The following message is a courtesy copy of an article that has been posted to bit.listserv.ibm-main,alt.folklore.computers as well.
[EMAIL PROTECTED] (Sebastian Welton) writes: > I have an original IBM Thinkpad. This is a small brown pocket notepad with > the word 'THINK' printed on the front and 'IBM' on the back (pn 520-6430 nad > 520-6431) still with the original paper pad inside but I think I'll keep it > as its > quite amusing showing people. I have a couple of the brown pocket notepads ... but i also have (round, clear, globe): http://www.garilc.com/~lynn/vnet1000.jpg it has gotten a little dinged over the years. the internal network http://www.garlic.com/~lynn/subnetwork.html#internalnet was larger than arpanet/internet from just about the beginning until sometime possibly mid-85. past reference mentioning the 1000th node http://www.garlic.com/~lynn/99.html#112 http://www.garlic.com/~lynn/internet.htm#22 the internal network was originally developed at the science center http://www.garlic.com/~lynn/subtopic.html#545tech the same place that originated virtual machines, GML, lots of interactive stuff. for recent slightly related networking post about a couple yrs earlier (1980) ... 300 people from the IMS group having to be moved to offsite location ... because STL had filled up (includes screen shot of the 3270 logon logo): http://www.garlic.com/2008m.html#20 IBM-MAIN longevity One of the interesting aspects of the internal network implementation was that it effectively had a form of gateway implementation in every node. this became important when interfacing with hasp/jes networking implementations. part of the issue was that hasp/jes networking started off defining nodes using spare slots in the 255-entry table for psuedo (unit record) devices ... typical hasp/jes might have only 150 entries available for defining network nodes. hasp/jes implementation also had a habit of discarding traffic where the originating node and/or the destination node wasn't in its internal table. the internal network quickly exceeded the number of nodes that could be defined in hasp/jes ... and its proclivity for discarding traffic ... pretty much regulated hasp/jes to boundary nodes. by the time hasp/jes got around to increasing the limit to 999 nodes ... the internal network was already over 1000 nodes ... and by the time it was further increased to 1999 nodes ... the internal network was over 2000 nodes. hasp/jes implementation also had a design flaw where the network information was intermingled with other hasp/jes processing control information (as opposed to clean separation). the periodic outcome that two has/jes systems at different release levels were typically unable to communicate ... and in some cases, release incompatibilities could cause other hasp/jes systems to crash (there is infamous scenario where a san jose hasp/jes system was crashing hurseley hasp/jes systems). The combination of the internal networking support started accumulating some number of "release-specific" hasp/jes "drivers" ... where an intermediate internal network node was configured to start the corresponding hasp/jes driver for the system on the other end of the wire. As the problems with release incompatibilities between hasp/jes systems increased ... the internal network code evolved a canonical hasp/jes representation ... and drivers would translate format to the specific hasp/jes release (as appropriate). In the hursley crashing scenario ... somebody even got around to blaming the internal network code for not preventing a san jose hasp/jes systems from crashing hurseley hasp/jes systems. By the time, BITNET started http://www.garlic.com/~lynn/subnetwork.html#bitnet they had pretty much eliminated shipping native drivers ... just the hasp/jes compatible drivers ... even tho the native drivers were much more efficient and had higher thruput than the hasp/jes drivers ... although the native drivers did continue to be used on the internal network (note these were *NOT* SNA). misc. past posts mentioning hasp/jes (including hasp/jes networking support) http://www.garlic.com/~lynn/subtopic.html#hasp -- 40+yrs virtualization experience (since Jan68), online at home since Mar70 ---------------------------------------------------------------------- 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
