On 3/27/18, 1:23 PM, "IBM Mainframe Discussion List on behalf of John Eells" <IBM-MAIN@LISTSERV.UA.EDU on behalf of ee...@us.ibm.com> wrote: > I'd quite forgotten about P370, in part because I don't know that it was > ever available external to IBM. It might well have been what the AWS* > things were ogiginally written for, and if so, I stand (well, sit, at > the moment) corrected.
They were widely used in BITNET II by the major interconnect points in the US and overseas (Rice, Cornell, Princeton, CUNY, Weitzman Institute, etc.) mostly to offload CPU and spool space usage (a lot of them ran VM/HPO, VM TCPIP and RSCSv2 just for that purpose) and they could be run on cheap hardware that would run a long time on a standalone UPS. I encountered a number of the oil companies in Houston that had them. I still have the Rice one in my garage; haven't fired it up in a long time. Maybe those came to the BITNET project via an internal source; don't remember, but they weren't that hard to get if you knew what to ask for. > It's not at all clear to me that it would > solve anything that is not already solved by retrieving the files from a > workstation file system (hard drive or DVD), either, and until someone > shows me how it would be, I don't know why I'd support the effort to do > it. (And, as it happens, it's not actually trivial to implement it in > the software delivery process.) I guess that's really the point -- as you know, AWSTAPE "volumes" are just that, a collection of files in a directory on a workstation filesystem plus a metadata file that specified the order of the files on the "tape". Writing a AWSTAPE image creates all the necessary metadata and packaging. You just have to pass the collection of files to the end user in some way, and that way is not media-specific if you have a care in how you do the metadata file. What's elegant about the way AWSTAPE is implemented is that the idea is that the process of making a product available in that form is a matter of copying already existing distribution and service tapes created using an already existing process to an emulated tape device *once*, nothing in IBM or in the customer site has to change in any way other than adding the copy step and the process of delivering the resulting files to the customer. Principle of least change; the people creating the content don't have to learn anything new, and the people consuming the content don't have to do anything new. The effort to produce an AWSTAPE file becomes "mount the distribution tape you already created using the existing process and do a copy of the tape to a emulated device", no new tools or new programming required for most of the people affected; it's just one extra step to copy the distribution tape to an emulated tape "volume" and that step can be heavily automated. Once that copy is done, format no longer matters; how you get those files to the customer is a detail; whether they come on a DVD or over the Internet or by covered wagon, there doesn't have to be a difference in format or tooling, just delivery methods.. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
