d5ve

I wonder how this sort of thing compares with asking claude to read a ctags file. I have git hooks set up to keep my tags up to date automatically, so that data is already lying around.

Don't the 12 Days of Christmas start on the 25th though?

I saw an interesting video on the same topic last week. By Brian Haidet (AlphaPhoenix) on reversing the GOL - it turns out that NP hard is hard! https://www.youtube.com/watch?v=g8pjrVbdafY

I had a run-in with the BOFH in real life many years ago as a university student where he was working. I'd messed up something network-related on a lab PC, and noticed that the NIC was now in promiscuous mode for some reason. I did a bit of idle poking around and pinged a few interesting hostnames I found. I was called into the the BOFH's office the next day and interrogated. He glared at me as I explained, and eventually let me off with a stern warning, though the PC was snatched from the lab before I made it back there and never seen again.

Damian Conway's https://metacpan.org/pod/PPR module defines a regex (in the commonly used sense) that can match perl source code - including perl regexes!

Edit: "The PPR module provides a single regular expression that defines a set of independent subpatterns suitable for matching entire Perl documents, as well as a wide range of individual syntactic components of Perl (i.e. statements, expressions, control blocks, variables, etc.)"

That version is missing some of the text. At least the paragraph starting "Perhaps I should explain to you here that although John Landy has". A bunch of other copies I found were also missing that paragraph.

I found an audio version on youtube https://www.youtube.com/watch?v=Y3-SrS15vb8 that had the missing paragraph (starting at 8 minutes in).

I use a Glove80 keyboard, and even after 4 months I'm not back to my normal typing speed! This has cost me a fair bit of productivity, and was stressful at times. Almost feeling like I'd suffered a bit of brain damage, where I could picture what I wanted to do, but be unable to type it out.

Having the modifier keys in the thumb cluster feels a lot less awkward than using my left-hand little finger at odd angles. And I like the fit and finish and customisation.

Overall though, I think I would have been better served by a regular qwerty, not columnar, split keyboard. Maybe my middleaged brain ain't plastic enough any more.

<small>Plus I think the benefits of columnar are oversold. If I naturally curl my fingers in and out, they kinda fan in and out, rather move in parallel straight lines. And the right hand side of a regular keyboard matches that movement pretty well. The left hand side is backwards of course.

I found a post mentioning "SuperBallistic Common Lisp" from 2015, perhaps that's where you remember it from.

SBCL goes SuperBallistic! I have hacked together an SBCL port for the IBM BlueGene/P supercomputer's Compute Node Kernel (CNK).

https://sourceforge.net/p/sbcl/mailman/sbcl-devel/thread/AAN...

    # Sort list by most common terms
    alias sorn='sort | uniq -c | sort -n'
    # Most common IP addresses:
    $ cat access_log | awk '{ print $2 }' | sorn

    # Last 30 modified files.
    function new() {
        ls -lat "$@" | head -30 ;
    }
    # I just downloaded something, where is it?
    $ new ~/Downloads

I've been using mosh since it was announced, but I periodically try Eternal Terminal.

However I havent ever managed to get ET running on Amazon Linux which all my AWS EC2 instances run. I've tried compiling, docker, app images and nothing has worked.

LAMP.

Linux, AWS, MySQL, Perl

HTML still works it turns out!

Before I used a password manager, I used three random word phrases, but with the first word lowercase, the second upper case, and the third (chosen to contain at least one of "aeio") with tr/aeio/4310/. This gave phrases that passed requirements like "must contain a mixture of upper and lower case" and "must contain at least one number". It also increased the search space for guessing. In some cases I had to chuck a "!" on the end to meet a "must contain a symbol" requirement.

"crystal lizard rekindle" became "crystal LIZARD r3k1ndl3" etc.

That looks neat, cheers dankco! I think there is a lot of room for these lightweight text-first systems.

I use a simple homebrewed system for my notes too, https://github.com/d5ve/noop which is just a search/edit loop, where you look for files and add/edit them.

https://github.com/d5ve/noop started out in golang too, but I ended up shelling out for so much stuff anyway, that I rewrote it in bash.

I figure that most humans alive today are the round-faced, more sociable, less aggressive strains. We've domesticated outselves because those more sociable and able to get along with their fellow humans have more children with better survival rates.

I've worked remotely (with a 12-hour time difference!) from a home office for 3.5 years. Initially my wife and daughter were home much of the time, but for the past 1.5 years it's been just me, as my wife is working part-time and my daughter is in childcare.

I wasn't a particularly social person originally, but working alone has made me even more insular, and less inclined to join in social situations. It ended up with me feeling lonely and unhappy. I was considering quitting my contract, although similar work is difficult to find where I'm now living.

So two weeks ago, I started working in a shared workspace, and I will try that for a few months to see if it helps.

My guess is that every person has a different buffer of how long they can do without regular human company during the day, and I'd exausted mine. Having a family helps, but may not replace professional company where you are discussing things about your dayjob.

I used the Ian Knot for a while, but then went back to one of the standard methods. The difficulty I had with the Ian Knot was when I wanted the laces to be tight. With the standard methods I was able to pull the laces tight, then keep the Starting Knot tight also as I tied the final loops. With the Ian Knot, it was hard to keep the Starting Knot tight whilst I tied the loops.

I'm working for a perl-based UK startup. It's the second perl startup I've worked at. Both have been very successful. The current one is about ten years old, and just starting to explore other languages now. Currently the whole infrastructure is perl-based, including public services, data-processing and inhouse tooling.

I think of online advertising in the same way as I now do smoking in a workplace, pub or restaurant - I can't believe people used to live like this. Using someone else's browser and seeing the distracting and intrusive ads feels a bit like trying to eat a meal surrounded by the smell and eye irritation of second-hand smoke.

I defined the following function to convert my bash aliases to fish functions, which eased my switchover a bit. I saved it as ~/.config/fish/functions/import_bash_aliases.fish

  function import_bash_aliases --description 'bash aliases to .fish function files.'
      for a in (cat ~/.bashrc  | grep "^alias")
          set aname (echo $a | grep -Eoe "[a-z0-9]+=" | sed 's/=//')
          set command (echo $a | sed 's/^alias .*=//' \
            | sed 's/^ *\'//' | sed 's/\' *$//' )
          echo "Processing alias $aname as $command"
          if test -f ~/.config/fish/functions/$aname.fish
              echo Function $aname is already defined. Skipping...
          else
              alias $aname $command
              funcsave $aname
          end
      end
  end

I read somewhere recently about companies buying up this MAC address data from individual shops and using it as an additional profiling source. If you walk into a store and end up purchasing something, then your MAC would be linked to any payment details. At which point, the profiling company would know who you are, and where you go.

I started writing a small proof-of-concept android app to randomise the WIFI MAC to make it more difficult to track a phone and its owner. The android API doesn't support changing the MAC, so it requires a rooted phone.

Mostly working code can be found at: https://github.com/d5ve/RandoMAC

It's made tricky by being unable to change the MAC whilst the WIFI is enabled, and re-enabling the WIFI resets the MAC back to the hardware one. Also by my total inexperience with android apps and java.

One thing I've found useful on linux is the following:

    $ tail -qFn0 /log/dir/*

You will only ever see lines from after you run the command.

promobay.org resolves to 108.59.2.74 for me, and www.promobay.org resolves to 108.59.2.75.

The first is blocked for me on Virgin, but the second works fine. This does look like a DNS issue on the part of TPB, rather than Virgin blocking promobay.

Even shorter, as -l appends \n to all prints:

    $ perl -MLWP::UserAgent -le 'print $LWP::UserAgent::VERSION'

One thing I've noticed with Agile projects at $work is that the processes better suit a "contractor" mindset, than a "fulltime employee" one. NOTE: I'm a fulltime employee, so view things from that mindset.

As a contractor (in theory), you're bought in to do a specific job, and then you leave - get done, get paid, get gone. This works well with Agile, where you can devote your entire attention to the project, and not get caught up in other things.

However, fulltime developers will find it more difficult to devote all their time to a project, as they're likely to still be maintaining some of their previous projects as well. In smaller organisations, you often end up being the go-to person for part of the system, and the rest of the business may struggle with these people being out of reach for weeks or months.

Another factor is that contractors have less reason to care about existing code and its design principles when designing and implementing a new project, and well as less reason to care about the maintainability of it. This leads to the software platform having a melange of different styles, further complicating the lives of those who have to support it later on.

To me, Agile just seems like contractor culture converted to religion.