Misc. Technical Ramblings and Nostalgia - share yours here

[khj]

Wow, I just realized that my history professor actually had a 500+ word obituary written for him in the New York Times. I knew he wrote articles and OpEds for NYT. He even has a very short Wikipedia bio that starts off:

Francis L. Loewenheim (June 27, 1927 – October 17, 1996) was an American historian at Rice University. He was one of the leading authorities on modern diplomatic history.

Turns out he actually edited a book on the Munich Crisis. Glad I didn't get an F for not mentioning it in my paper!

[zepherusbane]

Jan 23, 2019 17:31:13 GMT -5 angore said:

I remember punch cards...oh the fun, running jobs late at night to avoid the crowds, etc.

I missed the punch cards, I'm just a little too young to ever have had exposure to them. I never even had to deal with any mainframes until, ironically, my current job I started 2.5 years ago. We have an IBM mainframe still running a bunch of stuff.

Jan 23, 2019 21:17:46 GMT -5 mourningdoves said:

Eventually, a guy I knew sent me a pack of discs with a bootlegged WordPerfect 5.1, and got on the phone with me and told me how to install it. He also told me how to set some buffers and whatnot so that I'd get it to run really well. I believe that his motive was that he was a writer for a zine I published, and he wanted me to get computerized so that the zine would look better.

WordPerfect sat there for a while, mystifying me, but for a while when I was unemployed, I just sat there for a few hours a day and learned WordPerfect. It sounds funny to say it now, but it was actually one of the more profound experiences I've ever had. I could erase all my mistakes without leaving a trace and without going through reams of white-out! I could use 30 different typefaces in every paper I ever wrote! I could type resumes and term papers for my friends and they'd give me beer! It actually made me feel like I wasn't alone.

I went through a series of life changes in the mid-1990s and it wasn't until I emerged back into the work force as a free agent that I learned Microsoft Word had taken over the world and eaten everybody's brain.

In the early '90s while I was still in the Army I learned Wordperfect and got pretty good at it. Around 94 or so we were forced to switch over to Microsoft Word and I remember how pathetically bad it was compared to Wordperfect. You couldn't even explain it to a non-WordPerfect user because they had never even imagined some of the features that normal users of Wordperfect had learned to take for granted.

Jan 23, 2019 22:17:28 GMT -5 mourningdoves said:

Jan 23, 2019 21:47:52 GMT -5 khj said:

I remember that!!! When the full screen editor EMACS got ported to our DEC/VAX, I was so relieved!

I used emacs for several years at my last job, and I loved it. Oddly, when I left that job and moved back to Windows from Linux, I installed emacs for Windows, but even though it was a pretty good port and I could move over my dot.emacs file and my word lists & recorded macros, I soon lost interest, and I hardly think of it any more. And it was like my life for years!

In my work, I had to learn unix early on and spent most of my early career writing software to run on various flavors of it such as SCO unix and HPUX. I was all about vi for editing my source code, never once did I ever use emacs.

Back in those days I was mostly writing C and Perl code, later I moved to Java. Today I'm using mostly R, Python, and Julia for languages but even so, I still use vi today, just yesterday I was using it to edit some python code on my mac. It's funny how all the keyboard commands just stay in your muscle memory after 30 years. I couldn't tell you what the commands are, but my fingers remember when they are on the keyboard. I'm way faster in vi than in anything that requires the use of a mouse. I have a tendency to use vi in a console window to work through getting certain types of things to work that lend themselves to lots of trial and error kind of code changes, then copy the code into my juypter notebook or RStudio or whatever more modern toolset I'm using.

[mourningdoves]

Jan 24, 2019 10:38:38 GMT -5 zepherusbane said:

In my work, I had to learn unix early on and spent most of my early career writing software to run on various flavors of it such as SCO unix and HPUX. I was all about vi for editing my source code, never once did I ever use emacs.

If you think the WordPerfect vs. Word shouting matches were bad, vi vs. emacs was like the Hundred Years' War. I remember being at a Linux meeting in Providence and getting to know a guy who worked for Equal Exchange, the fair-trade coffee and tea company that works out of southeastern Massachusetts. I was babbling away about emacs and it took him like five minutes to admit, in a halting voice, that he used vi. I said, "I've heard vi is really great, but they're both unique, so I guess a lot of it's what you get used to." He was stunned that I didn't slug him. We got along great after that.

Emacs uses a lot of "chords": keystroke combinations and sequences that, if you transcribed them into musical notation, would look like Shostakovitch getting into a bar fight with Frank Zappa. Vi has that mode-switching that some people couldn't get a handle on. When I first joined the Internet, I had to use one of those email clients, Alpine or something, that had modes, and I didn't have an instruction book, and it drove me bonkers. Again, a few hours of training & practice, and it probably all becomes second nature. Those old-skool text editors work faster than I can type, something I'm not sure I've ever experienced in any Windows program.

I think the only time I cried in front of a computer was trying to do something in Word 97 and no matter what I did, Clippie popped up. Or were those bitter tears from trying to do a mail-merge in Word 6? I'm convinced that the only reason Word got so popular is that Excel really was better than Quattro Pro or Lotus, and number-crunchers tend to be higher up in the corporate hierarchy than letter-bangers.

[JerryB]

Hi

I usually worked on mainframes. The first was in college and was an IBM 1620. No disk storage. To run a program, you first loaded a stack of punch cards which was the compiler and then a stack of punch cards which was your program.  After a few seconds, or minutes, it printed an answer.  The printer was essentially a typewriter in a big box.The machine was top end at 16K memory, most were 8K 1620

The first micro computer was a Radio Shack TRS80 (affectionately called a Trash 80 by mainframers).  I wrote a full screen editor for it, In either Basic or assembler, I forget which. TRS 80

Some of the input devices I have used are patch panels for programming IBM accounting machines, paper tape, punch cards, magnetic tape, and all types of disk drives and disks.  The TRS 80 used cassette tapes.

When I switched from electronics engineering to computers, I worked for an architectural firm that had an IBM 1130 with a whopping 32K memory.  It also had removable disks that held 512K 16 bit words of data (you could have 2 of those or two 2315 disk packs).  That machine was essentially created for the scientific community and engineers.  Fortran I and assembler code were the only languages.  The printer was state of the art.  It had 120 print wheels.  The wheels spun until the characters were aligned and then the entire print head "thumped" against the paper.  Once a week we moved the printer back to its starting position as the printer moved from the printing action 1132 printer    1130

Believe it or not, these machines were lightening fast at 5 - 10 millisecond clock rates (today measured in nano or micro seconds).

Today everyone talks about Gigs of memory or disk storage.  In yesteryear, that would be 1 or more rooms of disk drives.  
Computer memory, for a large IBM 360, was anywhere from 8K to 64K , not all usable since part of that was for the operating system.  

Today, people don't realize we went to the moon with that technology and large corporations ran their businesses with that technology.  With such small storage devices and computers, developers were ultra clever in "crunching data" for storage and developing applications with a minimum of code.   

As a junior and full electronics engineer, I worked with micro-chip development and am totally amazed in how short a time technology went from transistors and a few devices on a 2 inch silicon disk to today's micro-processors.  My wife's smart phone has more power than some mainframes had.

Oh well, enough reminiscing for now.  As I look back, it was interesting to see the advancement of technology.  I am sure people in the 1800's felt the same way: horse and buggy to automobiles, radio, telephones, etc..

Added Note:  I have seen ads whereby one can  purchase old mini computers (now called mainframes) even today.

Jerry B

[angore]

In my college days, we just had the college mainframe. This was the beginning of the calculator phase so no slide rules like in high school.

There was really no such thing as a personal computer and everything was done on typewriter. The computer science guys had some interactive workstations.

As an engineering student, I learned PL/C (easy version of PL/1) rather than Fortran when I did finite element work and other required tasks. When I started work, we used IBM 3270's (VM) attached to IBM 360's. This was primarily for word processing and email with no real computing needed for what I did. It was not until the IBM PC came into existence. The 3270's were replaced by an IBM PC with a 3270 card so we could be local or attached to host.  Word processing was all done on the host using IBM Book Master (a markup language) and this continued well into the late 90's.  We would edit on PC then upload to process to print on IBM 1403 printers then later IBM 3800 printers.

When people started using Word the documents started to vary dramatically from person to person since it was all WYSIWYG and styles were not required to be used unlike using the mark up language/

[Gordon Lee]

Greetings Fellow SB Boarders
.
"This 'telephone' has too many shortcomings to be seriously considered as a means of communication." ~ Western Union internal memo, 1876
.
"Heavier-than-air flying machines are impossible." ~ Lord Kelvin, president of the Royal Society in Great Britain, 1895
.
"No matter what he does, he will never amount to anything." ~ Albert Einstein's teacher, 1895
.
"Everything that can be invented has been invented." ~ Charles H,. Duell, U.S. commissioner of patents, 1899
.
"There is no likelihood man can ever tap the power of the atom." ~ Nobel Prize-winning physicist Robert Millikan 1923
.
"Who the hell wants to hear actors talk?" ~ H.M. Warner, Warner Brothers, 1927
.
"There is not the slightest indication that [nuclear energy] will ever be obtainable. It would mean that the atom would have to be shattered at will." ~ Albert Einstein, 1932
.
"The problem with television is that the people must sit and keep their eyes glued on a screen; the average American family hasn't time for it." The New York Times, in an article about the debut of television at the 1939 World's Fair
.
"I think there is a world market for maybe five computers." ~ Thomas Watson, chairman of IBM, 1943
.
"There is no need for any individual to have a computer in their home." ~ Ken Olson, president, Digital Equipment Corp., 1977
.
"640K ought to be enough for anybody." ~ Bill Gates, 1981
.
"I predict the Internet . . . will soon go spectacularly supernova and in 1996 catastrophically collapse." ~ Bob Metcalfe, fournder of 3Com Corp. and inventor of Ethernet technology, 1996
.
You don't say . . .
.
Respectfully,

Gordon Lee

[youpiao]

Jan 25, 2019 14:44:06 GMT -5 Gordon Lee said:

"The problem with television is that the people must sit and keep their eyes glued on a screen; the average American family hasn't time for it." The New York Times, in an article about the debut of television at the 1939 World's Fair

Unlike, say, a newspaper, where one must sit and keep their eyes glued on a sheet of paper.

[zepherusbane]

I think this post qualifies as a "technical rambling", he he.

This past week I was dealing with frustrations trying to upgrade my Linux installation (a heavily modified ubuntu studio with a low latency kernel for audio recording). The partitions I created when I initially installed a couple years back as a dual boot option with windows were not large enough for the next release and the drive where I created the partition didn't have room to expand it. I could only create a completely new partition on a different disk.

This has taught me not to believe what space requirements distros list for base operating system partitions. If I had left the distro mostly as initially built I would probably have been ok, I didn't anticipate that I would start using it as my primary operating system at the time, meaning I ended up adding a ton of additional packages that weren't part of the initial distro. The need to upgrade to a new version also reminded me to stick to a rolling release distribution too.

I found a new version of a different Linux distribution I liked better that came out on January 28th. It's called Makulu Linux Core, based on Debian. The "look and feel" in this distro is amazing. It's also really easy to use and customize despite it being based on Debian instead of Ubuntu which tends to be easier. This isn't like Windows, Mac, or other Linuxes, it's something that kinds of blends much of what I like best of all three worlds - which is great for me since I use a Mac every day for work and have a Windows/Linux mix for my home machine.

I took a 100 gb partition for root and started adding all the same software I had in my ubuntu studio instance. It's new enough that when I installed jack, it asks if I want to use the real-time features that can be turned on in the kernel so I don't need a special low latency kernel to handle audio recording priorities, that was one of the main reasons I was using Ubuntu Studio as my main distro. Out of all the packages that come with Ubuntu Studio, only a few aren't in the Debian or Makulu repositories, and it doesn't look like any of the specific packages I have used up to this point were missing. I was able to install all those packages quickly after installing the OS.

Of course, those of you who use Linux are no doubt aware of how it's a pain to get your menus looking exactly the way you want to and I had to struggle with mine as I always do when first setting up a Linux install. I wanted my audio editing, video editing, graphics editing, etc. tools sorted the way I wanted them, not the way the defaults put them. In Linux, this means a bunch of pain to merge in your custom menu and directory files. I cheated and pulled .directory, merged menu files, and icons from my Ubuntu Studio instance, but despite this, I still had to spend hours tweaking stuff.

Even though I got my audio/video/graphics menus looking right, I still don't have some of the core menu categories showing the way I want. For example, I installed several hamradio related software packages but they are all showing in various non-ham places like accessories, other, education, etc. (not together). Software makers create a file for menus for each application where they suggest categories, but if your menus don't have one that's shared or if the developers didn't categorize things the way you would have done, you end up with software showing in lots of different menus that might not match your expectations. It's not easy like it is in Windows where a blunt force approach makes it easy for the end user by just creating shortcuts wherever.

I don't know how someone who doesn't know how to get around the Linux filesystem in a terminal window could ever customize their menus, you can't just drag and drop stuff around, on a menu, I wish someone would write some code for that for Linux. If I hadn't spent years working in Unix on the command line and later Linux I would never have been able to do it. Also, I find that the vi editor is critical for editing all these configuration files related to the menus, although I suppose someone could use emacs or some other command line editor if they really wanted to.

I'm thinking of trying to create menus for another whole category of software I was thinking of loading too, for security testing (Kali Linux packages mostly), but that will be for another day. Kali has its own set of problems, not the least of which being that for many of the tools on that system to be useful you need a "quiet" operating system or you'll spend all your time being distracted by the noise from your own machine. You also have to add their repositories to get almost all the software since almost none of them are availble in the core Debian or Makulu repositories. Most of those tools have to run as root too, which is really irritating if you are using gui tools. Even for command line interface (CLI) tools, if you sudo and enter your password, it only lasts for a short time before you have to do it over again. Many of those tools get broken by other dependency installs when you are loading all your other software that isn't part of the Kali distro too. I've found when testing security at work that using Kali from a Virtual Machine isn't a real option either as many of the tools simply don't work right when tunneling through the host machine.

I probably need to just create yet another Linux install to boot from, since I'm up to a tri-boot system, maybe going to four isn't so bad. The issue will be finding more space as I'm starting to fill up both drives on my machine.


In case you want to learn more about the operating systems I mentioned in this post, here are some links:

ubuntustudio.org/ for Ubuntu Studio
www.makululinux.com/wp/core/ for Makulu Linux (I like the "core" version, they have others) Watch their video if you want to get an idea of what it's like.
www.kali.org/ for Kali Linux

[JanetC]

My Daddy used to work for Sperry Rand-Univac. He built some of those mainframes that took up a whole room. Sometimes they would have Family Day in which we could go to the plant and see what he did. I remember coming home with a whole pile of punch cards. When I was 16 or 17 I decided to clean out the back of the closet (where I threw everything I didn't know what to do with). I found over 1000 of those punch cards from all the vists to Dad's work.

[JerryB]

Hi janetc

As the song goes, "I remember it well". In the dark ages I programmed a Univac 1100 in Fortran and some assembler code. I also interfaced it with a Tandem mainframe and a DEC computer. In those days, one got a basic machine (usually leased as they were very expensive) and then wrote their own software to make it do what you wanted. Today, everything is off the shelf. Those were the fun days

As a note, I can still read non-binary punch cards.  In fact I still have an IBM template for the task.


Jerry B

[youpiao]

Feb 12, 2019 6:34:36 GMT -5 JerryB said:

As the song goes, "I remember it well". 

Jerry B

I don't remember that song. 

[angore]

I still have a reel to reel tape recorder and tapes I had made back in the 70's, My first TEAC reel to reel died after 30 years so got another one similar of ebay. I never thought cassettes were as good as a reel to reel running 7½ips and much less fragile.

My first receiver was  Kenwood tube amplifier.  It consumed tubes. I eventually replaced it with a Pioneer SX727 receiver.

[JerryB]

Hi youpiao

Have you ever seen the movie "GiGi" en.wikipedia.org/wiki/Gigi_(1958_film)

One of the songs: "I Remember It Well" – Sung by Maurice Chevalier and Hermione Gingold   YouTube

Jerry B

[JerryB]

Hi  angore

My friend was working in an audio lab.  Between us we designed and built a quad amplifier.  He did the audio and I did any digital stuff needed.  We had a mutual friend who built server amplifiers and he built 4 high quality audio amplifier output stages for us.  This thing had all kind of switches and "doodads".  The nice thing was if an interface was need for a piece of equipment, a board was designed and built and plugged in.
It had a minimum of 50w per channel. At full blast you could hardly sit in the same room. 

Today I have a all Technics discrete component (tuner, pre amp/amplifier, tape deck, turntable) and, believe it or not, a Radio Shack 8 track player. Also Bose speakers.  The amplifier is a 100 watt per channel monster.
 Technics is really the high quality end of Panasonic.

The best audio equipment has a tube based output stage.  Try and find tubes today.  The "young uns" have no clue what a vacuum tube is. I think only a mention of them is made in beginning electronics classes.

Jerry B

[JanetC]

Feb 12, 2019 17:50:13 GMT -5 JerryB said:

The best audio equipment has a tube based output stage.  Try and find tubes today.  The "young uns" have no clue what a vacuum tube is. I think only a mention of them is made in beginning electronics classes.

Jerry B

My husband and I are both ham radio operators.  Out in the shed we have a whole bunch of those tubes that used to go into a radio he used to own.  They are not cheap at ham radio flea markets because they are so rare.

[frog]

JerryB I still use a tube amp and refuse to part with it even though it weighs a ton.My Musicman was the first year they took over the Fender line.It says 130 watts but it has upgrades.

[JerryB]

Hi

The power supply for the amplifier was external and was bought at an Army surplus store. My friend had it near his tape deck and found that tapes were being erased due to the magnetic field that supply put out. He put the supply in his basement and ran power cables to the amp.

I bought the 8 track player as my wife's friend gave her a carton of 8 tracks and we needed something to play them on. However, the tapes were old and falling apart. I once worked at CBS labs as an electronics technician and knew how to fix the tapes. I would open the cartridges, unwind all the tape, reattach it  to the hub and rewind it.  Then, I transferred the music to a tape cartridge.  That was easy because my tape deck (Technics) had capability of 2 cartridges and some editing features.  In all, I probably recovered 20+ 8 tracks,  I lost one when the cat started playing with a pile of tape.

Jerry B

[JerryB]

Hi Janet

I once thought of getting my ham license but never did. As an electronics technician I worked at an aerospace company where the CEO was a ham operator. He had designed and built a high quality radio. Anyone that worked in the company that was a ham got one free.

I had a friend, he moved away, that was a ham operator. The observation I make about hams is that they are extreme pack rats. They will squirrel away anything that might be useful.  My friend took the old TV antenna from the roof as it had a rotator  and he needed one for his antenna

Jerry B

[zepherusbane]

Some of you mentioned being hams. I started studying for a technical class but I have not made any useful progress, I always get distracted by my many other interests (such as stamps!).

I have a cheap handheld that I got to learn with, I don’t have a real antenna at this point, just the stub that came with it and a longer, but still pretty limited range add on for the handheld.

Have any of you done much with software defined radio? I have been going a little crazy installing dozens of packages related to ham on my new Linux install.

There is a Linux specialty distribution just for ham that I used to get a feel for what is out there in the open source world, amazing support in the Linux community exists for ham related software. Everything from actual software defined radio that talks to your antenna via a serial port, log programs, antenna design, even Morse code trainers.

Check out Andy’s HAM Linux if you want to download an iso to try it out- it’s free. It uses a different menuing system than the one I was talking about a few posts earlier in this thread. I did create ham menus for the newer style I am using.


Dale

[zepherusbane]

It's been over a year since I posted in this thread, but I had a topic that fits perfectly into it so I'm bringing it back.

So I know many of you use Stiener album pages, well I do for most of my albums, but I never know which country I'm going to work on next and so I wanted to just download all of his pdfs so I wouldn't have to go log in each time. After all, the files themselves aren't really that big, but there are a whole lot of them since he tends to make new pdf's for each year's new stamp additions for each country. I think it's up to well over 4,000 individual pdf files now.

As a world collector, I'm still foolishly optimistic enough to think I'll eventually get to all of these countries and print out pages. On the download page, there is just one giant list of everything one file at a time, there isn't a bundle that has everything available to download at once. Initially, I started downloading one file at a time using the directions he has on the page (right click on the link to the pdf file, save to local machine). I got through maybe 50 files before realizing there had to be a better way so rather that trying to download all, I started by just getting what I needed immediately and decided to think about how to get all the files in an easier way.

Well, I decided to write some code to do the downloading for me. I wrote a program using the R language that logs in to the web site and then downloads each file one by one. If this doesn't qualify as technical rambling, I don't know what does....

I cheated and rather than coding something to parse out the pdf links directly, I just manually copied the column that has the links into a csv file and read that in to the R program to know which files to download, the main the program had to do was to simulate a browser and to log in to the site with my credentials so the program could actually see those links.

If you are familiar with R, you know it's a language specialized for Data Science, one thing you often need in data science is to scrape data from web sites or to manipulate data in csv files (turning those files automatically into data frames for easy use, etc.), so the tools to do this are pretty simple in R.

I use R Studio to interactively run R code, it's a good tool for iterative development.

If anyone is technical enough to want to try this and wants some help to do it, let me know. You have to have R and R Studio (both free open source tools) installed, and there are some libraries you have to install once you have R Studio running, but since this is normal stuff for R the tools to do those things easily are built into the tools. Here's a screen shot of R Studio while I was running this.


Here is the code, but you would also need to create the csv file containing the pdf links (these links are listed on the download page). That csv file needs to be placed into the directory where your R runs.

library(Rcrawler)

br <- run_browser()

Data<-ContentScraper(Url = "http://www.stampalbums.com/worldwide_list.asp?", XpathPatterns =c("//h1","//article"))

brs<-LoginSession(Browser = br, LoginURL = 'http://www.stampalbums.com/worldwide_list.asp?',
                 LoginCredentials = c('yourusername','yourpassword'),
                 XpathLoginFields = c('/html/body/div/center/table/tbody/tr/td/form/div[1]/center/table/tbody/tr[1]/td[2]/input','/html/body/div/center/table/tbody/tr/td/form/div[1]/center/table/tbody/tr[2]/td[2]/input'),
                 XpathLoginButton = '/html/body/div/center/table/tbody/tr/td/form/div[2]/center/p/input' )

# Read in the contents of the csv file to a data frame object.
urllist<-read.csv("urllist.csv")

#the lines in the csv file should all have full links to a pdf, similar to this:
#final_url="http://www.stampalbums.com/pages/south_africa2005.pdf"
final_url=""

i=0
for(i in 1:nrow(urllist)) {
 if (i<1807) {next} #get past the ones I've already downloaded
 #read each row and convert to a string with just the text of the url in it.
 final_url = toString(urllist[i,])
 print(final_url)
 #create a variable with just the filename stripped off the end of the URL to use for saving the file.
 filename=tail(strsplit(final_url,"[/]")[[1]],1)
 print(filename)
 #download the file passing the link and destination filename.
 download.file(final_url, filename, mode = "wb")
 #break
 #if (i>3) {break}

 #sleep between each download for a bit to be sure not to overload the server.
 Sys.sleep(2)
}

In R # means it's a comment, you'll note in the code some "break" commands commented out; if you use this be sure to test on just one line first and uncomment the break and comment the download command so you can see that you are getting the expected link and filenames. I also tested with 3 lines so that's why the second break was there.

Also, I skip the first 1807 lines in this example, that's because my login timed out after downloading that many files and I wanted to start off where I left off rather than try to download them all again.

You'll also want to comment skip  - if (i<1807) {next}  - line if you want to start from the beginning of your file.  Note also, by default R expects there to be a column header in the csv and skips that automatically when reading out data (meaning you would skip the first file if you didn't have a header in your csv).

The various xpath items in the code are what make the magic possible, those tell the program where to insert the user name and password and how to click the button on the page to submit them.  The values for those xpath variables you can get by "inspecting" a web page in Google Chrome, then hovering over the elements in the developer tools that inspecting opens to find the spots where you need to enter data, then you can right click on that spot to copy the xpath info.  Screen shot below: