rer@hpfcdc.HP.COM (Rob Robason) (07/19/89)
I am involved in an investigation into providing better portability of 4.3 BSD applications to HP-UX by providing all non-obsolete 4.3 BSD libc, libm and libmp routines plus header file compatibility. The intent would be to provide an environment which would provide more straightforward applications porting for the vast majority of 4.3 applications (hardware dependencies excepted). I am able to use 4.3 commands as one set of reference source code to test this environment, but would also be interested in soliciting user applications which would benefit from these features. I would not be able to guarantee at this time that I would successfully port these applications nor that my investigation would result in a product. I am looking specifically for code which is fairly self-contained, is hardware independent, and has presented porting problems in HP-UX (though not all these attributes are necessary). If you have an application that needs some unique 4.3 BSD capability, or set of capabilities, and would be interested in participating by providing the source and associated data for your application, please respond to me directly by email with the following information: your name and organization U.S. Mail address including mailstop email address phone number brief description of BSD features required by your application have you attempted to port the application to HP-UX using libBSD.a would you require a non-disclosure agreement for the source Please do not send source now. I will notify you if your application is needed. If you are a BSD advocate, this is an excellent opportunity to help HP move HP-UX toward being a more Berkeley like system. As you know, we have made other efforts in the past by adding the McKusick file system with long file names and symbolic links, Berkeley Job Control, ARPA/Berkeley networking services, vi, csh, more and many other BSD features. The fact that HP-UX is based on the BSD kernel makes it possible for us to consider adding even more BSD functionality, in addition to our full SVID compliance. Please respond to: rer@hpfclg.hp.com or hplabs!hpfcla!rer or write Hewlett-Packard Co. Attn: Rob Robason M/S 99 3404 E. Harmony Rd. Fort Collins, CO 80525 or call Rob Robason (303)229-3763 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! This message does not constitute, in any way, a commitment on the part of Hewlett-Packard Co., or any other party, to provide the features or capabilities described. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
paul@mecazh.UUCP (Paul Breslaw) (07/25/89)
rer@hpfcdc.HP.COM (Rob Robason) writes: >I am involved in an investigation into providing better portability of >4.3 BSD applications to HP-UX >If you are a BSD advocate, this is an excellent opportunity to help HP >move HP-UX toward being a more Berkeley like system. >The fact that HP-UX is based on the BSD kernel makes it >possible for us to consider adding even more BSD functionality, in >addition to our full SVID compliance. So why don't you be terribly Berkeley-like and put *ALL* the UPM documentation in one place. You know, so that I don't have to look for a folder with a title such as:- ARPA Programming and Network Services (NFS) For Berkeley and Other Things We Borrowed From Other People References/9000 when all I want is to check the type of the 4th parameter of recv(2). How are poor novices supposed to know that socket(2) was not invented by AT&T, and therefore is to be found in a different part of the documentation. Why does login(1) have a manual entry that mentions login -r rhost and /etc/hosts.equiv, but I have to be a clairvoyent to guess that all the real information is to found in ARPA/Berkeley Services References Pages? Paul Breslaw Mecasoft SA Guggachstrasse 10 8057 Zurich Switzerland tel: +41 1 362 2040 e-mail: ...mcvax!cernvax!mecazh!paul
davidb@Pacer.UUCP (David Barts) (07/26/89)
In article <350@node17.mecazh.UUCP>, paul@mecazh.UUCP (Paul Breslaw) writes: > rer@hpfcdc.HP.COM (Rob Robason) writes: > . . . > and /etc/hosts.equiv, but I have to be a clairvoyent to guess > that all the real information is to found in ARPA/Berkeley Services > References Pages? > How about having a Master Index for the entire documentation set? More new ideas: - How about including a true set of termcap(3) emulation routines. (The ones from ATT are a joke.) - Troff(1) (and postprocessor) or ditroff(1) with LaserJet support. - The Berkeley -me {nt}roff macro package. (This may require shipping two versions of {nt}roff, as I think -me and -mm are written for different versions.) - Most important of all, the games ;-) David Barts Pacer Corporation davidb@pacer.uucp ...!fluke!pacer!davidb
burzio@mmlai.UUCP (Anthony Burzio) (07/29/89)
In article <240@zircon.UUCP>, davidb@Pacer.UUCP (David Barts) writes: > - The Berkeley -me {nt}roff macro package. (This may require > shipping two versions of {nt}roff, as I think -me and -mm > are written for different versions.) The SUN me files work fine. I just ln -s'ed them over from our SUN computers. > - Most important of all, the games ;-) Actually, games are a great way to test softare portability. The differences I have noticed are the random number generators and the timing routines. A couple other games complain about undefined constants out of the IO drivers. Compile a bunch of the games from the archives and you'll find all the BSD-SYSV differences very quickly :-) ********************************************************************* Tony Burzio * HP-UX needs some games... Martin Marietta Labs * mmlai!burzio@uunet.uu.net * *********************************************************************