From owner-sigcse-members@LISTSERV.ACM.ORG Tue Jan 18 14:53:52 2005 Sender: SIGCSE Member Forum From: Bob Lechner Subject: Re: [SIGCSE] Is it CS/Engr "soup" or Engr/CS "art" To: SIGCSE-MEMBERS@LISTSERV.ACM.ORG RJLRef: $PH/COOL-FAQ/isRDBfromCSorSWEng050118.txt I totally agree with Scott. But I believe Relational DBMS's belong on the SE side as well. The predecessor of RDBs was Bachman's Integrared Data Store, the (exceptionally well-engineered) first network database. Bachman the engineer also invented the Data Structure Diagram, to which generalization was added by the Smiths. I watched Peter Chen learn DSD's from Bachmann at HIS before Chen proposed his ERD cluttered with diamonds and ellipses. RDBs bungled along until the mathematician EFCodd provided its set-theoretic foundations. Bachmann, Codd and Chen did not win Nobel prizes, but all three won ACM's Turing award. RDB's are one of many examples of invention preceding explanation, and complexity preceding new theory. In other words, art precedes science. As another example, consider Microsoft Windows, whose first version was developed by (mighty crafty) hackers. It grew big much faster than it grew smart. But this month's Microsoft Academic Days revealed tremendous R and D team depth at MicroSoft Research. The guru's on such teams surely defy CS.vs.SE sub-classification. Bob Lechner CS Dept UMassLowell lechner@cs.uml.edu > From owner-sigcse-members@LISTSERV.ACM.ORG Tue Jan 18 13:34:05 2005 > Reply-To: Scot Drysdale > Sender: SIGCSE Member Forum > From: Scot Drysdale > Subject: Re: [SIGCSE] Is it CS/Engr "soup" or Engr/CS "art" > > > I've heard it said -- mostly in jest -- that Comp. Sci. gave us LZ > > compression, Relational DBMS, Binary Search etc. while Sftw. Engr. > > gave us Google. > > That is interesting, considering that Google was founded on an > algorithm (the page-rank computation) which is being continually > updated, and some fairly deep algorithms work is used in building > the index (parallel-processing using hundreds of machines for many > hours, and how do you organize it so that if a machine drops out > the whole thing doesn't have to be re-run?), networking (millions > of inquiries being routed to tens of thousands of computers), and > other areas. Talking to former students who work there and interviewers > who come to campus gives me the impression that Google is a great > consumer of CS results and talent, and is very interested in hiring > people with a strong CS background. > > Scot Drysdale >