andrewn@syma.sussex.ac.uk (Andrew D Nimmo) (08/29/90)
While checking out a problem for someone, I discovered that /usr/lib/crt0.o (and gcrt0.o and mcrt0.o) had the following makers - bind mcrt0.o -mak This object was made by the following: ASM, Rev 9.10, Date: 1988/05/09 18:06:14 BST (Mon) C compiler 68K Rev 6.7(232) - BETA TEST, Rev 0.00, Date: 1989/06/21 15:02:00 BST (Wed) ld, Rev 10.01, Date: 1989/05/10 20:27:26 BST (Wed) Should it concern us that a BETA TEST C compiler was used, and (more importantly) has this compiler been used for other system executables? Andrew
thompson@PAN.SSEC.HONEYWELL.COM (John Thompson) (08/30/90)
> While checking out a problem for someone, I discovered that > /usr/lib/crt0.o (and gcrt0.o and mcrt0.o) had the following makers - > > bind mcrt0.o -mak > > This object was made by the following: > ASM, Rev 9.10, Date: 1988/05/09 18:06:14 BST (Mon) > C compiler 68K Rev 6.7(232) - BETA TEST, Rev 0.00, Date: 1989/06/21 > 15:02:00 BST (Wed) > ld, Rev 10.01, Date: 1989/05/10 20:27:26 BST (Wed) > > > Should it concern us that a BETA TEST C compiler was used, and (more > importantly) has this compiler been used for other system executables? I wouldn't be bothered too much by it. In theory, beta software has been tested out well enough that you want an outside source to abuse it. This seems to imply (IMHO) that the developers are comfortable with it. Additionally, you run into a chicken-and-egg problem eventually. Something needed to be made with a beta release at some point in the past. Should we trust it? What about anything that's been made after that? To answer your second question, I'd assume that other stuff was made with beta software. What _I_ found was stuff made with ALPHA compilers! Specifically, the /lib/clib (ts 1989/10/05 14:46:55 CDT) was built with $ bind //gollum/lib/clib -mak This object was made by the following: bind, Rev 6.10, Date: 1989/03/06 18:36:30 CDT (Mon) ASM, Rev 9.10, Date: 1989/07/21 10:17:09 CDT (Fri) Pascal compiler PRISM=>68K Rev 8.7(226) - ALPHA TEST, Rev 0.00, Date: 2015/09/05 0:58:26 CDT (Sat) Pascal compiler 68K Rev 8.7(127) - ALPHA TEST, Rev 0.00, Date: 1989/05/09 13:06:29 CDT (Tue) C compiler PRISM=>68K Rev 6.7(301), Rev 0.00, Date: 2015/09/05 0:58:26 CDT (Sat) C compiler 68K Rev 6.7(299), Rev 0.00, Date: 1989/08/08 13:11:00 CDT (Tue) John Thompson (jt) Honeywell, SSEC Plymouth, MN 55441 thompson@pan.ssec.honeywell.com As ever, my opinions do not necessarily agree with Honeywell's or reality's. (Honeywell's do not necessarily agree with mine or reality's, either)
derstad@CIM-VAX.HONEYWELL.COM ("DAVE ERSTAD") (08/31/90)
>Should it concern us that a BETA TEST C compiler was used...
We found some things the other day that were compiled with ALPHA
versions of the compilers!rehrauer@apollo.HP.COM (Steve Rehrauer) (09/01/90)
In article <9008311320.AA02135@umix.cc.umich.edu> derstad@CIM-VAX.HONEYWELL.COM ("DAVE ERSTAD") writes: >>Should it concern us that a BETA TEST C compiler was used... > >We found some things the other day that were compiled with ALPHA >versions of the compilers! I wouldn't be too concerned, though I'll admit it does seem odd at first glance. Internal users of our compilers, e.g.: the OS folks, tend to be pretty conservative about consuming what we (meaning us compiler folks) produce. With good reason, for we often (shhhh! don't tell anyone! :) break things in our normal, workaday mode of fixing other bugs and adding enhancements. Usually such regressions are caught quickly, but sometimes not -- I don't build & boot an OS after every single compiler change, for example, so it may be several weeks before I catch a problem that manifests itself in the OS but not any other code in our test suites. So if, say, a Beta compiler is "good enough" for a given application, the internal users may not grab an updated copy for some time on the premise of "better the bugs you know". After all, most "known" bugs will manifest themselves only in a particular situation, and if your source code doesn't "trip the wire"... Certainly no one here is building and shipping applications cavalierly, using compilers they know to be unstable with their code. -- >>"Aaiiyeeee! Death from above!"<< | (Steve) rehrauer@apollo.hp.com "Spontaneous human combustion - what luck!"| Apollo Computer (Hewlett-Packard)