hpr-knowledge-base/hpr_transcripts/hpr2852.txt

Episode: 2852
Title: HPR2852: Gnu Awk - Part 16
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr2852/hpr2852.mp3
Transcribed: 2025-10-24 12:13:19

---

This is HPR Episode 2008-152 entitled Genoaq Part 16 and is part of the series Learning Ork
It is hosted by Dave Morris and is about 43 minutes long and Karim an exquisite flag
The summer is winding up the Genoaq series
This episode of HPR is brought to you by archive.org
Support universal access to all knowledge by heading over to archive.org
forward slash donate
Hello everybody and welcome to Hacker Public Radio. This is the 16th and final episode of
the Learning Ork series and with me tonight I've got Robert, also known as Be Easy. Hi, yeah.
Hey, how's it going, Dave? Great to be talking to you in person for once.
Yeah, yeah. This is so weird isn't it? We've been doing this now for a fair number of years.
I made a note when we started 2016 so that's a good three years so that we've been doing this.
We've never spoken to each other just emailed and whatever. It's really nice doing it this way.
It is and actually now that I have mumbles set up on this on my desktop here,
maybe I'll have an opportunity to join the community meeting every once in a while.
Oh, that would be lovely. Yes, yes, the more the merrier really and get different opinions on stuff
rather than just can and myself going on and on and boring ourselves or what I'm boring the audience or whatever.
Well, I personally really like the consistency of having you guys do it but every time there is a
special guest that comes on it is it does break it up and it makes it more interesting so. Yeah,
yeah, it's nice having just different outlooks and different voices on isn't it? Yeah, I agree.
So I thought we ought to start by talking about why? Why are we finishing this? Why are we winding up
this series? We've we've not reached the end of the walk but I think personally that it we've
probably reached the sort of practical limit on this. What's your feeling? Yeah, well, I think for
for my personal use, we've probably reached my practical and my knowledge of it maybe two or three
episodes of mine ago and maybe two or three episodes that you've done to just because we started
going into some of the edge cases and some of the more the non-standard functions, if you will.
And I think it's great to learn that stuff and for me, if I get to a situation where I need to go
into one of those rapid holes, I always have the documentation but it is good to have a recorded
message now to to to be able to go back to. Yeah, yeah, I think so. My thinking was that most people
use ork, maybe in the same way as they use said, just as a thing that takes a piece of text
coming in from one direction, does some stuff to it, filters it in some way and passes it onto
the next step. And as we're just on the edge of getting into talking about
ork has a full-blown language, which it is, it is though I would I would maintain that it's it's
shows its age quite a lot and it's maybe not the best thing to be using for for a lot of functions
but it's brilliant, absolutely brilliant for for doing the stuff that we've been talking about
pretty much up to now. I agree and when you think about when ox started and the type of power it
could give before some of these other languages have come around, it really was a really unique tool
in the toolbox for a really long time. But since I've been using computers really heavily,
it was great to learn, but yeah, it is I think the way you said it was perfect when you're talking
about how, you know, if it's in the middle of another command that you're piping around,
it is great to use ork. And actually I've been in a little mode right now where I've been trying to
simplify and simplify and simplify and sometimes I've been taking out ork and putting in cut
when I don't need ork just because, you know. Yeah, yeah, I know it'll be. You can do things like
just feed it into ork and then output field one. Whereas you didn't really need to use ork
ork is not particularly heavy, but it's heavier than say cut where cut would enable you to just
chop out that one field, you know. So yeah, I know what you mean. And then the same thing with said,
I think and we're going to go into this in a little bit, but some of my favorite projects were
a combination of set and knock together for cleaning up files. Yeah, yeah. So should we go on
to that? I think we've probably made the point that we feel that we've taken it as far as we really
should. And as you said, if you need more, the manual is actually very, very good and there's lots
and lots of pointers to it and you can find your way through it quite easily. So I would suggest
going there. What we wanted to do was talk about how we got into using ork our personal experiences.
Do you want to start that off? Yeah, I can. Well, I think it all really starts with my experience
getting into getting into Linux, which I guess I started on Ubuntu 1104 a little while ago. I guess
that's 2011. When I was given a computer that had Ubuntu on it and it was the only computer I had
at the time. And so I had to learn it. But then it turned into learning some of the things that you
learn on the internet and when it came up to work sticking with Linux. And then I found myself
into some really interesting projects where I had to do data cleaning. And some of my favorite things
that I had to do were around cleaning up dirty like CSV files, other types of text files were set
in knock were really handy. And one thing that you could rely on being on a computer would be
all can said. And so that I was able to do that without having to bring on a whole bunch of other
programming languages, especially on when I'm on some projects where I'm using computers that
are not mine. And so it's been really great for things like that. I was just looking at one of my
favorite ones. I just did a little bit of history search and I was working for a big pharmaceutical
company on this on this project. And I was looking for all of the words in a certain CSV file
that were uppercase and turning them into lowercase. And you know, I just found I just found one of my
scripts where it was just several lines of of that like cleaning up cleaning it up and say cleaning
up and all cleaning up and say cleaning up and all just and then go back and back and back and back
until we got the the finalized versions of the files because you know sometimes you can't control
the data input that you get. All you can do is deal with it when you when you have it. Oh yes,
yes. So yeah, that's my that's my probably my biggest experience with it. But I've been using it on
using AUK on and off. Well, I guess almost every day at a certain level just you know, just to do
things here and there since 2011. But I've really been moving to other programming languages to do
more of it more recently and we'll talk about that later. Yeah, that's your experience. Yeah, okay.
Well, I go way back. I was working at a local university here just to just a mile or so down the
road from about 1981 through till I retired 10 years ago. And there was one point where we had
a Vax cluster running Vax VMS operating system and that was a pretty good. They used to call it a
mainframe though. Some people would call it a mini, but you know in today's terms it was a
a number crunching centralized machine that certain people got access to. But we had very little
in the way of text processing features within it. So you had to write a program to do things. There
weren't many tools, but there was a move because it was very popular in the university scene across
the whole of the world really, particularly in the US and in the UK. And so people were actually
moving Unix tools onto it. And I saw that this was happening and found that GNU AWK was available
quite a different from what we have now, but in the 80s and also said it was available. And so
I installed them on this machine and was able to use them for some of the sort of applications
that I had a particular thing that I did. And I really got to be careful. I don't go on and on
about this because it was a thing I did for many, many years and I could talk to you to death of it.
Was we didn't have a way of creating accounts for students as they came into the university.
So new students would come along and they'd, oh I need to use the computer for my course and
requires me to do for trying or something. So they'd have to walk to our computer center and say
can I have an account please? And some guy would sit at a terminal and type in all the details
and stuff. And it got the number of people who needed this increased as the years went on.
And it became an enormously laborious process. So I took on the task of automating it by getting
the data from the student registration computer and using it as the basis of code which would
generate accounts for them. But the data quality, as you were saying before, was pretty bad.
And I don't know how they managed to get it quite so bad, but you know you'd get people coming in
and there would be student records of becoming in that said they'd just been born this year.
Which is fairly unlikely for university students. Somebody'd forgotten which year to type in.
They've maybe got the date of birth right, but they've forgotten which year.
And so spotting things like this, flagging them back and fixing them was a major thing.
And all can said we're just fantastic for that. And dealt with that really, really well.
And then the data was passed on to other programs that we developed for doing the
nitty gritty of account creation. So yeah, and then a bit later on, I won't go on to
a great length here, because I'm talking about 28 years of doing this sort of stuff.
Wow. I bet I'm not going too far. But in the late 80s and early 90s, we started moving towards
Unix. And we had a whole range of Unix versions. This was really before Linux was coming on to
the scene. And Auckland was there, of course. And was getting very much used. We were learning how
to do pipelines. And we weren't using bash at that time. I don't think it existed in the earlier days
that we were using the Born Shell and the C shell and the TCSH shell and so on. But you still
use your org and said there. So certainly did a lot of that for all manner of uses.
And later on, we moved to Linux quite extensively with a lot of servers running
initially with Fedora and later Red Hat. And of course, Quark was very much the thing to use.
And I ended up in some cases teaching people a bit about how to use them.
Particularly regular expressions. People didn't seem to grasp the concept of regular expressions,
which very much are being in an awkward state. And so that's something that actually
got me into it as well as the regular expressions. I do have a background. I know you do a lot of
prod. I have a little bit of pearl experience. And that's what got me really into regular
expressions before I got into Linux. And so it was a little bit easier for me to be able to go into
some of that stuff. But yeah, for a lot of people, it really is just a foreign language.
It's hard to wrap around your head around. Oh, yeah. It's really hard to read and understand.
When we moved to Unix systems, there was a point at which I had an old trick's workstation on my desk.
And we ran our mail system. We had a mail system which was private to just us in our department.
And I used to get a lot of email because I had configured various of the service to send me
reports in emails. I would get thousands of messages a day. And so I got into using a thing
called prop mail, which is a very, very clever tool for filtering your mail. You know, checking
its subject, its form and to and or and the contents of the body of the mail and stuff,
all using regular expressions. So I got, I became quite skilled comparatively anyway in regular
expression use by doing that. And of course, that's that moved on to the other tools that used
regular expressions big time. So yeah, so I certainly feel pretty comfortable. In fact, I find
Ork and said a little bit limited in terms of what they can do compared to say pearl.
Pearls gone way ahead of some of those things, isn't it? Yeah, sometimes, I mean, I pretty much
stay away from Pearl for the most part now. But sometimes I do have a time where I do just pick
Pearl just because it is, especially for Pro 5, I haven't really done too much in Pro 6,
but getting in Pro 5 and getting really deep into the regular expressions has been something
I've done for over a long time ever since my bioinformatics master's program.
Yeah, because Pearl used to be very popular in that sort of area, didn't it? It was used in a lot
of genetic analysis and that type of thing. DNA sequences and stuff was processed with
pearl in the earlier days. Yeah, and so if you're ever dealing in some old code base of a bioinformatics
pipeline, you'll see Pearls still. A lot of times people are moving to R or to Python now
in the bioinformatics world, but you still see a lot of pearl out there. Yeah, yeah. I certainly see
Pearl a third, I still write in Pearl quite a lot because I just feel more comfortable in it
having done it for so long, but trying to move on to Python and stuff like that. But yeah,
it's still about, it's still about, it's still quite common actually when you look behind the scenes,
but it's not, it's not a thing that anybody would volunteer to learn and go and write. These days
it's more legacy stuff, I think. Yeah, I agree with that. So yeah, so yeah, that's pretty,
I mean, I could go into a lot more detail about how I got to war, but I'll shut up now, I think.
But was there anything else you wanted to add to your experiences?
Not really. I mean, I've kind of gone on about it in some of other episodes where I talked about,
you know, coming out of a university and going into molecular diagnostics and a couple different
big companies in here in the US and then moving over and getting into, oh, there's my son. Hey,
what? Hold on. Yeah, getting into some of the, some of the other topics like biome format is later
on. And that's when I started getting really into more computer programming. Before I was more of
a computer consumer, maybe a super user of certain applications that we used, but not really,
not really someone who's going to make changes or do customizations. But since, you know,
getting introduced into Pearl and R and then later C really got me going on that side. Then
moving over and I started consulting and then really getting into a lot of computer programming.
And oh, my goodness, my son is over here coughing in my ear. He does sound very well.
Yeah, he's home sick today. So you're going to have to bear with him a little bit. And he
hears me talking now and he's all excited to hang out with me. But yeah, it's been it's been a really
long road to where I am now in my career. But it's been, you know, it's still moving more,
more and more away from some of the bench lab work and moving more and more to automation
and data analysis and now they call it data science. When I first started, it was just called data
analysis or bioinformatics. But yeah, that's that's kind of where my career has taken me and
and I pretty much stay in those areas nowadays. So I do some web development as well. And some
application enterprise application customization as well. But I try to stay away from those as
much as I can. So are you a biologist by by training them? Yeah, I'm a molecular biologist by
training. So, you know, DNA RNA, that's stuff by training. And then also bioinformatics, which
is a combination of that with the, you know, automation and computer systems. Yeah, yeah. I have
a I have a degree in biology, but not the sort of biology that you've done because mine was in the
late 1960s, early 70s. So, so it was like sort of traditional biology taxonomy and stuff like that.
So, so yeah, but yeah, it's interesting. I've certainly been to a few talks on bioinformatics.
I went to some at Phosdem, the big conference in Belgium. And they are really, really fascinating.
So yeah, it's the I'd like to understand more about that sort of stuff.
And actually, that's one area that's it's been really kind of lucky providence for me,
because a lot of bioinformatics is done on Unix and Linux. And so learning all that while still
on Windows was kind of difficult. And when I moved over to Linux, I'm like, oh, you know, you
mean like pearls are already installed? Like, okay. Yes, it's great. My daughter recently got her
degree last year. She got her degree in biology. And she used a quite a lot, I think, you know,
to do statistics more. So it's for the computation stuff. Yeah, most of my bioinformatics has been
mostly in our in my program too. Yeah. So yeah, I've only doubled with it in helping her out.
It's got a lot of power. I know. Yeah, I think that's actually a pretty good segue to the next
topic that we wanted to talk about, which is what are some of the tools that we use now instead of
using ARC? Yeah, yeah. We were going to talk about what bits of what we've left out, but we'll
come back to that one in a minute, shall we? The, um, I just mentioned that, um, well, I really
said it. As far as I'm concerned, I moved from, from, uh, Gork on the back cluster that I was managing
to Pearl, which is version four, and then later to version five of Pearl. And this, this was not
a difficult move because Pearl is very much derivative of, uh, the tools like Ork and Bash and
um, said and, and all sorts of other things. I think, uh, the, um, the guy who wrote it, whose name
suddenly escaped me, uh, who designed it, had been using these tools a lot and, and sort of came
up with ideas that effectively glued them together into, into Pearl. Yeah, um, I, I know the name
two that I'm not remembering either. It just come back. Larry War. Yeah, that's it. He's a really
interesting guy. Not a originally computer scientist, so this is why Pearl is, is, is unique,
because it's come from somebody's, he's come from an entirely different place.
Well, well, that's often been the criticism of, uh, Pearls, is that it is our right only language.
Well, yeah, but then any language is, if you, if you wish, I mean, if you want to write something
that's incredibly obscure, you can do it in any language, uh, it's my, my feeling. Maybe not
but I don't know. Haskell is pretty hard to understand anyway.
Yeah, I've, I've taken a couple of times. I've really been enjoying the series on Haskell
because it is, it has been quite a mystery to me for a long time. Yeah, yeah. I have a friend who's
a computer scientist and whenever I mention Haskell, he said, oh, you should learn it. You should
definitely learn it because it's the, it's a wonderful language. But he's coming at it from a PhD
in computer science, you know, and I think he maybe sees the world somewhat differently from
the way I do. You know, I, I think that is it, because everyone that I know who really likes Haskell
comes from a, uh, either a, um, either from pure computer science or some type of other, um,
you know, philosophical stance on it. It's mathematical prompts or something like that,
which, which is great. And I'm sure it has a lot of great use cases, but for the work that I've
done, well, and also just some of those stuff that I do, there's already great ecosystems around
some of the other languages, which is really why I choose them. Yeah. So what would your, what would
you be recommending as an alternative to walk? Well, I have a two-part series on XSV if you're
dealing with, uh, delimited files. And that's been my go-to for, um, recently. It has a lot of power,
especially I'm sitting here right now on my, um, my desktop computer, uh, which has, uh, 12 cores.
And so the ability to, like as a 12 core, I say, having the ability to go multi-core and, and go
through a file that has, I just did one earlier this morning that has a five, over five million
rows in like three seconds. And, and get a summary, you know, the thing that we did where we said
for, you know, to get a group by count, it does it in like five seconds on five million rows.
So it's written, it's written originally in Russ. And it has a really, um, it has a really nice
syntax for the most part. It's not a full-blown language. It really is just a command line tool.
But I use that for, for just command line interactive, uh, looking at, at, uh, delimited files.
And then if it's more involved in that, then I get, I either pull out R and I've been really
into the tidy first lately, which if anyone knows ours, it's kind of a newer development in R,
not, not base R. There's, there's a community around, I think started by a Hadley Wickham,
a community around making a more fluent interface to dealing with, uh, to dealing with R.
And, and so, um, on the R side, I've been doing that. And then also using, um, Python and the pandas
package, um, both NumPy and pandas. And I've, I've, I, I used to be really good with R and I would
write things that I knew had to be in Python. I would write it in R first because I was just more
comfortable there. And I, in the last two years, I really made it an effort to really learn pandas
really well. And it's, it's been great. The things you can do with it. And now that there's another
program called DASC, which is another library where you can, one of the criticisms of Python is that,
is that a lot of times you're, you're working with the global interpreter lock and things end up
being single threaded. But with DASC, it's really easy to turn those same with the same pandas
functions that you're already running. But we're doing a read all this data, group it by and give
me the mean. It's able to do those things on, um, multicolors with the same syntax that you're
already used to using pandas. So, um, okay. So that's, that's using, uh, multiple processes
around the multiple threads. Well, DASC can do it either way. So, um, it depends, you can set it
up to say, well, for this task, if, if you leave it up to it, it will choose which one it thinks is
best by looking at the data. But you can say at the beginning, you can declare, I want to use threads
and then use threads or I want to use cores and use cores. Um, and, and so I've been messing with
those things a lot and, you know, it's been, it's, it's what I primarily do nowadays. I just gave a
talk a couple of months ago and I'll be doing that talk again a couple of months from now and I do
live coding with a Python where I'm turning, uh, I'm turning problems like business problems and a
data of business problems and turning it into actionable events, event driven process, uh,
process development and decision making. And so being able to, it's one of the things I think, um,
um, I think you talked about, um, the creator of Pearl, he came from an environment where he really
wanted to have like feedback and be able to see the results back and, um, and, and the same thing is
true on the Python side where there's an interpreter where you can just pull it up and it's kind of
its own shell and you can put in the data, you can see what the results are and you can go back
and forth with it and, and, and make changes to your code really fast. And so that's where I've
been, um, hanging out mostly is, is really in Python, but sometimes also in R when I, um,
when projects require it. That's very cool. Yeah, yeah. I, uh, I, as I said, I did most of my data
manipulation in Pearl finding it to be an amazing way, amazingly better than anything else
it was available at the time, but, but yeah, I can see that I should really be, be rethinking
that sort of stuff when I need to, need to do that type of thing in the, in the future. So yeah,
I've been looking at Python a little bit myself. And, um, in fact, I was doing a tutorial on,
on Rust the other day just to see what sort of a language it is and, and that in itself is quite good,
but it's not specifically a data manipulation language, but it's a, it's a pretty amazing
language show if you've ever dabbled with stuff like C, C will catch you out very, very easily,
but like Rust is, is protecting you a lot. And, uh, that makes it good in my, uh, it's far as
unconcerned. Yeah, I've, I've seen a lot of, uh, really interesting things going on both in the
Go and Rust communities. And I think those two languages are going to have more and more of
influence, uh, on the overall landscape of programming in the, in the near future. But, uh, you
know what? Python is not going anywhere. I don't know if you pay attention at all to like the,
either the stack exchange or the jet brains, um, surveys that they do. And Python is one of the
most popular languages in the world now, um, you know, right up there with C and Java. So,
and it's because it's being used, it's, it's a general purpose language that, um, although it,
you can, you can write things in C and have them, and have them, um, be wrapped in a Python wrapper,
which means that you can bring some of the speed that you, that you lose from having a, an
interpreted language. Like, like, Frances pandas and, and NumPy, they are written in C. And some
of, and, and SciPy is written in C and Fortran. So a lot of the power that you get from those
compiled languages you can still have access to and Python. And also, there's been like a lot of
community effort around it. Uh, it's, it's the reason why I try to standardize around it is because
it's a one of the languages that you can, that there's a huge community around both the data science
and the web development. The community. So I can write a web application and write, uh, data, uh,
analysis for that, um, for that, uh, web interface in the same language, which makes it, you know,
really easy to have a team of people all working on the same code base. Yeah. Yeah. I do see this
happening and, uh, feel like I should, I should be following, following your lead. That's considered,
I just, um, um, reluctant to start learning anything else in some respects. Yeah. So, well, you
know, without, without a reason to, it's kind of hard to just pick up something just for fun. Yeah.
Yeah. Yeah. It's, we're redesigning some of, uh, EHPR and stuff, um, going to have a different
database, which I'm putting a prototype together for now. And, uh, we're going to move to a static
website, um, eventually. And there's quite a lot of tools involved in keeping, uh, the HPR stuff
going, many of which I've written and, uh, cross a rule in Pearl. So, uh, when I, when I, uh,
resign, when I bow out, I get my, my pension from HPR, I'll, uh, I don't, I don't envy the
poor soul has to take it over because if you're not into Pearl, then Pearl can be a little bit
scary. I don't know. Try to write it so it's not, but there you go. Yeah. I'll just take it out.
Maybe, maybe we should get on that sooner rather than later. Yeah. Yeah. Yeah. It's something I'd
think about at the moment. In fact, um, how would you, uh, how, given a, given a, a concept
which you would implement in Pearl, how would you better implement it in something else? I was looking
at, um, um, what's the, what's the other language that's equivalent to Pearl I'm having,
having brain parts at the moment? It's one Python. Uh, not Python, but the other one, the,
small, small talk. No, I'm thinking of the one that was effectively derived from, uh, from Pearl.
Takes quite a lot of its ideas from it. Uh, you got me on that one. Just, uh, it'll come back to me
just when it's inconvenient to, to say it, but, but yeah, there, there is, there are other languages
which, uh, is right there, but I think, I think the point you're making is that Python's really, uh,
where at the, at the front in, in terms of a lot of this type of stuff. So makes,
makes sense to, to be moving in that direction. Well, if you don't, if you already know Python,
and you want to learn, and you want it to do something with, uh, data manipulation, you,
you don't have to go anywhere else. I think that's definitely clear. One of the things about Pearl,
though, is that, uh, and this is less of an argument now because a lot of the, the modules and
libraries that have been available for Pearl, and there's vast numbers up, and they're very easily
accessible. Um, a lot of them are not being maintained as much as they were. So, but it was the
argument at one point, if you needed to do a particular thing, there was always a module that
would do it, you know, um, for example, I did quite a lot of manipulation of, uh, MP3 in
organ, and audio files are very, very, so it's putting tags on them. Uh, you know, the ID3 tags in
MP3 and that sort of stuff. And it's not easy to do in Python. I tried to write a Python, uh,
version to do this using a thing called taglib, but the interface to taglib from Python is quite
basic. Whereas the one from, uh, Pearl is, is a lot more sophisticated. So, um, you know, but, um,
you know, there was, there was another library that I was using recently for doing that same thing,
and, and you know what, I'll, I'll look for it and send you to you an email.
All right. That would be good. That would be good. Yeah. Yeah. Taglib is quite popular, though. It's,
it's embedded in a lot of tools that you'll, you'll find out there. But, um, it's, uh, yeah.
Yeah. The one I'm thinking of, I think it has ID3 in the name somewhere. I gotta go. Right.
Right. But this, um, taglib will handle any audio format. Everything from Opus through to
Flak through to MP3 and even speaks. Um, so that, that, it's written in C++. So I do want to be
writing C++, but I'm not going to do that. Not in my time of life. Yeah. That's understandable.
Um, shall we, shall we just talk about what in org we've left out? Or did you have more to say on
them? Yeah, let's do that. Yeah. Okay. So I'd put together a list of things that, that sprang to
mind when there, there's sort of pre, prewritten notes for this episode, which I jotted down,
which will probably get changed a bit before we release them. But, um, we reached the point where
we could have been talking about how you write your own functions in the org. But the way it's done
is so weird and clunky. Even the author says, uh, says something. Not very, not very nice about his
own implementation. So, uh, I think we, we agreed both of us that we wouldn't go down that
particular road. Yeah. And I've, and I've, like I said, whenever I had to go down that way,
I would choose a different language to do it. So I think we've made the right choice. And,
you know, balancing what we think would be useful to people versus, um, what we had time to do.
I think we've made a good decision there. Yeah. Yeah. It's a shame because we begin to write
functions in a, in a language is, is a cool thing to be able to do. But they are not nice to,
to use in an org by my reckoning. Anyway, all the authors, for that matter, that's Brian Conegan,
who, who's saying that? So, oh, oh, just so, just so it's on the record. The, the file, uh, I just
found it while we're, while we're talking mutagen is the name of the, is the name of the program,
mutagen. Okay. It's the sound of me writing it down on this paper. Okay. Okay. Oh, and the language,
I was trying to remember, and isn't memory weird? Uh, is Ruby? Have you looked at Ruby? Oh, yeah,
yeah. Ruby's, Ruby's great. Um, I'd have done some things nowadays. A lot of the Ruby's,
like, Ruby on Rails. But I do see a lot of, uh, programs, especially on the Linux side where
there's a lot of, uh, you, you look at, uh, you go down, you know, app install something,
and then if you do a witch on where it came from, it's actually a Ruby library. So there is,
yeah, it's pretty, it's still pretty popular. Yeah. Yeah. There's quite a lot of power in that,
and I quite like it's Polish nature as well. And it also uses, uh, curly brackets, which,
which is cool, not indentation. Yeah. Uh, we don't have to, we don't have to get into that argument,
right? I'm avoiding it. I'm avoiding it. I'm not going to, I'm not going to go that, just,
just look that way, but I'm not going to go that way. Yeah, that's right.
The other thing I'd written down here was multi-dimensional arrays. If you want to get into
some quite complex data manipulation, it's quite nice to be able to build matrices,
or three or four dimensional arrays, or really complex structures, arrays of hashes,
or hashes of arrays and that type of thing, which you can do easily in Perl. I think you can do
something similar in Python as well, can't you? Yeah, that's what, that's what NumPy is all about,
that libraries is for multiple dimensional arrays. And also, there's also, if you really want to
get really big ones, that's what things like tensor flow for. Yeah, yeah. I tend to do that a lot,
a quite like concept of being able to have a, have build structures where you start off with a,
with an array, and then the elements of the array are arrays, or hashes, or whatever,
or procedures, or whatever else you want to do, which is something you can do in C, and to
some extent, other languages, Pascal, was one I used to use a log, you can do that type of thing.
But, yeah. Is that even something I knew you could do in Perl? I mean, in, in, in,
arc? I don't, you could do multi-dimensional arrays, but you can't do anything very sophisticated
with them. Okay. But, GNU Walk has got a lot better than it was in the early days when I was using
it as far as I was concerned. So, my notes say other languages can do this better. So, I think I
would keep away from it personally. As you said, choose another language. If you hear that. You can
get into internationalization, which is all about, you know, writing stuff, which will, is, is
sensitive to what language you're currently using, what environment you're in. But, that's,
it does exist. Somebody has put a lot of effort into adding that to GNU Walk. So, when you say
internationalization, is that mean, for instance, when you're doing a print, a print F using
commas instead of periods in vice versa? That type of thing. But also, I think you can do things
like change to different character sets or different languages. So, you can, you can build something
which has got multiple translations of a given message and switch between them depending on what
your environmental settings are as regards language. Oh, wow. I think so. But, like you can in
many languages will let you do that. I think that's available in, in a walk. But, I just
looked at it and ran away. So, I might have misinterpreted that. Yeah. And I think once again,
it's one of those cases where there's, there's plenty of other languages that you could use to do
that too. Yeah. There is a debugger in the walk. And it looks pretty good. I've not tried it
myself. But, it wasn't something that we wanted to get into. So, like I said, to go and look at it
and thought, wow, again, somebody's put a lot of effort into writing this. But, who's using it?
Who's going to use it? I feel that it's, it was maybe a little bit misguided. I don't want to
criticize all, because it's, it's a wonderful thing. But, but in these areas, it's maybe, maybe
overkill. Well, I think the idea is to turn it into a full, a full-blown language. And to do that,
you have to do things like internationalization, make functions. And I think the next topic as well,
which is writing extensions and CNC plus plus. Yeah. There's a full API that you can write to.
And they do come with the installation of GNU org. And there's some quite nice things in there. I
did spend a half an hour looking at them. In fact, I think I installed one or two just to see
how they would behave. So, you can put quite high-speed additions into the into the language.
But, again, not really a thing for this series, I feel. Yeah, I think that's right.
So, that was the end of my list. Was there anything else you wanted to say under this heading?
I think that's pretty much it for me. I think, like you said, there was a lot of things that we could
have covered. And we didn't. But I think this is a good place to stop. Yeah. I think we've reached
the point where we've described, hopefully, enough of org to make it really useful in the sort of
command line bash script type context, but not to give you something that you could sit down
and write some really complex program in. You could do if you really wanted to. You can do it
instead if you really want to, but why, why bother? How do I ever want to do that?
Wasn't there recently a claim that said is a complete programming language?
Yeah, I think it is. Yeah, I did a series on said quite some time ago and did look at some of the
example programs that come with it. And some of those, they just, you can do some very powerful
things, but understanding is really hard to do. I certainly had some trouble working at what
it did. Yeah, writing a web server and said, I can just imagine what that looked like.
It'd be locked up at the end of it. I think you'd need to be.
So once next was my last point, and I'm doing all the talking at this point, but one of the things
I want to say was we had said along the way that we were going to try and turn these notes into
some sort of a combined document that we could put on the HBR site and archive.org. And I've
started on this while back actually as consolidating all the notes and stuff. And so I'm going to
carry on with that because I think it'd be a useful resource just to have everything together in
one place with an index. So you can find stuff if you want to. But there's no time scale at the
moment for this, but I would imagine within the next sort of year or something like that,
where you should be able to do it. I'm happy to do most of it, but it'd be quite nice if we could
bounce a few ideas off one another along the way. Hey, yeah, no, whenever you get to a point where
you want to share it, just, yes, I don't know over to me. And you can go through it together.
Yeah, that would be fun. That would be fun. Because it's been an interesting voyage this. I
never had originally considered doing a joint series, but it just sort of worked out this way.
And I think it was your idea to do the series, wasn't it? Yeah, it was, but I knew I definitely
could not do it myself. There's some topics. And actually, I've learned a lot along the way about
off that I didn't that wouldn't have learned without having this experience. So it's been good.
Yeah, yeah, it's been a lot of fun. It's been a lot of fun. I definitely, if the opportunity arose
again, and you know, and somebody who wanted to do a joint show, I'd be very keen to get involved
because it's been been fun to do. So yeah, it's been been good. So thanks very much for everything
you've done. It's been been a great time. And thank you, Dave. I mean, you've been kind of a mentor
to me on HBR. And so thank you for all of your work that you've done on HBR. And I could
help to continue to listen to you and then we'll continue to have conversations behind the scenes.
Yeah, that's great. Okay, then we'll say goodbye at this point then. All right, bye everybody.
Bye, this would be easy signing out and I continue to hack.
Yep, definitely do that. Bye-bye.
You've been listening to Hacker Public Radio at Hacker Public Radio. We are a community podcast
network that releases shows every weekday Monday through Friday. Today's show, like all our shows,
was contributed by an HBR listener like yourself. If you ever thought of recording a podcast,
then click on our contributing to find out how easy it really is. Hacker Public Radio was found
by the digital dog pound and the infonomican computer club and it's part of the binary revolution
at binwreff.com. If you have comments on today's show, please email the host directly, leave a comment
on the website or record a follow-up episode yourself. Unless otherwise status, today's show is
released on the creative comments, attribution, share a like, 3.0 license.