On Thu, 29 Dec 2005 20:00:30 -0800 Gerhard Adam said: >>But given the resources at their disposal (and without making any >comments on the current inventory levels of returned z900s) it would be >easy for IBM to >put up a well.configured z900 doing precisely this. > >>I think there would be marketing capital in it, too. "See, we've given >every keyboard jockey who can get to a phone unlimited access to MVS - >and none of >them have broken it yet." > >>Unless, of course, it can be broken. > >>It's incredibly difficult to get concepts like this through at IBM >because at every turn you have to make a business case to a manager >whose incentive >>program is different to the manager that let you get to him. > > >I think this argument is oversimplifying the issue. If it were simply a >matter of a CPU and an operating system, then perhaps it wouldn't be so >difficult, but it involves considerably more than this. Consider that >not all z/series processors support all the instruction sets, so >immediately that restricts some assembler language uses. What >subsystems should be installed? CICS? DB2? Who supports all this stuff?
Well, we support it for our students, plus the central Scholastic Alliance hub. It's what we do. Don't know what your employer does, but that's what we do, and it's part of my job description. Do we offer every option? No, but we support 99% of what people want for the Scholastic Alliance. We supported the recent Mainframe Programming contest, supporting lots of students in parts 1 & 2 as TSO users on the hub. Non-technical issues prevented us from supporting part 3, which was separate z/OS systems for a couple hundred students. >What about configuring the network (TCP/IP & VTAM)? Does this also >"promote" the MVS knowledge? Where does the DASD come from? Tape >access? How are USERIDs assigned? Security? If not, then what happens >when somebody clobbers YOUR data sets? Who will handle the SMP/E? Who >will ensure that somebody else doesn't simply corrupt the operating >system since the data sets are accessible for modification? Interesting questions that all need answers, but ones we've been dealing with for many years. We offer students here Linux servers that they can completely destroy. Different operating system, but the issues aren't that different. > >Of course if all the normal security processes have to be in place to >protect the integrity of the environment, then we are certainly not >dealing with a simple implementation and it certainly couldn't be made >available to just anyone with "access to a phone". Well, we come pretty close with our Linux support for OSDL. If co-locators can do it for Windows and Linux, are you claiming it can't be done for z/OS? > >In short, the issue of supporting a bunch of unknown users is neither >trivial, nor cheap. In fact, a strong argument could be made that every >systems programmer with a "sandbox" LPAR already has his/her system. True, but the discussion is getting new people into the mix. People that are not systems programmers today, or those that are but want to do something that isn't controlled by their employer. > >Anyway ... My two cents Like others have said, SHARE is probably a good place to make your feelings known. This isn't really a technical problem. > >Adam /ahw ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
