lee/hpr-knowledge-base
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Initial commit: HPR Knowledge Base MCP Server

Browse Source 
- MCP server with stdio transport for local use
- Search episodes, transcripts, hosts, and series
- 4,511 episodes with metadata and transcripts
- Data loader with in-memory JSON storage

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
This commit is contained in:
Lee Hanken
2025-10-26 10:54:13 +00:00
commit 7c8efd2228
4494 changed files with 1705541 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

461
hpr_transcripts/hpr3376.txt Normal file
View File

@@ -0,0 +1,461 @@
Episode: 3376
Title: HPR3376: Making books with Linux - part 2
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr3376/hpr3376.mp3
Transcribed: 2025-10-24 22:23:15
---
This is Hacker Public Radio Episode 3376 for Monday, 12th of July 2021.
Tid's show is entitled, Making Books with Linux, Part 2.
It is hosted by Dave Morris and is about 49 minutes long and carries an explicit flag.
The summary is, Part 2 of a discussion about how 2HPR hosts create books.
This episode of HPR is brought to you by an honesthost.com.
Get 15% discount on all shared hosting with the offer code HPR15.
That's HPR15.
Better web hosting that's honest and fair at An Honesthost.com.
Hello and welcome to another episode of Hacker Public Radio.
I am Andrew, also known as McNalloo, and I'm joined today by Dave.
How are you doing?
I am great, thanks.
It's a lovely day in Edinburgh and I believe it's a very nice day in Glasgow as well.
Yes, I'm sitting here in the dark looking out at the bright sunshine.
I should also mention that if you listen carefully we will have other guests.
Mainly a great tip to which is outside.
If you listen carefully you might hear him chopping your wee.
The bird sound is not so bad here because I've shut the door.
But if I were to leave the door open, the back door open,
you would hear lots of sparrows and starlings and everything.
I think it's good.
I like the ambient qualities of the background noises and stuff,
but sometimes it's not appropriate.
No, not I think it's great pleasant today and it's not fake.
It is real birds or what you might hear in the background,
but knowing them they probably all shut up now we started recording.
So what we're going to talk about today is, well this is sort of loosely connected to a show
that we recorded a week ago.
I don't know how far apart they'll be separated when they're released.
But last time when Dave and I talked he was quizzing me and I was talking about some document
creation scripts and methods that I had used and so we're now switching roles.
You're going to hear less from me and more from Dave on some interesting ideas and scripts
that he has been developing for, well I guess cataloging,
and HPR making notes and HPR episodes would that be a fair?
Yes, yes, it's my, the principle behind this is that I have developed a mindset that says,
I'm never a good at remembering things and things are done in order and stuff like that so
I like to write either recipes, bit of paper that says, do this, do this, do this, in this order
or I write, write bits of code to do it and that's what I've done in the HPR case,
you know, so I don't suddenly forget, oh no, I've got to put tags on that show.
Oh no, it's terrible and that sort of thing, I'll try and make them up on the fly and stuff.
So I have scripts which do a lot of, a lot of what I need to do to make an HPR show,
which I also use when a similar, I've got a similar set that I use when people send
shows into HPR and particularly when people send in shows with pictures, which I encourage,
pictures and, you know, bits of code or whatever that is this and again, much to be encouraged,
but it takes a bit of management when it comes in as a text file with possibly not even
intending it to be marked down. So it needs to be turned into marked down and the pictures
all stitched into it or whatever and I've written scripts to do that as well. So as you do, you know?
Yes indeed. In fact, I've, you know, sought your advice and how best to do that as a contributor
as well, I have a call and yes, I can't remember whether I got it right or not. You're a
two plate to see, I think. I try to minimize your luck. We haven't yet made a definitive
document of any sort. I mean, the process of putting one together that says, if you're sending in
what I call assets with you, just because I wanted to work. And that seemed like an appropriate
one. With your show, then could you please do it in this sort of way and then my scripts and
stuff can easily sorted out and then I can do a bit of automation of sticking the
sticking the links and so forth into the into the show. And actually, before we go into the details
your scripts, is it worth maybe just briefly describing the format of show notes with images
that you would perform? And how would you if some show notes arrived and they're all neatly done
mark down and somebody has included an image in the show notes and uploaded an image when they
with their show. What structure would you like to see in the mark down? Ideally, you go, oh good,
this is no work for me. Well, it's fairly simple actually. If all those things that you just said
are true, because mark down has the capability of referencing an image, it can be really
relatively straightforward to simply, you just need to know the URL of where your picture will go.
And we do, it's pretty obscure, but it is in the documentation. What we do is we have a
directory where all the shows go, all the audio goes, the actual notes are in a database.
So if we have assets, if you don't mind me using that term, then we create a sub directory,
which is named HPR1234, whatever your show number is. And in it, we put the various extras
and stuff. Ideally, we want you to put an index in that directory because the way that the
web server set up on the HPR machine that server is, but if you have a directory and you don't want
people to browse it, you need to put an index.hml file in there. So yeah, it's not that simple, it's
in our website, but given that you know that your picture is going to be, you know, the HPR
site slash blah, blah, blah slash HPR, so and so, and so, then the name of your picture.jpeg
or .png or whatever it is, then you could put that all together and I make it. So however,
most people have don't send in that level of markdown. There are some people who do brilliant job,
some people who send in what we originally asked for, which is HTML, but you know, get again,
the links right in that can be can be quite daunting. So what people tend to do, and this is
something that sort of evolved really, is that they say they write their text, not in markdown,
we're just in plain text, but plain text in markdown, the base level are pretty, pretty similar,
aren't they? The way people lay out text, you know, with a you might put a header in as a separate
line and you might put a make a list by putting a hyphen at the start of each line and so forth.
So if they send it in with some sort of meta thing in it that says picture one here please,
or something like that, Dave, could I picture one in this place or something, something like that?
What always be me, but you know what I'm saying, that something that makes it plain that here
is where picture one is to go, then I can actually turn that into a thing which I can turn into
to markdown with the links in and I can compute where they're going to be and get the order from
your order and so forth. So that's that's where we are at the moment, does that make sense?
Yes, yes, that does, that does make sense, yeah, that's fine. So anyway, so that was a bit of a
inauguration I threw in there. So what is it specifically that you'd like to talk about
with the work that you've been doing? Well, it's two things really, there's the whole infrastructure of
how to how I found it useful to put together notes for HBR and there's also what I'm trying to
do in terms of a book, call it book is a bit of high-faluting really, it's a consolidated
number of of show notes in a more readable form. That's why we started from really when we
were talking about this, but we've sort of gone into the area of HBR show notes a little bit,
which is I think quite relevant and interesting hopefully anyway. So if I mention what I do with
show notes in a general sense, I have found that it makes life a lot easier for me if I
interspers my notes with a sort of another markup thing. For years and years and years I've been
using a pearl tool and there's a equivalent called template toolkit. It's what do you call
these things? It's a templating system I suppose what you call it. They exist for all sorts of
languages and independently of languages I think as well. The principle of it is that you put
a tagged item into your text, which is seen by something that pre-processes it. So the pre-processing
would go through the file and say, oh look here's one of my tags, what am I to do? And in some cases
it's merely a variable being created or could be code being created. There's all manner of
capabilities that exist in there. What I often do is to make these template variables to contain
long convoluted URLs and I put them all at the top of the document and then when I come to make
a link to that URL or this in marked area tends to be two cases where you want to make links. One is
if you are just making a list of links to the things. The URL part is where you put your URL but
there's also a thing where you might want to put things which are multiple references I think they
are where you say you know in document ABC and then you put in square brackets a number after it
this is in marked down and then what happens is the word gets highlighted and a link is made to
an item that you put at the end of your marked down document tagged with that number or anything
actually but in there you have to have the URL so because I know I would screw up the
writing the URL at first time and then the second time and so I make template variables to hold it
and then I can just put the template language to allow the item to be inserted.
Yes that makes sense. I've used something, I haven't used it but there's something similar going on
behind Pelican which is written in Python. I forget what it's called now but yes and I know
that the Ruby for equivalent of Pelican uses something similar. What's the one that you're using?
Did you see that? Maybe I missed the name. It's called template toolkit. It's usually referred
to as TT2 and there's also a Python version of it now that it came from Perl originally and it's
was originally used to enable you to make generic HTML so you'd write a piece of HTML that said
put the footer here, put the header here, the footer here, put the body here and then those
things would be references to other files which just got pasted in at the time that the process
are processed. It would work for static pages as well but it could be used for dynamic pages as
you sort of fit. That's where that sort of stuff came from and loads of other people that are
there. It's in PHP, it's in all sorts of languages, that stuff, style of idea.
Yes, I think it's either called pigments or ginger or ginger to the Python one.
Yeah, Python 10 one. I've never used it but I just noticed it as a dependency when I was
comparing pelican ones. I used to use it at work quite a lot because we serviced the mailing list
feature we run mailman eventually. So we offered a mailing list feature that
there were mailing list that sent messages to all students and all staff and that sort of thing
but there were also the capability of people who remember the staff at the university to request
a mailing list to go to their research group or whatever and we created a form that they needed
to fill in on the web which sent a standardized request in and then I wrote things that took those
forms and did the necessary and also sent back or the plate mail reply. Sometimes it was said,
you know, I just made it simple so you could just say send boilerplate or something.
You're not authorized to have a mailing list or the boilerplate which so yeah, I've done it.
This is all done with just sort of command line scripts that you take a mail message,
chop it up into pieces, rebuild it through the template with boilerplate stuff and so forth.
So I found that template toolkit was fantastic for that sort of thing because you could make
umpteen different, you just have one script and then you just told the script, use this template
and that caused it to do one thing and this template made it do another thing. So it's
remarkably useful for all sorts of things like that. All right, yes. I didn't realize this seems
to be the ancestor of the things I come across and it really went quite so far back as well.
Yeah, it's, I think I was probably playing with this in the early 2000s, that sort of timescale
when Perville really was the language that you did all this sort of stuff in. But yeah, it's
it's very convenient. I find myself writing scripts of a general nature these days where
I put a template capability into it. So if I want to report out of it or something,
then I can just say run against this template and it just throws whatever data structure it's
got and that comes a form which might be HTML or CSV or anything you like. Depending on how you've
written the template, the work is done in the template, but you're just throwing it a bunch of
data. I mean, yeah, I'm sure you're well aware of how this sort of stuff works, but just in case
people are not fully up to speed with this type of stuff. It's a it's a fun way of doing things.
Yeah, no, it's definitely worthwhile explaining that because actually I'm not that familiar with it
for starters. Secondly, sometimes you know, we're listening to HPR shows, some dimension,
something as if everyone's going to understand it, but that sounds interesting. I don't understand it.
Of course, we know that Ken would be there. Do it show, do it show.
Yes, yes. These templating systems are quite good. The template toolkit that I know the best
has got the ability to, loads and loads of plugins, for example, so you can, if you wish, put
a database plugin into it. So within your template, you can do database queries and you can
then format those queries into markdown in a document, which then gets printed in a pretty form,
or it's pretty as you can do with markdown, which has a few limitations in that sort of area.
But it's just really convenient to be able to do that sort of stuff. I developed my
show note creation stuff, where I often do shows about how to do things in bash or whatever,
and then I do some example scripts, bascripts. And what I do is to have the bascript available to
the infrastructure that creates the notes. And I can put a reference in the template that says
at this point, run the script or show the script, maybe as part of the text, and then run the script
and show the output from the script. So that's a couple of template calls. There's a macro type
facility, which will actually make external calls and that type of thing. So that has been
amazing and useful, because if you're ever doing things where you're running a script and showing
the output, the tendency to tweak that script, forget to put the tweak into the notes, and then put
the output. And then you look at it and think, how on earth did that script ever produce that
output? Well, I do, anyway. And if it was the script itself that made the output, then it's a lot
easier. I mean, my son's doing this MSc and computer science, and they use Jupyter Notebooks.
So Jupyter Notebooks, yeah, quite a lot, which has that principle in it, but you can put your,
I think they have to submit their work sometimes with their program that they've written in the
Notebook with the output, which I think the reader can just run and see what it does. So, you know,
what I've done is very basic and comparison to the way the world has moved to do these types of
things, but it's just a fun thing. Yes, yeah. Yeah, it's, I certainly have a pandemic introduced me
to the Jupyter Notebook stuff. I wish the Python thing, isn't it? Yes, I believe so.
Yes, and I was like, just check out and remembering the right thing. And I was quick and
traced with that and the models he was building. It was all really easy for me to inspect and for
the learn with. Yeah, yeah. I think so. I think I went to a talk given by the Edinburgh Linux,
who's a group a couple of years ago now. Wow, times for the learners. But the guy who was managing
the Edinburgh University, Jupyter Notebook's engine for Edinburgh and a whole bunch of other
universities, I think, as a service was there talking about, you know, the sort of things he was
being asked to do. So you can build notebooks for different languages and this type of thing
is incredibly impressive indeed. Yes. So, so when a, what is your hoping to have produced
when you get to the end of us, if you're or is it just a sort of long, ongoing task?
The production of HBR show notes has been through a long, long development where the goal was
simply to make something that I could use and was was free from, fairly free from, fairly resilient to
the Draftlady, you know, running it wrongly. And so me being the Draftlady, which had more or less
achieved actually. So I do things like I start writing some notes and I, in the majority of shows
that I produce, I make a long set of notes, which is pages and pages sometimes and separate that
from the notes you see when you actually bring up the show on HBR that the, you just get an overview
in the short form of it. So I start writing the long thing and because the way I've built the
template, it auto-generates the short stuff as it goes, you know, so there's a, as long as you type
your overview in a pre-assigned block. Sorry, my chair is very creaky today.
That's okay. It's better than the birds. No, no, no, prefer the birds actually.
So long as you put the overview in the right place and or you put your links in the right place,
then you're not with an overview and links in the short notes and you have to do anything to it.
In fact, if I do do something to it, then I'll break it probably. So things like that I do.
And I also determine what, what files are going to be associated with the thing and that's a sort
of incrementing process. So that I've decided I'm going to have a script here which demonstrates
that feature I've been talking about. And then, and it's some examples that run and produce
output dish to prove that it does what it's going to do. So I would put those together and put
them in a, in a place where the software can find it. Each of my shows has its own directory
associated with it. And then I make a sort of parts list manifest whatever you like to call it,
which just link it just has the paths to the various files. And then I've written a pre-process
of that goes through all of this stuff and feeds the template. The template then generates
markdown and the markdown is given to Pandock. And now comes HTML with all of this stuff
hopefully stitched into it. So when incidentally you were saying about monitoring the development
of the program of a text, I mean, you're using ENTR. How are they supposed to pronounce that?
Yes, that's right. So you told me about that years ago actually. And yeah, sorry, I interrupt you do.
Well, I was just going to say that you've recently suggested a better alternative. Because I
would do it because you could instruct me to page refresh which you couldn't do with Chrome and
Firefox and other browsers that I could interact with them at command line. And then I could
cause use ENTR to trigger the page. We do need to do page refresh. So I could literally type
in the markdown file. As soon as I saved it, ENTR would pick that up and tell me do they to
to fire a page refresh. So I would, you know, every time I saved it, I would get an instant preview.
But was it the fault in browser you said can do this with those together?
Falcon doesn't need to be told. It seems to monitor. And whether it's only files, I don't know,
because, you know, in my case, I would have created a file in my file system that contains the
the HTML that I'm going to send to HVR. And Falcon just says, oh, it's changed, right? And then it
just refreshes it as it changes. So it's doing something like what ENTR is doing all by itself,
which is I don't know whether you can turn it on or off, but it's on by default with me. It's
really good. Yeah, and Falcon, it's built with a key. And the clue being that it's a, I think,
when we in Tuxtam would have viewed chaos. And then that's when you prompted you to mention
Falcon, because this Falcon was included with chaos. So the key is an indication that it's part of
key to E. Yeah, I think so. I think so. It's, I'm not sure what it derives from.
One of the well-known browsers, I think. I think it offers, I think it offers sort of tab better
tab management. You remember how I think Chrome and Firefox had really fancy capabilities that let
you have lots and lots of tabs that you could go and group together and go back and fish or anger
them. Then they took all that out, which I missed enormously, because managing tabs was a great
thing for, as far as I was concerned, I think Falcon's got some of that in it. So, you know, it's
relatively easy to manage the things that you're working with through Falcon, which is good.
Yeah. Well, I'm just looking because I assumed that, I mean, the browser, I assumed that Falcon
replaced Conqueror, which is the, the browser that KDE has not used for that long as I've used KDE,
so KDE 3.0 and something like that. But in my new Slackware, which has KDE with Plasma 5,
it has, I still have Conqueror, but that may be because I haven't actually installed, I haven't
removed old packages, so, but Falcon is there. I hadn't noticed the Falcon must have appeared
when Plasma 5 came into Slackware, and it says that it's a QT web browser, previously known as
Copzilla QUP,zilla, or QT web engine for rendering. So, it's distinct then from both Chrome and
Firefox, it's separate to that, man. Yes, I'd forgotten that. I'm confused because I've got so many
browsers, I fiddle around with, yeah, yeah, yeah, and it doesn't have quite the powerful features of
tab grouping that some do, but, you know, that's by the buyer, really. Yes, and, well, I mean,
the fact that I've already got it installed means the future I'll be using Falcon to do this
job because it replaces both Enter and majority, neither of which, I believe, come with Slackware, I had
compiled myself, which wasn't both very easy, to be honest. But if I've already got Falcon to do
the job, I might as well just use that from now on. Yeah, see how you get on with it. It might not
contain all the browser features that you want. I'd mainly use it to point at HTML files that I'm
developing, but it seems to be great for that. It's very responsive and it's got pretty much
everything I want, hopefully the same with you. Yeah, or actually, now I see that I still have to use
Enter, because the other thing Enter will do is fire the script that turns the background to HTML,
so Falcon can't do that, but it's got something of clever hook in it. I don't know. I'll go and read
because if it can do, yeah, you'll have it all. Yeah, yeah. I think I said last time we spoke that
I use make to do that, so make spots that there is work to do. I run make out of them, so I do
something like make all and it just goes and says, oh, you've updated your notes. I will just go
and those notes as well, the usual stuff and it's got rules for doing all the pan docking and
putting through template engines and all that sort of stuff. So yeah, yeah, so that's
that's quite, I actually made it, I haven't quite got to the point where it uploads the show yet,
but that would be my ultimate goal. I think it can be too, please, if I did that.
My original bash script, which I was going to share with the world, did do that because
it could use FTP to upload stuff, but now we've got a form. I could obviously write to the form
with curl or something, but I have never got around to developing it because you can't really
poke around to somebody's form to develop your automation behind the scenes without causing
a spare amount of grief, I would have thought. Yeah, so I suppose one thing that could be done
as, rather than using curl, the way you described it, the way to do it is create some kind of
rudimentary API on the HPR website so people, because the community is a kind of community that
would like, so I'm sure many, I would do it. I would like to submit my shows through an API,
that would work for you. Yeah, it has been discussed actually. It has been discussed. I'm not
sure how far we've got with it. I think we suffer from failing to develop in the way
the way that you'd like, because everybody's so busy doing their own lives and doing their own
jobs and stuff like that. It's proved to be really difficult to get a huge lot of development
on HPR done. I tend to be hacking around on things myself, but just a fairly small level,
not the big redesign of API level. Yeah, so I mean, it comes up to ideas as relatively easy,
actually implementing as a team consuming. Yes, and certainly, I mean, HPR is quite a loose group,
and more formal groups, like a lot of trustee's and company boards, not even that formal, but
you know, people have to come together and get something done and have legal responsibilities.
I'll always make a big fan of, if you're going to suggest something, don't just expect
somebody else to do it, you know, suggest it with some idea of how it might be done, and
possibly with the idea that if you think it's a really good idea and nobody else is going to do it,
then perhaps you should do it. The downside of that is that inhibits people proposing ideas,
if they think they have to execute with an idea they've got, but on the other side, it stops people
coming up with great ideas, which require a finished amount of time to implement.
Yes, yes, having worked in schemes where there's a project manager who says,
oh, what we really need is this, and walks off thinking that that somehow has magically made
that piece of software come to exist. Without really any other thought about resources or
people or anything, then yeah, I think I was working in an environment that wasn't really well
well set up for that sort of thing. Being a university, universities are not all that brilliant
than that sort of thing. Anyway, yeah. Yeah, I've just given you to mean a website,
has a big button on it, and you press it, and it says, create world peace, you know,
it's easy to implement in that website. Yes, it's a little bit more difficult to get into
doing anything. Yes, the thing that's happening behind it is something's been written to a file,
had a world peace request on this date at this time, right? That's why I've sorted then.
It's world peaceful, equals true. So yeah, so yeah, really my message regarding documentation
and stuff is that I've come up with the thing that lets me make HPR show notes relatively
easily, of the type that I tend to do, which is long and long-winded and detailed, and it's
it handles shortened and punchy as well, but it's not the thing that I originally visited,
which was something I could hand over to the world and say there you go, it's just it's just
been personalized to the empty degree and it's weird, and you know, people go, what is this
nonsense? Who would you ever do that and stuff? So yeah, it's still worth hearing it, I think,
because even if somebody's reaction is that, you know, what I've done, I mean, this has happened
to me before. I mean, I'm not really a software developer, as such, you're not professional,
one am I do, obviously, right software, and I have made money from doing it over the years,
but I don't regard myself as a professional software developer, just as somebody who
plays cricket this Saturday isn't a professional cricketer, but there wasn't an occasion where
I'd written this thing, and I think it was, you know, I used MediaWorkie, which is what, you know,
Wikipedia was based on, and there was a phase where like every other website had set up with MediaWorkie
and was doing something, and I was just, I was written PHP, I think, and there was a little hooks
you could put into it, so I don't know these clever things with the PHP, but it quickly, you know,
to extract numbers from a database and display them in tables and do some elementary processing,
to make it more interactive than a usual wiki would be, what we're doing this project is not
so important. Anyway, it quickly became a nightmare, you know, it sounded like a very good idea,
but because I was just cobbling it together myself in my spare time sort of thing, it quickly turned
into a nightmare to maintain, and then this guy came into the company, and the first thing he looked
at was this, I asked him to look at it, anyway, oh my god, this is awful, how would I have to maintain
this, and I went, actually, I don't really know how to anymore, so he immediately said, well,
this is how you should do it and then went in and did it much better than I did, you know, so I think
that, you know, that can happen, can't it, if you share your stuff, show somebody else, and then
they realize, with benefit of hindsight, your work is then valuable because it can show somebody else,
and how they think, oh well, that's what you want done, I see what you want done, I see how you do
it, but now I'm going to go at it from my way, you know, I can be good. Yeah, yeah, no, you have a
point there, you have a point, I spent a lot of my working life creating things to solve
immediate problems, and often they were conkey things that worked, but were not, it was hard to
expand and whatever, when I started working at the university, there was no automation, that
let students get accounts on the central machines, it was a guy who sat in his office all day long,
and students came and knocked at the door and said, can I have an account on the machine, please,
and he said, yeah, and he'd sit there and type stuff into some bits of stuff, and that would
come a bit of paper that said, there you go, there's your account, off you go, but we were dealing
with, you know, thousands of people and remote people and stuff like that, so I ended up writing a
whole, there's a name for these things, but it's an account management system that created all these
things, but it was done to my way of doing it, you know, it was a case, oh, we need this right,
quick, quick, quick, you've got a day to fix it, right, it's fixed, and you look back and
eat it, oh my god, what happens if that bit of string breaks and it all falls to the floor,
and I handed that over when I retired, I spent months explaining how this stuff worked,
and handing this stuff over, none of which they were going to run, they were going to
re-implement it in Oracle thingies, Oracle offers a similar product that lets you hook
into your student database and create accounts and give people resources and roles and control
and stuff, so you know, it's a big booming area these days, so yeah, so there was big handover
of knowledge, not of software, it was the knowledge handover that was the useful thing,
did you listen to everything, I said, but they did listen to 90% of it, so that was good,
well that is good, because one thing I was, I mean I had my university education at Glasgow
and worked there for one and all over the years, and I was quite shocked, and away when I left
in the 90s, in the late 90s, everything was done in Linux, Unix systems, and lots of clever
people that you get in universities, new, new about computer, especially in the physics department
and some other departments, to including computer science, and the computing service itself,
I came up with like homebrew solutions, and these homebrew solutions were used throughout
different universities, certainly around the UK, some internationally, and I thought, you know,
that really was really good, but when I came back, I didn't work for Glasgow University again,
until maybe 10 years later, when I came back, it was all Microsoft, you know, all this
sort of, all the stuff that they developed for themselves had gone, and the university had bought
into some, I don't know what it was, teams or something, whatever, you know, exchange, mail servers,
and all the rest of it, and I was really quite shocked, you know, to be honest, that, you know, and
of course there were all complaining that they couldn't update things, and things weren't working
quite the way they wanted it, and they couldn't change it, and the university was paying lots of
money, so there's all the grumbles, and I think maybe I've got rose tinted spectacle, but I don't
remember seeing that, you know, because the people that developed a lot of these systems, like you
developed your system, were there at the university, and you could go and talk to them, you could go
and ask them, and you know, and you ended up with something that's much more bespoke to how that
university worked, so, and yeah, so I was a bit disappointed, but I don't really know enough
about to see how bad or good a thing it was, but it wasn't, it wasn't a good thing, it was, it was a
bad thing. Well, having been in a university while that process was happening, there was a,
the way that universities were funded changed a lot over my time work in there, it used to be
the central, central funding for certain, particularly computers, I mean, when I
started work there in 1981, then all of the central computing facilities were paid for by
government through a, through a body set up specifically to do that, and that I think
was shut down in the early 90s that sort of time, I don't remember precisely, they were called
the computer board, they had, you know, sort of a fair amount of money, we got a million pounds
in 1987 to buy a replacement system, which was due to last but a seven years before getting more
funding, so it must have still kept going beyond those seven years, so it wasn't sort of mid-90s,
maybe the thing pitted out, and of course universities at that point were on their own, you know,
use your own budget to buy your computing fees, and nobody, at least in my experience,
who were managing these places, knew what the hell to do, you know, they were,
the first reaction was, we'll bring in consultants, and the consultants came in, rubbing their
hands of Glee because they had this wonderful bunch of people who knew nothing about what they
did know a lot about in the, in the main anyway, at least at the management level, and you know,
they bulldozed all sorts of nonsense into, into the way things were, were set up, because the
the answer had to be that you bought stuff from the big boys, you don't, you don't go, there was,
there was a thing, you know Simon Fips, he was, he was working for Sun, when Sun went for much more
open source with their, with their stuff, and then Sun got bought out by Oracle, Oracle wasn't it,
and he left around about that time, and he went to work for a company called ForgeRock. ForgeRock
was a company that was making the sorts of student, well, account management, things like I was
describing, so, you know, you had to have a source of who the people were from your HR system,
or from your student record system, whatever it was, and you fed the data into the system when
it spat out, whatever you told it to, through filters, so to make accounts and give people access
rights and all of that sort of stuff. And we looked, there was, there was a strong recommendation,
but we look at ForgeRock as the way to do things in the university, but the management
would not touch it, because, you know, here's this small company making this product, whether it's
good or bad, we don't care, but how they still going to be here next year, you know, five years time,
Oracle, on the other hand, we know we'll be there forever and ever and ever, you never get rid of them,
you know. So, yes, yeah, you've got your, yeah, it's like the, you know, people wanting security,
well, being in prison is pretty secure environment, I would do it, that would do it.
Oh, we've maybe vatured off into the subject of a feature, HPR episode there, that's actually
quite interesting. I don't know if it's one, you know, talk about one, I certainly have my, you know,
my contact with it too, from a different capacity to yours. Well, yes, I, um, the whole business of how
UK universities functioned, really just from my limited experience of it. So, um, I was in a managerial
role at that time, and I was not happy with what the way things were going, so I was quite keen
to take the early retirement offered me in 2009 to get out, but one of the things, and his, his,
I've never, I've always hesitated to mention this, but I'm going to mention it now,
when you get a deal, but on universities who created this, they say, okay, we've got all these
people, all of these an old guy, I was in my 50s at that point, so, oh, and he's only in fair
bit, uh, right, let's offer him the offer he can't refuse to get out, right, get your pensions
starting up early and off you go, and so I come to, to, to say this, and they really do put the
pressure on, and, um, they say, but before you get your, uh, leaving bundle, whatever,
your, your free stapler and stuff, um, that wasn't right, but, uh, um, then you have to sign this
non-disclosure agreement that will prevent you from saying anything nasty about us into the
future, so I had to sign this, and that, to me, was a strong indication that things were well
off the rails in that organization, I was very pleased to get out of it, so, uh, so here's me going
against this non-disclosure agreement, but, but you don't know what the university has to do, so
that's fine. Yes, that's true. Yes, we haven't mentioned that. No, indeed not. Well, I do think
that, because I have, not personally, but I have friends and did my, my own parents, who, uh,
some older than you actually went through something very similar, and a lot of the things that you
described are familiar stories, uh, that they told me, and so, uh, yeah, so I may, that may be an
interesting thing. Well, yeah, it would be, it would be an interesting thing to, to talk about,
and the whole, the whole subject of, if you have a, a number of employees,
the, which, a very large proportion are, very skilled, highly motivated. What's your best approach
as a manager to keeping them doing what they do? Well, do you tell them that they're all
bunch of idiots, and not to be trusted, and you're going to build this infrastructure that says,
okay, every week you come and report what you did, and we, we score it, and then if you get a low
score, then you're very, very naughty, and you have to do much better than following week.
You build that sort of a structure for that sort of group of people. Then, do you get a good result?
This is, this is a very interesting question. I'm sure there are many people who farm or
skilled in this subject, and I am just being a victim of it, but you know, it's a, it's a,
an interest, and it happens all over the place. I mean, the demotivation of skilled and
trust, potentially trustworthy people by failing to trust them and treating them like, you know,
cogs in a machine is well known to be a destructive force in the world of work these days.
Indeed, yes, and I mean, the way I viewed it as, that could have practiced that you just describe,
perhaps, like, target-driven culture of management, it was imported from the commercial role,
or from what academics like to call industry, or the academic school. Yeah.
Everything that's not actually made industry. So commercial companies tried this, and it failed,
and weirdly, the consultants who are peddling it went into universities, I guess, because,
and sold it to them too, as you described. But anyway, I am going to have to go.
Yeah, yeah, yeah. So, but I do think I'd like to talk more about that in some future episodes.
Yes, yes. Well, it would be an interesting subject. I guess how there would be a few people
listening who might be, might be interested, might have their own experiences to share as well.
As regards the making of books and stuff like that, I think, I don't know, is there any more mileage
in this subject? We've covered, in general, quite a lot of stuff here, so we could just leave it
at that point. I don't know if we want to get into any more detail, do we? No, no, I think I've
said something. I said all that I would see in the last episode, and I was enjoying hearing what
you described too. So, yeah, no, I think, for the thing being that said, but you know, I might go
off and get some ideas to try and, you know, those templates are not something I consider using
myself, but I think maybe I might. Well, it would be interesting maybe to revisit this subject
months down the road, just to say, you know, we've both taken a different look at the problems
that we were kicking around and come up with some different approaches, so let's have a little
chat about those, that might be some mileage in that. Yes, yes, definitely. So, I guess all
that really remained, unless it's any of the else you would see Davis just to say goodbye, and thanks
to everyone in HPR for listening. Yep, that's me. I've finished 12 goodbye to everybody, and
thanks Andrew. It's been been fun having these chats. You've been listening to Hacker Public
Radio at HackerPublicRadio.org. 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 HPR
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 founded by the digital dog pound and
the infonomicon computer club, and it's part of the binary revolution at binrev.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 live, 3.0 license.