I would propose we move the most advanced and critical equipment and associated personnel first. Even if this involves rocket science its not impossible, here in the USA we do rocket science. I would not move the buildings, coffee machines, soda machines, barrooms :-) Yes , this is a significant part of Taiwan's economy. Remove the technology makes Taiwan less attractive. Bold action is called for in the face of coming war.

As long as we are throwing money at this. Perhaps it would be prudent to bring all of TSMC, people and equipment, here to the US. It would involve some immigration, but its the good kind. It might be the best way to help keep it out of Chinas hands.

Ah, there lies the rub. Privacy protection from aggregating in the past was provided by the volume of data and the cost. Now it is less expensive. I favor a constitutional amendment to protect against it. Especially with respect to advertising, politics, and medical.

They must be following Boeing management philosophy: you don't need engineers, just MBAs cutting costs an driving up the stock price. That's how you get amazing engineering results!

OMGosh, Just fund the blue states. Tell customers in red states, they asked for it. Red state congress members will approve right away, am I wrong?

OMG! If you actually *measure* performance you can give concrete results to employees. To do this you must have specific goals and a measure for evaluating each goal without MBA-bull like "better communication with collogues". You were assigned to complete these programs, modules, routines, what have you and you did that, good. You created X bugs and fixed Y bugs. etc. etc. Nice thing about this is that you turn around and pass this up the chain in summary form and looks good for the whole group review because the higher ups love to see number even if they don't know what they are really saying. Add some min, max, and averages, maybe a chart and you have them amazed. Management can be managed sometimes, but only sometimes. I recall decades ago being in a high level meeting which asked how we can get more minorities and women hired into engineering positions? I asked why not hire every minority and women engineer that applies and find out how they do. There are so few of them it can't hurt and chances are they had to work so hard to get there they ought to be good. Oh, we could never do that. Even when you looked at the numbers of women applicants vs the hours interviewing all new candidates the numbers looked good. Too radical. I once made a job offer to a guy without consulting HR that had been rejected by all the other interviewers, it made HR quite angry, went all the way to the CEO. My answer to the CEO: anyone with that GPA from Stanford should be hired on the spot. The guy was an outstanding hire. These HR folks do need to keep changing it up so as to keep there jobs (how may throw-away slogans did you create this quarter?), just smile and get back to programming; you can always use feedforward as something to laugh about over beer.

My practice was, wherever possible, if you compile a component with RUN_TEST define you get an executable that tests the component. Sometimes a subdirectory needs to be provided with input data. The documentation then includes the output of a test run. The code reviewer also looks at the test. Results from a test engineer' review are then are given back to the developer and their supervisor so that inadequate tests are detected and learning can take place. When UI is involved you can rely on standard UI script test engines that test engineers maintain. Old school, I guess.

I remember PL/C with nice string handling and garbage collection. I made an 360/40 timeshare a few terminals with 80 col editor and job submit using virtual card reader. Perfectly fine language, where did it go. I think C was just to portable.

Cruelty is the Cornerstone of modern consrrvatism IMHO

Time was when you dialed in the channel and unit number and hit the IPL button. Who remembers putting in the address of tape controler and a read command to boot the UNIX distribution tape? Oh, these new fangled computers!

While in high school in San Diego we had access to an unfriendly 360/20 and a wonderful 1620 at t he local IBM office on weekends. I wrote my first Fortran program to accept equations at the console and rerturn results. Loved when I typed "X=1/3" and "X" to have it respond "0.33333" late one night. In my garage I never got much beyond three flip flop circuits built from CK722 transistors, doen't count as a computer. Most fun was running XENIX on M68000 in 1980, my favorite PC.

Many students get an education in Algebra and Trig that is cookie cutter: you solve this problem this way, etc. They "pass" those courses without a full understanding of the subject in a way that makes it difficult to "get" calculus. The problem extends all the way down to basic arithmetic in US schools. Theres no point in advancing to Algebra if you can't do basic add/mult/div or understand magnitude. Culture also to blame with allowing "I just can't do math" is ok.

Switching back from "personal" computing to "time share" as we have done and called it "the cloud" is THE tragedy of our time. Sure, some things need centralization but surely not everything. Yes, some big hardware is needed for big problems. The real motivation for this giant step backward away from personal computing is money. Also, I think, fear; take e-mail, I can't send email from my computer to my neighbor without involving some mail server in the cloud, and that is because server technology is "too complicated" for the average user. Blame the user, no blame the software vendors. [FLAME ME? I know that my desktop can send and receive email thanks to having a domain name of my own and a mail server, but it's not trivial at al].

What ever happened to cold fusion?

It is wise to pay the amount of tax owed by law. Few (if any) pay more than required. Change the laws to change the tax these guys pay. But be careful that what you wish for may not be the outcome.