Comments by "TeeKay" (@teekay_1) on "Asianometry" channel.

  1. I had originally taken my company to Solaris over AIX and HP/Unix because the hardware was a fraction of the price (although the HP hardware was really much better). Once Oracle bought them, we greatly accelerated our move to Linux, moving an entire major corporation from Solaris to Linux in 18 months.
    11
  2. The electrolytic capacitors in 1970's era stereo equipment are generally in good shape despite being 50 years old.
    10
  3.  @sv1201  They probably knew they could get their boss to pay it out of his slush fund.
    4
  4. @ I've recapped a bunch of receivers, and generally all the caps I've removed are near spec. Nonetheless, I recap with high quality Nichicons and in most cases the transistors in the preamp and the transistors in the power supply (for obvious reasons). I don't touch the transistors on the amplifier boards because in many cases they're hard to match.
    2
  5. I believe the 68000 had 32 bit instructions, but it 's data and address lines were only 16 bit (which is why most people still consider the 68000 a 16 bit computer). It wasn't until the 68020 that the 68K line became truly 32 bit, and even then it required the 68851 MMU to do hardware process protection. I also seem to recall that Sun did a 80386 version, but it didn't' make a lot of sense since it required different binaries, and the x86 market wasn't clamoring for a unix variant, since OS/2 was the hot ticket when it was released.
    2
  6. When this was first announced (the merger of Honda and Nissan), most people said "I hope Nissan becomes more like Honda rather than Honda becoming more like Nissan", which is somewhat unkind, but you understood the sentiment was roughly correct.
    2
  7.  @sillystuff6247  Sun had a great vision, but the embedded Java market never took off for two reasons.... embedded hardware wasn't fast enough to run it effectively, and the run time licensing was pretty expensive compared to 'free" for embedded "C" being used at the time. The only place I'm aware it's used today is in Cable Cards and BluRay players. Then Oracle came along and screwed Java up. Now Javascript rules the front-end, and back-end Java is still free because of the available from Google & Amazon have their own versions/
    2
  8.  @nasarazam  I always felt Borland's stuff may not have been as sophisticated as MS's stuff (particularly the class browser), but it was tight, and it didn't require 6 CD's to load.
    2
  9.  @calkelpdiver  I recall that Novell did worse; they bought Unix, which was actually free. Novell also screwed up coming out with Netware IV, which was a great product, but their customers didn't want it, and all their certified guys had no idea what to do with it. But yes, Borland screwed up when they bought Ashton Tate. Similar to Novell, their heyday was dBase III+, and moving to IV, their customers really didn't want it, and the developers for dbase III didn't understand it.
    2
  10.  @nasarazam  No doubt, and a lot of people did. It wasn't 100% dbase III+ compatible though, and even thought it offered a lot better speed, and better programming tools, it was essentially a niche. As I recall, MS bought Foxpro as an alternative where it promptly died after a year or two (although it looks like it's still supported).
    2
  11. Correct on both counts. I remember the BYTE magazine review of the Sun 80386.
    1
  12. Instead of 1's and 0's, soviet computers used "Da" and "Nyet". Jokes aside, the Russians always excelled at Math, but that's not enough to build a computer. You need the industrial and semi-conductor engineering to build fabrication plants, and improve computer speed and accuracy. My guess is they fully understood how to implement things like Euler/Improved Euler/ Runge-Kutta methodologies on computers long before they actually had computers to use them on.
    1
  13. Agree. I believe I still have a Borland x86 assember package (unopened) around here somewhere.
    1
  14.  @sillystuff6247  It was initially; there were several significant apps written in the early days of Java. My point is that Java's premise of write once, run anywhere only came about because Sun gave the whole thing away. As to Oracle, no, Oracle is not a reasonable company. If you find yourself choosing Oracle for anything, it means you have a basic misunderstanding of the Oracle business model. I will give Oracle credit for one thing; by taking Google to court over Android, they accidentally established precedent that you couldn't patent or copyright the verbs used in any program language. Finally as to the JRE and JDK, both of those are available legally to any company because the source code is available to anyone, and if you use the Amazon, IBM or Google versions, you'll be up to date.
    1
  15.  @leland818  " science is separate from nationality or politics" If only that were true. When weak men proclaim themselves as "The Science" and scientific illiterates claim "I believe in the science", then we've moved back to the 16th century.
    1
  16.  @stephenhill8790  Well, Communist Soviet Union was pretty sh*tty to it's people, so it's not something to celebrated.
    1
  17.  @spartacus17.  hahahahahaha. That's funny.
    1
  18. I think you'll find that motherboards of garden variety PCs are the least of the worry about electrolytic capacitors; usually they get too slow before capacitors become a problem. It's more of the embedded devices and power supplies that fail because of capacitors (particularly switching PS's). They're not hard to repair, but once they fail, consumers usually buy another one because board level diagnostics are more than the value of what needs to be fixed.
    1