The IBM breakthrough supercomputer Roadrunner is to be be decommissioned. From breakthrough in 2008 to
obsolete five years later.
Paul Pierce’s Historical Computer Collection One of the most extensive around outside of a museum.
The beginning of the end for mainframes?
” … She noted that mainframes still have their place in computing where it does not matter if “end-user interfaces are clunky and somewhat inflexible,” when there is a need for a system that delivers “extremely reliable, secure transaction oriented business applications.” …
I got my start on mainframes, punch cards, magnetic tape reels, magnetic core and all that. I learned to program in Fortan II on an IBM 7040 and soon after, started programming on the newish IBM 360s.
I hate to see them go but they are now a far cry from the “Big Iron” of that day. Now the really Big Iron are the supercomputers with 19″ rack after 19″ rack in rows. Physically as big or bigger than any of the old mainframes.
August 12 is the the 30th aniversary of the introduction of the IBM Personal Computer. A machine that changed the face of personal computing see:
“Its the oldest still working Seagate drive in the UK”
Seagate reckons it has found the oldest working Seagate disk drive in the UK, a 28-year-old ST-412 disk drive from 1983. Here
I remember these, 10 Megabytes and quite expensive at the time but for disk I/O a significant speedup from a floppy drive. I may still have one somewhere. I am sure I have the later ST-225 20 Meggers.
On Intelligence: How a New Understanding of the Brain will Lead to the Creation of Truly Intelligent Machines is a book by Palm Pilot-inventor Jeff Hawkins with New York Times science writer Sandra Blakeslee. The book explains Hawkins’ memory-prediction framework theory of the brain and describes some of its consequences. (Times Books: 2004, ISBN 0-8050-7456-2)
I saw Richard Stallman (of Emacs and GNU FSF) at Virginia Tech where he spoke at invitation of the ACM. He spoke on copyright, what he perceived as the problems with it and how he would fix it. (Basically he proposed reducing the copyright term substantially to 10 years, a number he thought about right but was willing to say that it should be tried and adjusted if appropriate). He also advocated dividing the copyrightable works into 3 classes and treating each class differently. The first class were reference works, textbooks and like (things you need to do your job). For these advocated essentially a GPL approach anybody could copy or modify and redistribute. The second class were “impressions”, works of opinion etc. These he would allow free noncommercial distribution but no modification. The third class was entertainment, music, video, art etc. These he would allow free noncommercial distribution but was ambivalent about modification saying there were good arguments to be made for modification and nonmodification (for example to preserve artistic integrity). Obviously he covered more, a few political comments some of which I agreed with and some I did not. Afterwards he sat down, answered questions, during which he took his socks off. All in all an entertaining and thought provoking session which perhaps got off to a slow start.
http://www.jaymoseley.com/hercules/ Hercules an IBM S/370 and ESA/390 Emulator than runs on Intel architecture. Of course it needs an operating system. Some of the older MVS, MVT, VM 370 R6 operating systems are in the public domain, but not the manuals. (Interesting that they are in the public domain perhaps they were not copyrighted at that time because it was believed that software was not copyrightable/patentable?
The second programming language I became proficient in was SNOBOL4 which stood for “StriNg Oriented symBOLic Language” or something similar, a bad case of mangling a name to come up with a cool acronym. SNOBOL4 was a really cool language. It is a string manipulation language with a very powerful pattern representation datatype and pattern matching which is quite a more powerful and easier to use than regular expression such as are used in Perl and grep. Snobol had tables. Tables were like arrays but the “subscript” was a character string instead of an integer. (Called hashes in some other languages). It was an interpreted language implemented as a macro processor and could sometimes be rather slow. Later a true compiler called SPITBOL which stood for ‘SPeedy ImplemenTation of SnoBOL” another name mangled to get a cool acronym. Spitbol produced quite speedy programs. Snobol inspired several follow ons, SNOCONE was a preprocessor for C that implemented Snobol. An implementation of Snobol patterns in Python called SnoPyl. The developer of Snobol, Ralph Griswold went on to implement another language Icon Icon never enjoyed the same degree of usage as Snobol probably because while it implemented a number of improvements over Snobol (more modern control structures for example), it implemented a different pattern string manipulation system. This pattern implementation may have had theoretical advantages over that of Snobol, but in practice it never became as wide spread as that of Snobol. I remember seeing recruiting ads for the NSA looking for CS majors/programmers with knowledge of Snobol, thinking about it Snobol would certainly be a natural to use for some aspects of code breaking.