1/5     Phew!  Quit that job just in time!
        http://biz.yahoo.com/ap/050105/transmeta_stock_2.html
        --lye
        \_ tell us about transmeta, like was it cool while you were there?
           is the technology really that cool? what did you work on?
           \_ The technology is very cool, but it has become a solution for
              a problem that doesn't exist.  The Pentium M essentially makes
              Transmeta's hardware moot.  The amount of engineering labor
              required to make code morphing really work is immense, and
              it is the kind of labor that PhDs are not fond of.  TMTA is
              staffed with lots and lots of PhDs.  I'm sure you can imagine
              the sorts of management drama that resulted from that.
              I myself worked on x86 verification, which essentially meant
              writing lots of directed tests in x86 assembly that tested
              thousands of corner cases.  In x86, virtually everything is
              a corner case due to legacy support, especially in floating
              point.  I also wrote and managed random testers.  I Learned
              A Lot, but it was Not Fun.

              I also learned that Linus Torvalds is a very nice man who drives
              like a complete maniac.
              --lye
                \_ what kind of car did he drive? and is he a nice guy?
                   \_ Yellow Mercedes Benz SLK32 AMG.  --lye
        \_ Did Ditzel go down in flames too?  I see that he's making $411K
           a year, but did he manage to sell his stock in time?
           \_ Ditzel was well equipped with all manner of golden parachutes.
              While I was there, he typically unloaded 100s of thousands of
              shares a month, which he was optioned at a very low price.
              This is all a matter of public record, I believe.  --lye
        \_ I think this is a very good move, even if they don't make it.
           The code morphing is very cool, I have always thought this would
           be a great match for a next generation of high performance chips.
           A brilliant way to help move away from the aweful x86 ISA. -jwm
           \- you're an idiot.  -tom
              \_ In case you were wondering, this response has gotten old, at
                 least for me. -- ulysses
                 \_ Someone (probably crebbs or erikred) has decided that
                    it's amusing to post that in my name.  I certainly
                    wouldn't use psb's "\-".  And I don't think jwm is an
                    idiot.  -tom
                    \_ dude, fuck off.  What the hell is wrong with you anyway?
                        I didn't post with your username.  It seems obvious
                        that erikred didn't either. Nor did I think it amusing
                        when someone did it in the middle of our thread.
                                        -crebbs
                    \_ Doh! I should have thought of that. -- ulysses
                    \_ tom, I promise you, I have never nor would I ever post
                       under your name. I have a soft-spot for loveable cranks,
                       so I'm not likely to alter or delete your posts, either.
                       Keep on trucking. --erikred
              \_ Tom is right on here, although he probably didn't need to
                 resort to insults.  Code morphing is "cool," but essentially
                 useless.  Is there a word for technolgy like this?
                 Slashdotware? --lye
                 \_ I recently suggested to google scholar that they add
                    an Erdos number calculator. -- ilyas
                    \_ That would fucking rock.  Could you post if they
                       respond?  I wonder if they would be more likely to do
                       it if they got more requests for it.  The only time
                       I sent a suggestion to google(about their calculator)
                       it was totally ignored.
                       \_ What was your suggestion?  I might be able to talk
                          to someone who can get it done.
                       \_ I made another suggestion to scholar a while back
                          that got put on a feature list, apparently.
                            -- ilyas
                 \_ I respect your opinion on this because you know
                    a great deal more than I do about this.  To me it
                    seems that it could be very useful, and in fact
                    practical in the sense that it could be used to solve
                    real problems.  But I am very interested in your take
                    on it as someone with experience.  --jwm
                    \_ It's a very long story - but essentially code morphing,
                       no matter how good it is at individual benchmarks,
                       never "feels" like "real" hardware from the native
                       instruction set that you are emulating.  Translation
                       time lag always leads to a jerkiness and the feeling
                       that the computer is "shifting gears."  This is on
                       top of the fact that getting the translator to do the
                       right thing in all cases while optimizing for
                       performance is extremely hard.  --lye
                 \_ Java is the term I use.
        \_ So what did you work on at Transmeta?