jbuck@epimass.EPI.COM (Joe Buck) (03/07/89)
Hello. Our company has a 68000-based product that was developed using the HP64000 system for emulation, compiling and assembling code, etc. The tools as provided by HP run on the 64000 as well as VAX/VMS on a 750. We did most of the development work under VMS with Eunice so we'd have make, SCCS, etc. There is an HP-proprietary transfer program that moves files back and forth between the VAX and the HP6400 system. My management wants to get rid of the VAX, and frankly, so do I. The problem, though, is continuing to support the code. There are several possibilities: 1. Junk the HP64000 (it's really ancient stuff) and go to a new type of development system. Problem: our product code is about 2/3 C and 1/3 assembler. If the assembler doesn't use the same syntax (there are quite a few different 68000 assembler syntaxes in use) or the calling convention for C calls isn't the same, there's a lot of conversion work. I could buy HP's whizzy new latest-generation emulator systems, at a cost of $40K per workstation or so (I haven't investigated this option in detail). But we are a small, cheap company. 2. Do our builds on the HP64000 itself, rather than on the VMS system. Problem: no decent editors, no SCCS, no make. One possibility is to do editing and SCCS on a Unix system, and transfer source files over to the 64000 for building. Problem: the HP transfer program is proprietary. 3. Get some kind of PC-based emulation system. We have PCs on an Ethernet with PC-NFS already in place. We have the problem of converting the code. It occurs to me that lots of people once used HP64000 systems, and also that such systems are pretty much obsolete. I'm looking for ideas and experiences other people have had. How did you handle this? Please respond by mail (unless you think a lot of people would benefit by a posting), and I'll summarize what I learn. Thanks in advance. -- -- Joe Buck jbuck@epimass.epi.com, uunet!epimass.epi.com!jbuck
algoss@hpubmaa.HP.COM (Al Gosselin) (03/08/89)
I am a field System Engineer for the 64000 line, so I have a vested interest in the direction you go. The 64000-UX system has all of the things you need, and the prices have come down drastically over the last few years. We have a new language system we call AxLS. This includes an ANSI 'C' compiler which is totally new and much better than the old compiler. There is a new assembler which is compatible with the Motorola standard. Shipped with these products are some conversion programs to ease the transition. These include both 'C' and assembler converters as well as some tools to allow you to call exis- ting assembly language code from the new compiler. If you buy the 9000 computer just for S/W development, you can use the High-Speed link to move the absolute to the 64000 system. You can also convert the 64000 stations to Instrumentation Card Cages for direct use with the 9000. This allows emulation from your HP-UX terminal or windows or whatever with neat fetures like the softkeys and "C" source code in your trace listings. Don't write off the 64000 just because it's old! It's still selling well and we've developed a good upgrade to HP-UX. Call your local Sales Rep. or me if you have questions! Algoss (203) 659-6067
raveling@vaxb.isi.edu (Paul Raveling) (03/08/89)
In article <2937@epimass.EPI.COM> jbuck@epimass.EPI.COM (Joe Buck) writes: > >My management wants to get rid of the VAX, and frankly, so do I. The >problem, though, is continuing to support the code. There are several >possibilities: > >1. Junk the HP64000 (it's really ancient stuff) and go to a new type of >development system. ... One alternative you might want to check is Kontron Electronics development system products. I'm haven't been in touch with their product line for several years, but would expect that they have a descendent of the FutureData Slave Emulator product line that would either communicate with a host or plug into an IBM PC or equivalent. Back when I was the MESS (Manager of Emulation Systems Software) at FutureData, the HP 64000's were our nearest competitor in terms of technical merit on 68000 emulators. Both systems were generally similar, with similar tools available for cross-compilation and cross-assembly. FutureData had an edge in the user interface for debugging via the emulator and in several emulation capabilities. Kontron bought FutureData and seems to be keeping a fairly low profile in this country, though I understand they're thriving in Europe. It's probably worth checking on what they offer now. ---------------- Paul Raveling Raveling@isi.edu
cramer@optilink.UUCP (Clayton Cramer) (03/09/89)
In article <7712@venera.isi.edu., raveling@vaxb.isi.edu (Paul Raveling) writes: . In article <2937@epimass.EPI.COM. jbuck@epimass.EPI.COM (Joe Buck) writes: . . . .My management wants to get rid of the VAX, and frankly, so do I. The . .problem, though, is continuing to support the code. There are several . .possibilities: . . . .1. Junk the HP64000 (it's really ancient stuff) and go to a new type of . .development system. ... . . One alternative you might want to check is Kontron Electronics . development system products. I'm haven't been in touch with . their product line for several years, but would expect that . they have a descendent of the FutureData Slave Emulator . product line that would either communicate with a host . or plug into an IBM PC or equivalent. They certainly do. (That was my baby at Kontron Electronics). The user interface is the best available in an in-circuit emulator today (of course, I'm biased). They can be reached at (415) 965-7020. . Back when I was the MESS (Manager of Emulation Systems Software) . at FutureData, the HP 64000's were our nearest competitor . in terms of technical merit on 68000 emulators. Both systems . were generally similar, with similar tools available for . cross-compilation and cross-assembly. FutureData had an . edge in the user interface for debugging via the emulator . and in several emulation capabilities. The gap between the Kontron product and their competitors isn't as dramatic as it once was, but it's still worth considering. Keep in mind that successful emulation of a particular target system is highly dependent on a variety of variables, including whether your target's quirks match the emulator's quirks. . Kontron bought FutureData and seems to be keeping a fairly . low profile in this country, though I understand they're . thriving in Europe. It's probably worth checking on what . they offer now. . . Paul Raveling -- Clayton E. Cramer {pyramid,pixar,tekbspa}!optilink!cramer Governments that don't trust most people with weapons, deserve no trust. ---------------------------------------------------------------------------- Disclaimer? You must be kidding! No company would hold opinions like mine!