View on GitHub

Quorten Blog 1

First blog for all Quorten's blog-like writings

When I was a student at the University of Minnesota, out of curiosity, I wanted to know how they managed their networked home directories in CSELabs. /etc/nsswitch.conf is where all the magic goes where they tie down to LDAP, right? Sure, so I looked inside there… and to my surprise, I did not find LDAP. Instead, I found vas4. What is this vas4? As it turns out, it is a more specialized authentication system specifically designed for large institutions with many customers, made by Quest Software. The modern name of the software is One Identity, but the old code name still sticks in low-level system configuration. At one point I erroneously thought that it was developed by Qwest, one of the spun-off “baby-bells” companies when the Bell Telephone Company was broken up from its old monopoly structure. Yeah, like for telephone customer account management? Nope, it’s not like that.

Unlike LDAP, vas4 never became popular in mainstream use.

20170911/DuckDuckGo vas4
20170911/http://rc.quest.com/topics/mod_auth_vas4/howto.php
20170911/https://en.wikipedia.org/wiki/Quest_Software
20170911/https://www.quest.com/

Read on →

Tricks on using Chef Inspec

2019-09-11

Categories: ruby  
Tags: ruby  

Useful Chef inspec hacks!

result = command(...).result
# Do this to make sure the command actually gets run.
x = result.exit_code

fconts = file(...).contents

You can use these for remote system management and receiving files. Why? Because… business reasons.

You don’t have to use all of Chef Inspec’s (RSpec’s) full DSL that they give in the examples, after all. The commands on their own, despite freeling undocumented and hacky, work just fine.

Do you want to print information to standard output when running Inspec? I don’t know of a sure way of doing this, but one way that worked well for me is this:

describe my.stderr do
  its('content') should be ''
end

Then if your standard error (for example) is not zero, then you will get the standard error messages printed out when running Inspec.

20190910/DuckDuckGo inspec expect
20190910/DuckDuckGo chef inspect expect
20190910/https://www.inspec.io/docs/
20190910/https://www.inspec.io/docs/reference/profiles
20190910/DuckDuckGo rspec expect
20190910/DuckDuckGo rspec reference
20190910/http://rspec.info/documentation/
20190910/http://rspec.info/documentation/3.8/rspec-expectations/
20190910/http://rspec.info/documentation/3.8/rspec-expectations/#Truthiness
20190910/DuckDuckGo chef inspec print information
20190910/DuckDuckGo chef inspec info message
20190910/https://www.inspec.io/
20190910/https://www.inspec.io/docs/
20190910/https://www.inspec.io/docs/reference/dsl_inspec/

Being “a world class software platform” (Linux, WebKit, Microsoft Office, and so on) appears to come at a considerable price in terms of code quality and bloat.

The OpenBSD community takes it upon themself that by being relatively obscure, they can put a greater emphasis on quality than is the standard for “world class software platforms.”

In essence, what I am doing is gradually putting together a miniaturized “BSD-like” operating system, again in hopes that doing so will allow me to have a greater focus on quality than is possible in the bigger, more mainstream projects.

Now, all that being said, another thing that is interested to think about in relation is the IKEA effect: the perception that things that you need to do some of your own effort are of higher quality than things bought readymade.

20190908/https://en.wikipedia.org/wiki/IKEA_effect

But alas, this is where the specifics of the facts come into play. In software development, if you are using “ready-mades,” that means you are basically restricted to using very large and complex software libraries, trying to tout in themselves that they are a software platform. The problem is that this kind of software simply just doesn’t fit on a traditional PC microcomputer. Even the most trivial of programs will be terribly bloated and exceed the memory capacities of your system!

Read on →

For my DSLR camera remote, I was wondering if I couldn’t get it to work because the battery got drained out because my push button repair may have turned out to be a bit too jittery. Good thing I got the one with the replaceable battery.

The first thing I did was measure the battery voltage with a voltmeter when it wasn’t working. 2.9 volts, yeah that’s lower than the earlier times when I measured at 3 volts exactly, but that shouldn’t be too bad, should it? So long as it uses a switch mode power supply. Do these tiny low current electronics use switch mode power supplies, or do they wire up to the battery directly? I searched around for this on the Internet, and after weeding through some StackExchange questions where people clearly didn’t know what they were doing, yeah, it looks like it is plausible that the coin cell electronics may wire up to the batteries directly. Also, the coin cell batteries themselves tend to have a high internal resistance, so it is important that you do not try to draw too much current from them, and also that you avoid continuous current draw. Suffice it to say, it’s safe to assume that coin cell battery electronics are more sensitive to battery voltage changes than AA battery cell electronics. Not to mention that the coin cell batteries themselves have a reputation for outputting a more stable voltage than AA battery cells.

20190907/DuckDuckGo does cmos clock battery use boost circuit
20190907/DuckDuckGo 3 volt button cell battery microcontroller boost power supply
20190907/DuckDuckGo do coin cell electronics use voltage regulators
20190907/https://electronics.stackexchange.com/questions/261124/coin-cell-regulation-required

COBOL is still a widely used and modern programming language? Yep. Wow, that sure is interesting. Despite all the badmouthing about COBOL, there is a lot of software written in COBOL and it can be deployed in modern environments and under modern paradigms.

20190907/https://www.eweek.com/development/six-reasons-cobol-has-survived-to-age-60

Wow, when searching around for the article from the Programming is Terrible website, I found this other interesting article about problems with the TCP protocol and possible improvements upon it.

20190907/https://programmingisterrible.com/post/69710734780/the-trouble-with-tcp-its-good-but-were-stuck

Then I found this other article about starting another community in light of harassment in a previous one.

20190907/https://programmingisterrible.com/post/62561520295/computer-anonymous

When searching for other subjects, I came across this page by accident, and it sure is interesting. gHacks, a once strong and mighty website, got hit really hard by some changes that Google made. Pretty soon, they were slipping from the page rankings, they were getting less traffic, ads were bringing less revenue, and it all over became difficult to keep things running. No longer could they have paid writers, or definitely not nearly as many of them. All of this, hinging around the random popularity trends that the Google search engine brought. Except for the fact that the Google search engine itself remained popular nonetheless, and some of the big and monopolous sites became ever more popular.

20190907/https://www.ghacks.net/2015/02/27/ghacks-is-dying-and-needs-your-help/

Nowadays, some Zoom video host websites are IPv6-only, and this causes problems for old dinosaur networks within enterprise companies. Seriously, I’d assume that they made such a decision out of pure economics. With IPv4 address exhaustion, an IPv4 address must be pretty expensive nowadays, is it? After some rough and tough searching, indeed, it purportedly get more expensive, and some companies are dropping their IPv4 addresses to save money. But, this problem notwithstanding, this is not speeding up the transition from IPv4 to IPv6.

20190906/Google which is cheaper public ipv4 or public ipv6
20190906/https://www.networkworld.com/article/3254575/what-is-ipv6-and-why-aren-t-we-there-yet.html