Pat Farrell

+ Follow
since Aug 11, 2007
Pat likes ...
Mac OS X VI Editor Linux
Merit badge: grant badges
For More
outside Washington, DC
Cows and Likes
Total received
In last 30 days
Total given
Total received
Received in last 30 days
Total given
Given in last 30 days
Forums and Threads
Scavenger Hunt
expand Rancher Scavenger Hunt
expand Ranch Hand Scavenger Hunt
expand Greenhorn Scavenger Hunt

Recent posts by Pat Farrell

I'm not really here, just popped in to see if folks are still trying to use Java and failing.
I do see a thread that has a number of posts from me back 9 or 10 years ago, the one on betting on emerging languages.

Clearly one could argue that Swift is emerging, or perhaps that it is just Objective-C-with-better-syntax. Its so popular in IOS and IOS is so popular that it probably is not really emerging.

It is good to see non-SQL databases. For decades, it was heresy to hint that SQL wasn't prefect.

I spend more time these days riding dirt bikes, rather than writing code. My preferred language these days is Google's go.
Mostly because I have always preferred being more at the system level than application level.
1 year ago

Paul Anilprem wrote:You are mixing two things. Banks deal in moving money. Money, which is produced by the govt. Innovation in money (manipulation i.e.) is the domain of the govt. Banks don't and can't do that. But they definitely do innovate in how that money touches people.

Banks don't move money, computers move money. I worked at CyberCash, we invented internet commerce. We made instant movement of money between merchants and consumer banks happen. The banks hated it, because they took 3 to 7 days to "move" the money, charging interest all along the way. While CyberCash is long go, PayPal is here. Most folks pay by electron, my house pays nearly all bills automatically, there is no need for human interaction at all.

What do you mean about how money touches people? I see zero innovation in banking at all. Woweezowee, US banks added smartchips to credit cards in the past year or so, something Euro banks had last century.

Please be specific, I am not following your generalized claims of innovation
6 years ago

Paul Anilprem wrote:Well, nothing wrong with having more number of banks. Whatever work banks do, they are paid for that work by the customers. More banks means more competition for that work and therefore better rates.

I strongly disagree with this. Sure, in a pure market world, more competitors means better rates. But banks don't compete on rates. And consumers have no clue what they are buying or how to compare costs. US banks are essentially an appearance of a market, with a 10,000 leg octopus, each leg having a sign indicating that its unique. But its more of a Borg.
There is next to zero innovation in US banking. They are all sheep, doing what the other banks do
6 years ago

Paul Anilprem wrote:Minting of bitcoins is a small part in the whole scheme of things. Its power comes from the fact that it cannot be inflated or deflated by any one. Second, it has no dependency on banks.

I can't understand why we still have banks. It is not clear to me what they do. I have one for two things: I have a safety deposit box for keeping critical paperwork like my marriage documentation, deed to my house, heirloom jewelry, etc. and they accept cash deposits. My broker-checking account doesn't offer those two services. Other than that, I don't see any value in what they do.

I also do not understand why the US has so many banks. Back when I worked with banks a lot, in the late 20th century, we had about 10,000 banks in the US (bank brands, not offices) while all of Canada had six. So if six is enough for Canada, its hard to see why we need more than say 100 in the US. Why do we have a hundred times that many? What unique offerings do they bring to the marketplace?
6 years ago

Campbell Ritchie wrote:Until about 100 years ago, money was backed by tangible assets; if the Bank of England wanted to have 1,000,000 £20 notes printed, they had to have £20,000,000 worth of gold stashed away somewhere. The King, if he decreed anything, didn't so much decree money as decree taxes to raise the money, which meant somebody else had to have the gold.

If you sprinkle the word "usually" and "approximately" thru that, I'll agree. It was rare for money to be backed up 100%. Google "Seigniorage"
the king could also mix in base metals to devalue the coins.  The history and the topic itself are not quite as black and white as a lot of folks think
6 years ago
what makes it better (or worse) than any other money? Sure, in the US, the Fed just invents money. In the olden days, they had to run printing presses, buying ink and paper. Now they just decide it exists and change a few zeros and ones, and voila, money. Before the US existed, Kings would decree money.

Bitcoin is IMHO a cool idea, a prototype, that reflects fresh thinking. The blockchain idea is far more interesting to me than bitcoin itself, and there are lots of other block-chain based moneys.

I'm always amazed that folks complain that if you lose your bitcoin wallet, you lose all your money. That is true of US cash as well. Lose your wallet with a bunch of $20 bills and they are gone.

I guess since most modern money is just zeros and ones floating thru the cloud, people have different expectations.
6 years ago

Junilu Lacar wrote: Are they too "That's what a Java programmer would do but Go programmers don't do that kind of thing."?  I'm curious because when I was learning Python, I'd often see comments like "Python is NOT Java!" and took that to mean "We don't do that kind of thing in Python!"

Well, GO is not Java, either. And I think the Python folks are really saying "that is not idiomatic Python"
not that "we refuse to do that kind of thing in Python" or even "no one does that style of code in Python"

I'm no GO expert, but yeah, I think your code is still not idiomatic GO. From what I've seen (and I can't write good idiomatic GO either)
most good go is more functional and makes more use of channels.  For example, this video from Rob Pike shows using channels for coroutines. The usage is subtle and nearly invisible.
Rob Pike on channels for coroutines

the designers of go are very smart and very experienced. They left out a lot of stuff on purpose, like the focus on objects and inheritance. Based on my Java time, I think that the OO fad has run its course.
6 years ago
Sorry to jump in with a negative comment, but the code posted looks like Java code that is accidentally written in GO. Its a long way from idomatic GO

For example, GO code doesn't use sleep, it uses channel. And the mutex usage looks straight out of Henry Wong's Java Threads book.

Don't feel bad, most folks write in their old language when learning a new one. But to see GO's strengths, you have to write idiomatic GO.
6 years ago

Bear Bibeault wrote:

kavin savvy wrote:It is possible of all the talk about GO replacing Java won't happen in web space on the server side (maybe need more time).

I think it is a complete given that this will never happen.

I have been looking at GO, playing with it, etc. for maybe six months. I am by no means an expert. I did spend about 20 years doing Java web stuff.

I don't see any way for GO to replace Java for things like servlets, JSP, etc. The design aim of GO is very sharp and specific. Kinda like Java's was in the mid-1990s. But Java has grown, layers of stuff added on to it, over and over. Its giant, and frankly a giant mess.

I'd love to see a new, modern language be invented to replace Java. But I don't expect it in my lifetime.
6 years ago

Paul Clapham wrote:They cost $450 at the time, if I remember right.

About that time, tuition for a term at my University was $300. so $450 was an insane amount of money
8 years ago

chris webster wrote: Still, mustn't grumble, eh?

You must keep calm and carry on.

Trying to get a tad back on topic, anyone have nostalgia for any early Brit computers? After all, they had Alan Turing himself right there. I know they had some early machines, maybe named Colossus?

(and I mean earlier and more powerful than a Sinclair)
8 years ago

Winston Gutkowski wrote: They heated the aircraft - fairly important at 70,000 feet - and they have better tolerance to temperature extremes.

Which is good at 70,000 feet any time of year, but really leaves a lot to be desired in the summer over Georgia or Turkey at low altitudes. I think most airplanes have a lot of sources of heat.

My guitar amp is tubes, cause they rock.

The first computers I used did not have valves, but they did have discrete transistors and miles of wire-wrap. In 1973, I used a new computer built of "integrated circuits" rather than individual transistors.
8 years ago

Winston Gutkowski wrote:Back in the mid-70's, the US managed to capture a 'Foxbat' - the latest Soviet interceptor, reputed to be capable of Mach 3 - virtually intact, when it's pilot decided to defect; and the American techs were surprised to discover that it still used valve technology for most of its systems.

1) valves are much better for high power RF radios
2) there was a theory that in a nuclear war, the EMF would stop all semiconductor circuits from working. Valve stuff would keep working. Said the threory
8 years ago

Stevens Miller wrote:BYTE? If you do, and you want a dose of nostalgia, don't type "" into your browser. Where it takes you is kind of depressing.

There was a time when BYTE magazine was the size of a phone book, maybe 3 or 4 inches thick. I think it, or one of the other main magazines came out twice a month because they had so many ads it would have been impossible to lift a copy.

All those ads paid for a lot of very good journalism.

Some time in the mid-90s, probably during the Intel Pentium era, the PC evolved to a commodity. Then the hardware industry consolidated to a few brands, Dell, Compaq, HP, IBM/Lenovo, and all the other vendors disappeared. No need for lots of ads, so the magazines shrank. Once the internet destroyed print media, BYTE plodded on, but was tiny and they could not afford to pay real journalists, so the stories were edited press releases.

The website and domain, have been sold a bunch of times in the past decade or so. Each time, someone thinks there is value in the brand, and implements some half-baked vision. The result is very depressing.

Of course desktop and laptop PCs are dead. All the cool kids are into smartphones and watches and Internet of Things.
8 years ago

Stevens Miller wrote:anyone remember the Rainbow 100 or Pro/350?

My group of bought about three of the Pro-350s and we bought hundreds of Rainbow 100s. They were interesting "personal computers".
The Pro-350 had a PDP-11 in its system unit, the Rainbow had a dual-processor system unit, a Z80 and Intel 8008. They had the same monitor and keyboard. They were designed at the time when the S100 bus was "the standard" and released about the same time as the IBM PC (original). The Z80 CPU was what the S100 bus computers used, and it ran CP/M. The Intel 8008 side ran Microsoft PC-DOS, as a safe bet incase the IBM PC actually got traction.

The Pro-350 with a PDP-11 was a massively more capable machine than any prior personal computer. And there were millions of PDP-11s deployed in all sorts of industries, with huge amounts of PDP-11 programs.

At this time, many vendors sold Intel 8008 based personal computers running PC-DOS, but you had to buy a specific version of the programs for your branded PC, there was no such thing as a "IBM PC Compatible" system. Then Compaq figured out how to reverse engineer the BIOS and used the fact that IBM corporate thought so little of the concept of a personal computer that they allowed the microcomputer group (in Florida, not Rye NY) to release the engineering specs of the bus and boards. So Compaq released the first "IBM compatible" personal computer. There were a few lawsuits, but soon lots of companies were selling "IBM compatible" machines, and the industry was radically revolutionised.

Actually, there was one more step. There were a lot of applications, such as DBase or Visicalc. They ran about the same on all of the competing personal computers, including the Rainbow 100. Then Mitch Kapor wrote his Lotus 123 directly to the hardware of the IBM-PC, ignoring the DOS drivers for graphics, disk, or keyboard. Since his code was specific to the hardware, and skipped all the general stuff, it was way faster than the other brands.

Sadly, for brands selling generic PC-DOS systems, speed mattered. Soon the market collapsed into only "IBM hardware compatible" systems. Killing off the Rainbow 100 was not worth crying in your beer over, but IMHO, the PDP-11 based system could have been cool and a huge improvement over the original IBM PC and even the brain-dead Intel 80286 in the IBM AT.

Note: I am not slagging the Intel 80286 unfairly, in the mid-1980s, Intel put up billboards all over Silicon Valley and Los Angelos advertising that the 80286 was braindead -- the suggestion was that modern systems should use the new Intel 80386 and not the old 80286.
8 years ago