprised that location on the Internet has become unavailable. Dead links have always been a problem as the Web is constantly changing. The "copy" on SF was placed there by PO when we were getting the document build process working for...

Thanks for these changes. I had not forseen that the tools would be invoked from other programs so I had them issue a (hopefully) useful error message and exit whenthey encountered an error that prevented them from continuing.

Have you looked at the Raise keyword? It might give you what you want (or maybe not :-( ).

Just a few thoughts on this proposal. I think that what threw me off initially was the choice for the operator - '/'. I'm a Windows guy so it struck me as a bit confusing at first. I think that perhaps a better choice might be '+' as it suggests that one is "adding" an element - folder or file name - to the path defining the .File object. With that in mind I have put together the following bit of code that you can use now pending the (possible) implementation of this feature request. Some things...

Thanks for that! That code has been there a long time so I guess no one ever tried it on a non-Windows system.

Thanks to Rick, we know that this behavior was introduced as part of RFE 777. There does not seem to be any documented request for this feature, only that MS recomended it. We also know that there are several examples of "bad" behavior resulting from its introduction. So, I agree that adding a function that would allow DISABLE_NEWLINE_AUTO_RETURN to be turned on/off seems to be needed. The remaining question is what the default/startup behavior should be. Whichever one is chosen, the potential for...

It appears this was introduced with 5.0.0 as I cannot reproduce it with 4.2.0. I would be interested to learn if there was a use-case that inititiated that change.