ken@argus.UUCP (Kenneth Ng) (07/09/86)
In article <6340MW9@PSUVM>, MW9@PSUVM.BITNET writes: > I would just like to take this opportunity to inform people > that there are VM systems on the net now. We've been on > for awhile and would now like to be considered in discussions. > UN*X doesn't rule the world. In fact, it sucks. (Sorry, that > was a flame and there is no more net.flame, boo hoo.) > Ok, when posting sources think about us vm folks. > And you vm folks, how about posting some relevant stuff! Like what from the Un*x world is relavent to the VM world? And vica versa? I don't think my REXX programs are much good on Un*x environments. Furthermore, if the programs are C, there is another problem: what the heck do you make the square brackets "[" and "]"? Of course we could more to a real language like Fortran or Ada, but that's another story. -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(unreliable) ihnp4!allegra!bellcore!njitcccc!ken soon uucp:ken@rigel.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet soon bitnet: ken@orion.cccc.njit.edu (Yes, we are slowly moving to RFC 920, kicking and screaming) Vulcan jealousy: "I fail to see the logic in prefering Stonn over me" Romulan: "Permit me the glory of the kill"
jsdy@hadron.UUCP (Joseph S. D. Yao) (07/09/86)
In article <6340MW9@PSUVM> MW9@PSUVM.BITNET writes: >I would just like to take this opportunity to inform people >that there are VM systems on the net now. We've been on >for awhile and would now like to be considered in discussions. >UN*X doesn't rule the world. In fact, it sucks. (Sorry, that >was a flame and there is no more net.flame, boo hoo.) >Ok, when posting sources think about us vm folks. >And you vm folks, how about posting some relevant stuff! Unix information should be (probably has been) put in net.unix. I suggest that this young person go post things to net.vm. If enough dinosaurs and whales join the net, net.vm might even get created. -- Joe Yao hadron!jsdy@seismo.{CSS.GOV,ARPA,UUCP} jsdy@hadron.COM (not yet domainised)
gelfand@valid.UUCP (Brooks Gelfand) (07/10/86)
> In article <6340MW9@PSUVM>, MW9@PSUVM.BITNET writes: > Like what from the Un*x world is relavent to the VM world? And vica > versa? I don't think my REXX programs are much good on Un*x environments. > Furthermore, if the programs are C, there is another problem: what > the heck do you make the square brackets "[" and "]"? Of course > we could more to a real language like Fortran or Ada, but that's > another story. > > -- > Kenneth Ng: > Post office: NJIT - CCCC, Newark New Jersey 07102 > uucp(unreliable) ihnp4!allegra!bellcore!njitcccc!ken > soon uucp:ken@rigel.cccc.njit.edu > bitnet(prefered) ken@njitcccc.bitnet > soon bitnet: ken@orion.cccc.njit.edu > (Yes, we are slowly moving to RFC 920, kicking and screaming) > > Vulcan jealousy: "I fail to see the logic in prefering Stonn over me" > Romulan: "Permit me the glory of the kill" There are several way to make the square brackets under VM a. The square brackets are part of the APL character set. If you are using a 3274 controller, simply generate the APL character set option and get the correct key caps for your 3179 or equivalent tubes. b. If you have Waterloo C, follow the instructions in paragraph 1.5.4 of the users guide (Version 1.3 dtd 31 March 1986). c. Use the ALTER subcommand to XEDIT to alter any chosen characters to the bracket symbols. The ALTER subcommand is described in SYSTEM PRODUCT EDITOR COMMAND and MACRO REFERENCE manual SC24-5221. d. Use the CMS SET INPUT command to translate any characters to the [ (X'AD') or the ] (X'BD') symbols. A complete description of the SET INPUT command may be found in the CMS COMMAND and MACRO REFERENCE manual SC19-6209. Brooks Gelfand
george@sysvis.UUCP (07/11/86)
> I would just like to take this opportunity to inform people that there are > VM systems on the net now. We've been on for awhile and would now like to > be considered in discussions. UN*X doesn't rule the world. In fact, it > sucks. (Sorry, that was a flame and there is no more net.flame, boo hoo.) > Ok, when posting sources think about us vm folks. And you vm folks, how > about posting some relevant stuff! > "UNIX is to operating systems as QWERTY is to typing." -me I heard that VM stands for Visceral Maggots. True? It's uncommon for an "OS" to SUCK down about 40% of the total available machine cycles for overhead, as does VM. That's why it wasn't tagged with an `OS' label, I'm sure. Who needs VM? Who cares? IBM didn't have an official `C' compiler as of some time ago so what are you going to do with sources that won't compile, and if they did, the I/O is handled totally differently? Batch is bitch, Butch. Try net.jokes instead of net.sources for your jollies. If VM is so great, why aren't all the "good" microcomputers [IBM/AT (-:] using it? "VM is to operating environments as barnacles are to ships." -me
ted@bcsaic.UUCP (07/12/86)
In article <6340MW9@PS> MW9@6340MW9@PS writes: >I would just like to take this opportunity to inform people >that there are VM systems on the net now. We've been on >for awhile and would now like to be considered in discussions. >UN*X doesn't rule the world. In fact, it sucks. (Sorry, that >was a flame and there is no more net.flame, boo hoo.) >Ok, when posting sources think about us vm folks. >And you vm folks, how about posting some relevant stuff! > >"UNIX is to operating systems as QWERTY is to typing." -me >------- >Michael S. Weiss >The Pennsylvania State University >MW9@PSUVM.BITNET I responded to this article in an e-mail message a few days ago. It has appeared on the net in this group now for the third time, so I believe I am justified in broadcasting a response. It would be preferable that the gentleman (and I use the term advisedly) making the rude comments consider that he is in facting riding upon a dinosaur while criticizing the modern world. With considerable experience using MVS, TSO, VM, and CMS as well as various Univac, DEC, and other operating systems I can say that Unix may have some shortcomings for some applications, but it's a long way in front of whatever is in second place!! If you want to join a group and participate in it, it is certain that the first thing you must do is make rude comments about the membership of the group. Who needs this kind of trash, especially in net.general? Surely not Usenet! TJ {With Amazing Grace} The Piper (aka Ted Jardine) CFI-ASME/I Boeing Knowledge Based Systems Center ...uw-beaver!uw-june!bcsaic!ted
bzs@bu-cs.UUCP (Barry Shein) (07/14/86)
In article <6340MW9@PSUVM>, MW9@PSUVM.BITNET writes: > I would just like to take this opportunity to inform people > that there are VM systems on the net now. We've been on > for awhile and would now like to be considered in discussions. > UN*X doesn't rule the world... > ...Ok, when posting sources think about us vm folks. Uh, UNIX runs under VM, what are you talking about?? What is a VM "source"? Unix? SOURCE DC X'83000000'? -Barry Shein, Boston University
ken@argus.UUCP (Kenneth Ng) (07/14/86)
In article <934@bu-cs.UUCP>, bzs@bu-cs.UUCP (Barry Shein) writes: > > In article <6340MW9@PSUVM>, MW9@PSUVM.BITNET writes: > > I would just like to take this opportunity to inform people > > that there are VM systems on the net now. We've been on > > for awhile and would now like to be considered in discussions. > > UN*X doesn't rule the world... > > ...Ok, when posting sources think about us vm folks. > > Uh, UNIX runs under VM, what are you talking about?? What is a > VM "source"? Unix? SOURCE DC X'83000000'? > > -Barry Shein, Boston University Quite true, VM thinks that Un*x is another application program. -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken soon uucp:ken@argus.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet soon bitnet: ken@orion.cccc.njit.edu (We are VERY slowly moving to RFC 920, kicking and screaming) Kirk: "Spock, the women on your planet are logical, that is the only planet in the federation that can make that claim" Savaak: "He's so....human" Spock: "No one is perfect"
jennings@onion.cs.reading.AC.UK (Richard Jennings) (07/14/86)
In article <6340MW9@PS> MW9%6340MW9@6340MW9.UUCP writes: > >UN*X doesn't rule the world. In fact, it sucks. > >"UNIX is to operating systems as QWERTY is to typing." -me >------- >Michael S. Weiss >The Pennsylvania State University >MW9@PSUVM.BITNET Just a thought: does vm enforce silly userids??? >All my opinions were bought at Opinions-R-Us during their fall clearance. Oh, yeah! (best signoff I've heard for quite a bit) -- Richard Jennings (incomprehensible jargon follows): _^_ jennings@sage.cs.reading.UUCP <o|o> / \ - remember space invaders??
ken@argus.UUCP (Kenneth Ng) (07/15/86)
In article <26@onion.cs.reading.AC.UK>, jennings@onion.cs.reading.AC.UK (Richard Jennings) writes: > In article <6340MW9@PS> MW9%6340MW9@6340MW9.UUCP writes: > > > >UN*X doesn't rule the world. In fact, it sucks. > > > >"UNIX is to operating systems as QWERTY is to typing." -me > >------- > >Michael S. Weiss > >The Pennsylvania State University > >MW9@PSUVM.BITNET > Just a thought: does vm enforce silly userids??? The user id's on vm systems don't have "silly" userid's any more than other operating systems, unless the administration enforces them. -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken soon uucp:ken@argus.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet soon bitnet: ken@orion.cccc.njit.edu (We are VERY slowly moving to RFC 920, kicking and screaming) Kirk: "Spock, the women on your planet are logical, that is the only planet in the federation that can make that claim" Savaak: "He's so....human" Spock: "No one is perfect"
jcollas@ll-xn.ARPA (Juan Collas) (07/15/86)
>In article <6340MW9@PSUVM>, Michael Weiss commands: >> I would just like to take this opportunity to inform people that there are >> VM systems on the net now. We've been on for awhile and would now like to >> be considered in discussions. UN*X doesn't rule the world. In fact, it >> sucks. (Sorry, that was a flame and there is no more net.flame, boo hoo.) >> Ok, when posting sources think about us vm folks. And you vm folks, how >> about posting some relevant stuff! Geez. What an monarchical tone to your message! In article <-121460420@sysvis>, george@sysvis.UUCP rebuts: >I heard that VM stands for Visceral Maggots. True? It's uncommon for an >"OS" to SUCK down about 40% of the total available machine cycles for overhead, >as does VM. That's why it wasn't tagged with an `OS' label, I'm sure. VM is a little different than your usual operating system. What it does extremely well is provide a user with what appears to be a standalone IBM/370. This means that systems people can do kernel development while other people are using the system. No need to wait around until your average late-night hackers have gone to bed. It also means that VM can run several different OS's simultaneously. OS's like UNIX, for example. It might be better to think of it as a "meta-" operating system. > IBM didn't have an official `C' compiler as of some > time ago so what are you going to do with sources that won't compile, and if > they did, the I/O is handled totally differently? Batch is bitch, Butch. > Try net.jokes instead of net.sources for your jollies. If VM is so great, why > aren't all the "good" microcomputers [IBM/AT (-:] using it? > > "VM is to operating environments as barnacles are to ships." -me Funny. I believe that Waterloo, Bell Labs (ATTIS), SAS Institute, Oracle all sell C Compilers for the IBM/370. Most stuff can port between UNIX and CMS. (Most, not all). Juan Collas, MIT Lincoln Laboratory PS: Maybe the two of you could take the insults off-line or something.
ken@argus.UUCP (Kenneth Ng) (07/16/86)
In article <-121460420@sysvis>, george@sysvis.UUCP writes: > > > I would just like to take this opportunity to inform people that there are > > VM systems on the net now. We've been on for awhile and would now like to > > be considered in discussions. UN*X doesn't rule the world. In fact, it > > sucks. (Sorry, that was a flame and there is no more net.flame, boo hoo.) > > Ok, when posting sources think about us vm folks. And you vm folks, how > > about posting some relevant stuff! > > I heard that VM stands for Visceral Maggots. True? It's uncommon for an > "OS" to SUCK down about 40% of the total available machine cycles for overhead, > as does VM. That's why it wasn't tagged with an `OS' label, I'm sure. Who > needs VM? Who cares? IBM didn't have an official `C' compiler as of some > time ago so what are you going to do with sources that won't compile, and if > they did, the I/O is handled totally differently? Batch is bitch, Butch. Strange, I've been on an IBM 4361 for almost a year and have never used a batch process, practically everything is interactive. Also, VM has many I/O capabilities which are better than Un*x. For starters, the equivalent of the Un*x write(1) on VM does not go splattering all over whatever else is being displayed on the screen. By default it is displayed on a seperate screen, or if you write the control logic, you can even intercept the wwrites and do whatever you want with them. > If VM is so great, why > aren't all the "good" microcomputers [IBM/AT (-:] using it? I believe one of the higher PC models does use a variation of VM. But VM on a single process machine does not really make much sense. Furthermore a lot of VM's power is in its hardware implementation, the paging and lookaside tables for example. Lastly, is there really a reason why programs posted here have to be in C? -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken soon uucp:ken@argus.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet soon bitnet: ken@orion.cccc.njit.edu (We are VERY slowly moving to RFC 920, kicking and screaming) Kirk: "Spock, the women on your planet are logical, that is the only planet in the federation that can make that claim" Savaak: "He's so....human" Spock: "No one is perfect"
herbie@polaris.UUCP (Herb Chong) (07/17/86)
In article <-121460420@sysvis> george@sysvis.UUCP writes: >I heard that VM stands for Visceral Maggots. True? It's uncommon for an >"OS" to SUCK down about 40% of the total available machine cycles for overhead, >as does VM. That's why it wasn't tagged with an `OS' label, I'm sure. Who >needs VM? Who cares? IBM didn't have an official `C' compiler as of some >time ago so what are you going to do with sources that won't compile, and if >they did, the I/O is handled totally differently? Batch is bitch, Butch. >Try net.jokes instead of net.sources for your jollies. If VM is so great, why >aren't all the "good" microcomputers [IBM/AT (-:] using it? the mindless drivel i won't bother replying to. the facts are wrong. a VM system can use 40% of the CPU. it can also use 2% or 90%. a unix system can do the same thing. unless you are specific about the work load, performance figures like 40% system state are meaningless. as it happens, some of the 4.2 VAX systems i used to work on before i joined IBM regularly spent 60% of the time in system state. it can happen and will happen and be perfectly normal given the work the system is running. whether IBM has a C compiler or not is irrelevant. as pointed out in another article, there are C compilers for VM. some of them are as stringent in compiling the language as lint is so there is no need for lint. batch is an option on VM systems. unless you want it, you don't get it. would you find Multics on a micro? whether a system runs on a micro and a super computer is not very meaningful as the two sets of customers are totally different and have different problems to solve. VM addresses a specific problem and is one of the VERY few systems that does so. it is not designed for micros and there is little advantage to having it run in a micro. having the same user interface is a different question from having the same system. for instance, the XT/370 and the AT/370 happen to run almost all CMS code and gives the appearance of running in a CMS virtual machine but it really isn't even though it is using the 370 instruction set. Herb Chong, IBM Research... I'm still user-friendly -- I don't byte, I nybble.... VNET,BITNET,NETNORTH,EARN: HERBIE AT YKTVMH UUCP: {allegra|cbosgd|cmcl2|decvax|ihnp4|seismo}!philabs!polaris!herbie CSNET: herbie%ibm.com@csnet-relay ARPA: herbie@ibm.com, herbie%yktvmh.bitnet@wiscvm.wisc.edu ======================================================================== DISCLAIMER: what you just read was produced by pouring lukewarm tea for 42 seconds onto 9 people chained to 6 Ouiji boards.
gelfand@valid.UUCP (Brooks Gelfand) (07/17/86)
> > > I would just like to take this opportunity to inform people that there are > > VM systems on the net now. We've been on for awhile and would now like to > > be considered in discussions. UN*X doesn't rule the world. In fact, it > > I heard that VM stands for Visceral Maggots. True? It's uncommon for an > "OS" to SUCK down about 40% of the total available machine cycles for overhead, > as does VM. That's why it wasn't tagged with an `OS' label, I'm sure. Who > needs VM? Who cares? IBM didn't have an official `C' compiler as of some > time ago so what are you going to do with sources that won't compile, and if > they did, the I/O is handled totally differently? Batch is bitch, Butch. > Try net.jokes instead of net.sources for your jollies. If VM is so great, why > aren't all the "good" microcomputers [IBM/AT (-:] using it? > > "VM is to operating environments as barnacles are to ships." -me Oh boy, sounds like we have a REAL COMPUTER SCIENTIST here. Having used both VM and UNIX I'd like to add my opinions. First -me is a bit behind times. VM use to take about 40% of the machine cycles on the old 370/145 about 12 years ago. Since then, the addition of microcode assists has reduced the overhead to about 10%. However, depending on the guest operating system, this overhead may be offset by the increased efficiencies in CCW translation (what the machine does to perform an I/O) and storage management. As to who needs VM and who cares - the people who are running it, that's who. There are many VM sites. As to the 'OS' label, VM is more than an operating system. When a user logs on under VM, rather than just being a process each user thinks he has his own 370 complete with control registers, general registers, PSW, memory and I/O devices as allocated in the directory. Thus you have your own Virtual MACHINE - hence the name VM. In this machine you IPL (boot) whatever operating system (OS) you want (and have access to) e.g. CMS, MVS, DOS/VSE, UNIX. You don't like these? There is a good assembler with CMS; write your own. The advantages of this system are obvious. If a program in one virtual MAchine does something that would crash the machine, only that virtual machine goes down. The other users continue to operate. The only effect they see is a slight increase in speed, since the crashed 'machine' is not using any cycles. Since VM handles all the I/O, it is an excellent system for people who who are debugging Physical IOCS. Vm will not let the user move the disk heads outside their alloted area even with PIOCS. Thus one user cannot corrupt anothers files by accident or even on purpose. Last but not least, VM will run on both the PC/XT370 and the PC/AT370. It runs slowly but it runs. Brooks Gelfand ...!hplabs!pesnta!valid!gelfand
george@sysvis.UUCP (07/18/86)
Just to set the record straight, I was a systems programmer in an IBM shop some years ago when VM was first announced. I installed a pre-release copy of VM for testing purposes and became quite familiar with it over some time. I'm sure that improvements have been made (there was room for them). So you can all now stop explaining to me how "it works". Cheez... I answered the original message in the same tone it was written. > Written 8:31 pm Jul 15, 1986 by argus.UUCP!ken > Strange, I've been on an IBM 4361 for almost a year and have never used a > batch process, practically everything is interactive. Also, VM has many > I/O capabilities which are better than Un*x. You ignored my (hard tested) number about VM's overhead. I wonder how many MIPS per user are needed by an interactive multi-os/VM system? Do you know? "Better" is mostly a relative word, so I won't get into a frame of reference discussion with you. Do all net readers know what an IBM 4361 is, Ken? (Please don't explain it to me.) > For starters, the equivalent of the Un*x write(1) on VM does not go > splattering all over whatever else is being displayed on the screen. > By default it is displayed on a seperate screen, or ... > do whatever you want with them. Gee, I hope there are no emergency messages to go through...No wait! I know! We'll just buy two terminals for each user (from IBM?). Wow! the thought! [BTW (seperate <> separate)] > I believe one of the higher PC models does use a variation of VM. But > VM on a single process machine does not really make much sense. Further- > more a lot of VM's power is in its hardware implementation, the paging and > lookaside tables for example. Gosh, you mean that $meta-million dollar machine has some hardware that $4K micros don't implement? How utterly absurd of those dastardly micros! What is the $$ cost per user of your 4361? > Lastly, is there really a reason why programs posted here have to be in C? No. Its just customary for UN*X systems. Portability and all that. ----------------------------------------------------------------------------- > Written 12:19 pm Jul 15, 1986 by ll-xn.ARPA!jcollas > Geez. What an monarchical tone to your message! "That which we see in others..." > VM is a little different than your usual operating system. I didn't think it was an operating system, did you? BTW -- its different FROM > It might be better to think of it as a "meta-" operating system. You think of it in your own terms. I will do so in mine. >> ...[me]... IBM didn't have an official `C' compiler as of some >> time ago so what are you going to do with sources that won't compile, and if > Funny. I believe that Waterloo, Bell Labs (ATTIS), SAS Institute, Oracle > all sell C Compilers for the IBM/370. Most stuff can port between UNIX > and CMS. (Most, not all). I didn't see IBM's name in that list. Still no official `C' compiler. This means that only some very small percent of the IBM shops compile C programs. Did you not tell me everything?
herbie@polaris.UUCP (Herb Chong) (07/19/86)
In article <607@bcsaic.UUCP> ted@bcsaic.UUCP (ted jardine) writes: >It would be preferable that the gentleman (and I use the term advisedly) >making the rude comments consider that he is in facting riding upon a >dinosaur while criticizing the modern world. With considerable experience >using MVS, TSO, VM, and CMS as well as various Univac, DEC, and other >operating systems I can say that Unix may have some shortcomings for some >applications, but it's a long way in front of whatever is in second place!! first, i agree that the person from PSUVM needs a little more tact but i think you should consult your history books a little to straighten out the facts. 1) OS/360 was designed in the early 60's, around and about 1964 2) VM/370 was designed in and around 1967 3) Unix was first called unix around 1970 if OS/360 is a dinosaur based purely on age then unix isn't far behind. unix is not a state of the art operating system. it may have been about 10 years ago but anything that has been around long enough to be a "commercial success" and has an ANSI standard pending is by definition ancient. i also note that you give no criteria for why unix is better. it is, i agree, a very good programming environment and testbed for OS ideas and stuff like that. VM/370 and it's descendants are much better for designing and building OS's that will run on bare iron because it simulates a complete hardware architecture. OS, in it's latest incarnation MVS, provides a very good system to build LARGE applications for transaction processing. if i had a 2K ROM monitor to manage a small display if LEDs, i'm not going to chose any of the above. what a system is intended to do determine what criteria are relevant. if you believed the marketing hypes in all the trade rags, you would think that merely the mention of unix would mean instant customer satisfaction. suppose i wanted to have an office automation environment that supported 15,000 simultaneously users. can you think of an existing unix system that could handle that? i can think of one that might. the VM/PROFS system (running on a dyadic processor) down the road has been at the 15,000 user point for nearly a year. suppose i wanted to do new compiler development? off the shelf lex and yacc are second to none of all OS's that i am aware of. this does not mean the tools are not available on other systems. unix may be great for the things that YOU want to do but to others, what unix isn't good at is what matters. Herb Chong, IBM Research... I'm still user-friendly -- I don't byte, I nybble.... VNET,BITNET,NETNORTH,EARN: HERBIE AT YKTVMH UUCP: {allegra|cbosgd|cmcl2|decvax|ihnp4|seismo}!philabs!polaris!herbie CSNET: herbie%ibm.com@csnet-relay ARPA: herbie@ibm.com, herbie%yktvmh.bitnet@wiscvm.wisc.edu ======================================================================== DISCLAIMER: what you just read was produced by pouring lukewarm tea for 42 seconds onto 9 people chained to 6 Ouiji boards.
taylor@glasgow.glasgow.UUCP (Jem Taylor) (07/21/86)
In article <455@valid.UUCP> gelfand@valid.UUCP writes: >As to the 'OS' label, VM is more than an operating system. When a user >logs on under VM, rather than just being a process each user thinks >he has his own 370 complete with control registers, general registers, >PSW, memory and I/O devices as allocated in the directory. Thus you >have your own Virtual MACHINE - hence the name VM. In this machine >you IPL (boot) whatever operating system (OS) you want (and have >access to) e.g. CMS, MVS, DOS/VSE, UNIX. You don't like these? >There is a good assembler with CMS; write your own. The advantages of this >system are obvious. If a program in one virtual MAchine does something >that would crash the machine, only that virtual machine goes down. The >other users continue to operate. >Since VM handles all the I/O, it is an excellent system for people who >who are debugging Physical IOCS. Vm will not let the user move the >disk heads outside their alloted area even with PIOCS. Thus one user >cannot corrupt anothers files by accident or even on purpose. During my undergraduate days, our site's 370/165, running Phoenix/MVT, had a hardware upgrade to a 3081. To allow system development under MVS, to bring Phoenix3/MVS up to the standards we were used to :-), without disrupting the users, the new machine ran VM which provided a virtual 370 to run MVT, and several virtual 3081's to run test, development, and release versions of Phoenix on MVS. Some 80 concurrent users used the "370", and I was lucky enough ( as an undergraduate ) to be allowed an ID on the pre-release "3081" ... using a pseudo-cardpunch/cardreader to move my files onto the 3081's filestore from the 370's, if I remember right. At the time, some 10 people might be found at peaktime on that 3081; eventually all the users migrated to this virtual machine and the disks were re-allocated to it, with the virtual 370 being withdrawn from service. When I last logged on, that virtual 3081's concurrent session limit was 200, with >100 logged on at the time. VM was described to me as a 'hypervisor', which sounds appropriate. I assume this is the same 'VM' ? -Jem. Give me a VAX I can run BSD Unix _and_ VMS on ... sorry, I forgot. VAXes are super-minis, not mainframes :-). -- JANET: taylor@uk.ac.glasgow.cs -o Jemima USENET: { uk }!cs.glasgow.ac.uk!taylor (==). Puddleduck
colonel@sunybcs.UUCP (Col. G. L. Sicherman) (07/24/86)
>Since VM handles all the I/O, it is an excellent system for people who >who are debugging Physical IOCS. Vm will not let the user move the >disk heads outside their alloted area even with PIOCS. Thus one user >cannot corrupt anothers files by accident or even on purpose. We recently got a 3081 running VM. When the 3081 itself crashed, EVERYBODY's files were trashed! I shall never understand how you can make a computer more reliable by simulating it on another computer. -- Col. G. L. Sicherman UU: ...{rocksvax|decvax}!sunybcs!colonel CS: colonel@buffalo-cs BI: colonel@sunybcs, csdsicher@sunyabva
ken@argus.UUCP (Kenneth Ng) (07/25/86)
In article <455@valid.UUCP>, gelfand@valid.UUCP (Brooks Gelfand) writes: > As to the 'OS' label, VM is more than an operating system. When a user > logs on under VM, rather than just being a process each user thinks > he has his own 370 complete with control registers, general registers, > PSW, memory and I/O devices as allocated in the directory. Thus you > have your own Virtual MACHINE - hence the name VM. In this machine > you IPL (boot) whatever operating system (OS) you want (and have > access to) e.g. CMS, MVS, DOS/VSE, UNIX. You don't like these? > There is a good assembler with CMS; write your own. The advantages of this > system are obvious. If a program in one virtual MAchine does something > that would crash the machine, only that virtual machine goes down. The > other users continue to operate. The only effect they see is a slight > increase in speed, since the crashed 'machine' is not using any cycles. > Since VM handles all the I/O, it is an excellent system for people who > who are debugging Physical IOCS. Vm will not let the user move the > disk heads outside their alloted area even with PIOCS. Thus one user > cannot corrupt anothers files by accident or even on purpose. > > Last but not least, VM will run on both the PC/XT370 and the PC/AT370. > It runs slowly but it runs. > > Brooks Gelfand > ...!hplabs!pesnta!valid!gelfand Heck, and there is one feature that I found interesting about VM, VM can be run on top of VM, which is great for playing around with the operating system internal, since you can't damage the real OS when your playing with your own. -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken soon uucp:ken@argus.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet soon bitnet: ken@orion.cccc.njit.edu (We are VERY slowly moving to RFC 920, kicking and screaming) Spock: "Captain, you are an excellent Starship Captain, but as a taxi driver, you need much to be desired." Savaak: "He's so....human" Spock: "No one is perfect"
herbie@polaris.UUCP (Herb Chong) (07/25/86)
In article <-121460412@sysvis> george@sysvis.UUCP writes: >You ignored my (hard tested) number about VM's overhead. I wonder how many >MIPS per user are needed by an interactive multi-os/VM system? Do you know? >"Better" is mostly a relative word, so I won't get into a frame of reference >discussion with you. Do all net readers know what an IBM 4361 is, Ken? >(Please don't explain it to me.) explain your workload. benchmarking in a notoriously difficult process and numbers are meaningful only in context. you do not give context. .... deleted stuff about terminals.... >Gee, I hope there are no emergency messages to go through...No wait! I know! >We'll just buy two terminals for each user (from IBM?). Wow! the thought! this dates you. for quite a while now, high priority messages automatically clear (and preserve the former contents of) your screen no matter what you are doing to display the message. when you clear the screen, the old contents is restored. it is not a screen manager but it's a lot better than write, not that anyone in their right mind would use write except on a dumb terminal. >> I believe one of the higher PC models does use a variation of VM. But >> VM on a single process machine does not really make much sense. Further- >> more a lot of VM's power is in its hardware implementation, the paging and >> lookaside tables for example. > >Gosh, you mean that $meta-million dollar machine has some hardware that $4K >micros don't implement? How utterly absurd of those dastardly micros! >What is the $$ cost per user of your 4361? you are comparing apples and oranges. how many MIPS per user are available during each time-slice? how many effective MIPS are available? i can format a 2,000 line document in 5 seconds. do that on your micro. they just can't be compared on the same terms. Herb Chong, IBM Research... I'm still user-friendly -- I don't byte, I nybble.... VNET,BITNET,NETNORTH,EARN: HERBIE AT YKTVMH UUCP: {allegra|cbosgd|cmcl2|decvax|ihnp4|seismo}!philabs!polaris!herbie CSNET: herbie%ibm.com@csnet-relay ARPA: herbie@ibm.com, herbie%yktvmh.bitnet@wiscvm.wisc.edu ======================================================================== DISCLAIMER: what you just read was produced by pouring lukewarm tea for 42 seconds onto 9 people chained to 6 Ouiji boards.
ken@argus.UUCP (Kenneth Ng) (07/26/86)
In article <494@sunybcs.UUCP>, colonel@sunybcs.UUCP (Col. G. L. Sicherman) writes: > >Since VM handles all the I/O, it is an excellent system for people who > >who are debugging Physical IOCS. Vm will not let the user move the > >disk heads outside their alloted area even with PIOCS. Thus one user > >cannot corrupt anothers files by accident or even on purpose. > > We recently got a 3081 running VM. When the 3081 itself crashed, > EVERYBODY's files were trashed! > > I shall never understand how you can make a computer more reliable > by simulating it on another computer. > -- > Col. G. L. Sicherman A system crash could wipe out any file system, Un*x, VM, DOS, etc. The purpose of VM is to mitigate the problems that a faulty user program could cause. A program running under a prived id can blow away any operating system I know of. -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken soon uucp:ken@argus.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet soon bitnet: ken@orion.cccc.njit.edu (We are VERY slowly moving to RFC 920, kicking and screaming) Spock: "Captain, you are an excellent Starship Captain, but as a taxi driver, you need much to be desired." Savaak: "He's so....human" Spock: "No one is perfect"
ken@argus.UUCP (Kenneth Ng) (07/26/86)
In article <1426@rsch.wisc.edu>, herbie@polaris.UUCP (Herb Chong) writes: > In article <-121460412@sysvis> george@sysvis.UUCP writes: ^^^^^^^^^^ gee, is that a vaid item? [ deleted material on VM vs. Un*x arguements ] Is something on your mailer broke? I read this about 4 times. > > you are comparing apples and oranges. how many MIPS per user are > available during each time-slice? how many effective MIPS are > available? i can format a 2,000 line document in 5 seconds. do that > on your micro. they just can't be compared on the same terms. > > Herb Chong, IBM Research... > > I'm still user-friendly -- I don't byte, I nybble.... > > VNET,BITNET,NETNORTH,EARN: HERBIE AT YKTVMH > UUCP: {allegra|cbosgd|cmcl2|decvax|ihnp4|seismo}!philabs!polaris!herbie > CSNET: herbie%ibm.com@csnet-relay > ARPA: herbie@ibm.com, herbie%yktvmh.bitnet@wiscvm.wisc.edu > ======================================================================== > DISCLAIMER: what you just read was produced by pouring lukewarm > tea for 42 seconds onto 9 people chained to 6 Ouiji boards. -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken soon uucp:ken@argus.cccc.njit.edu bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet soon bitnet: ken@orion.cccc.njit.edu (We are VERY slowly moving to RFC 920, kicking and screaming) Spock: "Captain, you are an excellent Starship Captain, but as a taxi driver, you need much to be desired." Savaak: "He's so....human" Spock: "No one is perfect"
gelfand@valid.UUCP (Brooks Gelfand) (07/28/86)
> >Since VM handles all the I/O, it is an excellent system for people who > >who are debugging Physical IOCS. Vm will not let the user move the > >disk heads outside their alloted area even with PIOCS. Thus one user > >cannot corrupt anothers files by accident or even on purpose. > > We recently got a 3081 running VM. When the 3081 itself crashed, > EVERYBODY's files were trashed! > > I shall never understand how you can make a computer more reliable > by simulating it on another computer. > -- > Col. G. L. Sicherman > UU: ...{rocksvax|decvax}!sunybcs!colonel > CS: colonel@buffalo-cs > BI: colonel@sunybcs, csdsicher@sunyabva No operating system, by itself, can imporve hardware reliability - VM, UNIX, or BRAND-X. If the hardware itself fails, the system comes down unless you have duplicate hardware such as a TANDEM system. When the hardware crashes you will loose work in progress - the files in core. However, you should not have lost any files on disk unless you were writing to them at the time of the crash. Or did you perhaps have a disk head crash? Brooks Gelfand
ronc@fai.UUCP (Ronald O. Christian) (07/29/86)
In article <503@valid.UUCP> gelfand@valid.UUCP (Brooks Gelfand) writes: >No operating system, by itself, can imporve hardware reliability - >[..] When the >hardware crashes you will loose work in progress - the files in core. >However, you should not have lost any files on disk unless you were >writing to them at the time of the crash. Or did you perhaps have a >disk head crash? > >Brooks Gelfand Not necessarily. Depends, for instance, on how often the disk is updated with directory information. One of the first things OS engineers discover is that keeping directory info in ram decreases overhead, speeds things up. A CPU crash, not a head crash, then causes loss of data because the directory (superblock, in Unix terminology) has old information. How well the OS can recover from this is a good measure of robustness. Some do pretty badly. Besides, I don't think that was the Col.'s point. I'm not sure how virtual machines are implemented, but it occurs to me that there is probably even more possibility of losing files (whole environments?) if the hardware (on which several virtual machines exist) gets sick. Ron -- -- Ronald O. Christian (Fujitsu America Inc., San Jose, Calif.) seismo!amdahl!fai!ronc -or- ihnp4!pesnta!fai!ronc Oliver's law of assumed responsibility: "If you are seen fixing it, you will be blamed for breaking it."
gelfand@valid.UUCP (Brooks Gelfand) (08/01/86)
> In article <503@valid.UUCP> gelfand@valid.UUCP (Brooks Gelfand) writes: > >No operating system, by itself, can imporve hardware reliability - > >[..] When the > >hardware crashes you will loose work in progress - the files in core. > >However, you should not have lost any files on disk unless you were > >writing to them at the time of the crash. Or did you perhaps have a > >disk head crash? > > > >Brooks Gelfand > > Not necessarily. Depends, for instance, on how often the disk is > updated with directory information. One of the first things OS > engineers discover is that keeping directory info in ram decreases > overhead, speeds things up. A CPU crash, not a head crash, then > causes loss of data because the directory (superblock, in Unix terminology) > has old information. How well the OS can recover from this is a > good measure of robustness. Some do pretty badly. > > Besides, I don't think that was the Col.'s point. I'm not sure how > virtual machines are implemented, but it occurs to me that there is > probably even more possibility of losing files (whole environments?) > if the hardware (on which several virtual machines exist) gets sick. > > > Ron > -- > -- > Ronald O. Christian (Fujitsu America Inc., San Jose, Calif.) > seismo!amdahl!fai!ronc -or- ihnp4!pesnta!fai!ronc > > Oliver's law of assumed responsibility: > "If you are seen fixing it, you will be blamed for breaking it." Ron, you are correct. VM keeps the disk directory in core when the disk is accessed. I think I covered that by saying that the file that is being writen to would be lost. As an example: I am updating File A. As i write to file A its directory entry (in core is modified). The directory entry of all other files remain unchanged. The CPU crashes; the in core directory is lost. When the CPU comes back up and the disk is accessed, the directory in core is the old directory previous to any modifications to file A. Some blocks in file A may be modified depending on the program that is modifying the file. Any blocks added to file A are back on the free list. This is how CMS would treat the file. Since it is possible to run many different operating systems under CP (VM), it is impossible to say what happened without `knowing the exact circumstances. Brooks Gelfand
craig@comp.lancs.ac.uk (Craig Wylie) (08/04/86)
In article <503@valid.UUCP> gelfand@valid.UUCP (Brooks Gelfand) writes: >No operating system, by itself, can imporve hardware reliability - >[..] When the >hardware crashes you will loose work in progress - the files in core. >However, you should not have lost any files on disk unless you were >writing to them at the time of the crash. Or did you perhaps have a >disk head crash? > >Brooks Gelfand Just joining in, I have never lost a file on a VM system. However when I was using VM I always thought that the file structure placed on top of VM (be it cms or system V equivalent) determined what happened, ie when superblock writes took place etc.. Back to the first posting in this series now, the poster asked that we bear in mind VM systems when writing sources. OK, what do we have to remember, tell us about the differences and similarities, what can you do on VM that you can't do in UNIX (under something that is common to both, there is no point saying that you can write EXEC2 (don't laugh -- they designed it a long time ago) or REXX. What is the difference between C on VM and UNIX. Is there a C compiler for VM/CMS .... Craig. -- UUCP: ...!seismo!mcvax!ukc!dcl-cs!craig| Post: University of Lancaster, DARPA: craig%lancs.comp@ucl-cs | Department of Computing, JANET: craig@uk.ac.lancs.comp | Bailrigg, Lancaster, UK. Phone: +44 524 65201 Ext. 4146 | LA1 4YR Project: Cosmos Distributed Operating Systems Research Group
ken@argus.UUCP (Kenneth Ng) (08/07/86)
In article <336@comp.lancs.ac.uk>, craig@comp.lancs.ac.uk (Craig Wylie) writes: > saying that you can write EXEC2 (don't laugh -- they designed it > a long time ago) or REXX. What is the difference between C on > VM and UNIX. Is there a C compiler for VM/CMS .... I know of two C compilers for VM/CMS: Waterlou and Whitesmith. After using them both I must say I prefer Waterlou far better. > UUCP: ...!seismo!mcvax!ukc!dcl-cs!craig| Post: University of Lancaster, > DARPA: craig%lancs.comp@ucl-cs | Department of Computing, > JANET: craig@uk.ac.lancs.comp | Bailrigg, Lancaster, UK. > Phone: +44 524 65201 Ext. 4146 | LA1 4YR > Project: Cosmos Distributed Operating Systems Research Group -- Kenneth Ng: Post office: NJIT - CCCC, Newark New Jersey 07102 uucp(for a while) ihnp4!allegra!bellcore!argus!ken !psuvax1!cmcl2!ciap!andromeda!argus!ken *** WARNING: NOT ken@bellcore.uucp *** bitnet(prefered) ken@njitcccc.bitnet or ken@orion.bitnet Spock: "Captain, you are an excellent Starship Captain, but as a taxi driver, you leave much to be desired." Kirk: "What do you mean, 'if both survive' ?" T'Pow: "This combat is to the death"
cd-peter@gustaf.UUCP (Peter T|rnqvist) (08/22/86)
>> Funny. I believe that Waterloo, Bell Labs (ATTIS), SAS Institute, Oracle >> all sell C Compilers for the IBM/370. Most stuff can port between UNIX >> and CMS. (Most, not all). > I didn't see IBM's name in that list. Still no official `C' compiler. ... There IS a C compiler from IBM for the IBM/370! Program number 5713-AAG or 5713-AAH (VM/CMS or MVS(/XA)-versions).