eric@pyramid.pyramid.com (Eric Bergan) (04/22/89)
In article <66983@pyramid.pyramid.com> csg@pyramid.pyramid.com (Carl S. Gutekunst) writes: >That's all anyone gets. The letter denotes a "checkpoint" release, which is >done when a change occurs that cannot reasonably be done in a PTF, but we are >not ready for a full release. Normally all PTFs will also be folded into the >checkpoint, too. When a checkpoint release is done, the previous release (as >in 4.4 or 4.4b) is no longer shipped. But, like a PTF, the checkpoint release >is not "announced" to the existing customers the way a full release is. You >have to ask for it. The philosophy of a checkpoint release is a little more than what Carl has stated. The primary purpose of a checkpoint release is to generate a stable, tested release that has the PTFs integrated in. PTFs are generated to fix problems reported from the field, and are tested to make sure they fix the problem and don't break anything obvious, but they are bug fixes, not real releases. A checkpoint release takes all the PTFs up to that point, and generates a release that then goes through the full QA procedure, including stress and regression testing. That way, we can document that no contention has slipped in between different PTFs, or between a PTF and the previous base system. A checkpoint release can also be used to integrate in changes necessary for new (or revised) optional products. The intent is to generate checkpoint releases on a reasonably fixed basis, about every three to six months, based on the the number and severity of the PTFs that have been accumulated. eric -- eric ...!pyramid!eric
bgg@yarra.oz.au (Benjamin G. Golding) (04/26/89)
In article <677@Terra.cc.brunel.ac.uk>, linda@cc.brunel.ac.uk (Linda Birmingham) writes: > 4) What state is Fortran 77 in ? I've been supporting the Pyramid f77 compiler for almost a year now and it's looking good, IMHO. There is one outstanding priority 3 bug related to an esoteric anomaly in the microcode; it's being worked on at the moment. All other outstanding SPR's are priority 4 enhancement requests. There are a number of extensions in recent releases of the compiler, for example unlabelled DO and DO-WHILE loops, etc (contact me for details). The code generator is producing better code too. Ben Golding <bgg@yarra.oz.au> Pyramid Technology Corporation (Australia).