Unix shell design [was Re: Yes, Unix can do batch renames - Re: Unix handles file extensions "very poorly"?]

Some of it was even correct! Okay, that's unfair. DEC documentation was actually quite good, at least back in the '80s when I worked with DEC-produced doc. Sometimes it even called out where things were specifically undefined, which I wish more interface specs these days did. But I was responsible for our lab generating something like five SPRs (that's what DEC called bug reports in those days) and the only one we ever got any response at all to was one which was my own mistake. As an example of one of the others, I created an extensive patch to add line editing to DCL. It broke PATCH; I had to split it into two patches, and linking them together was..awkward. We submitted an SPR about this undocumented limit in PATCH. (I could patch DCL because we had source...on microfiche. I spent about a week poring over the 'fiche and constructing that patch.) Never heard a word back about that bug. Not even a FITNR or a "this is actually a doc bug, the limit should be documented". /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B