Talking Postgres with Claire Giordano

CLAIRE: 00:00:05
Welcome to Talking Postgres. It's a monthly podcast for developers who love Postgres. I am your host, Claire Giordano, and in this podcast, we explore the human side of Postgres, databases, and open source. And that means we try to figure out why do people who work with Postgres do what they do and how did they get there? Thank you to the team at Microsoft for sponsoring this community conversation. And I am so excited to introduce today's guest to you. Peter Cooper is our episode 28 guest. Peter is joining us from the UK. And if you don't know Peter, he was formerly chair of several O'Reilly conferences, conferences like Fluent and OSCON Europe. And today and since about 2010. So what is that? 12 years. It says 2010, but that's more than 12 years. We'll have to talk about that. That Peter has been founder and editor-in-chief of Cooper Press. And it's a company that publishes about seven weekly newsletters that reach almost, it's about 460,000 developers and engineers, including JavaScript Weekly, Front End Focus, Ruby Weekly. I think that was the first one. There's a newsletter for Golang and for React and Node. And my favorite, my absolute favorite, is Postgres Weekly. And one of the things I love most about Postgres Weekly are the insights and perspectives that get shared in the very short blurbs that link to new content, new articles, new videos, things that have been recently published. And it's a newsletter that's industry-wide. So it covers Postgres content from all over the Postgres project, around the globe and around all the different companies. So with that long preamble, welcome, Peter.

PETER: 00:02:03
Hey, how's it going?

CLAIRE: 00:02:04
It's good. I'm so glad that you're here. Today's topic?

PETER: 00:02:08
It feels like we've known each other forever.

CLAIRE: 00:02:11
It does, but we've only met in person for real once, but actually probably twice. [Yeah.] I think the very first time I saw you, maybe I introduced myself, but you probably don't remember. But yeah, I'm glad you were actually in the San Francisco Bay Area a couple of weeks ago. And that kind of led to today.

PETER: 00:02:33
Yeah, you finally managed to convince me.

CLAIRE: 00:02:35
Although I am so happy after years of trying. I don't know if you know this, but Michael Christofides, who is one of the hosts of another Postgres podcast called Postgres FM, he sent me a note a number of months ago saying, Ah, Claire, I think it would be great to get Peter Cooper on a podcast. [LAUGHS] And I think your podcast makes the most sense. So make it happen. And hopefully he knows this is happening and he's going to listen to it afterwards.

PETER: 00:03:03
Yeah I know his show, I promote it quite often. So yeah, fantastic.

CLAIRE: 00:03:07
So the topic for today is 12 years of Postgres Weekly. [Yeah.] Oh, so how does that work? I get it. You've been working on the newsletter business since 2010, but maybe Postgres Weekly didn't happen for the first couple of years. Is that how the math works?

PETER: 00:03:25
Yes, that's correct. So Postgres Weekly didn't turn up until quite late in 2012, I believe. I don't have the number in front of me but I know that it was during the Postgres 9.2 era so that kind of puts it into perspective a little bit.

CLAIRE: 00:03:42
Okay. And when you were in elementary school, I actually don't know what they call.

PETER: 00:03:47
Yep. We're skipping right back.

CLAIRE: 00:03:50
We're skipping right back. We're going to the beginning of your origin story. And you know how you have to... [When I was born...] When you were born, how about where you were born? I'm presuming you were born in the UK.

PETER: 00:04:02
At least I can say something interesting about where I was born, so I was born in Greenwich pretty much very close to the Meridian line, so I know, you know, I was born at nought degrees, and actually have pretty much lived on that line, like the nought degree Meridian line actually runs through my current town I live in. I'm much further north now in a place called Lincolnshire but we have the line run through our town so I've always lived on that Greenwich Meridian line so I'm a GMT child all the way.

CLAIRE: 00:04:31
Okay. Well, I don't think my question applies to when you were born. I want to fast forward to like elementary school. But you know how you have to fill out those pieces of paper to bring home to your family that say like, my favorite color is this and my favorite thing to do and my favorite food. And usually one of the questions is, what do you want to be when you grow up? So when you filled out that, what do you want to be when you grow up? Did you say, I want to be a publisher that focuses on developers and giving them really useful content? Is that your answer?

PETER: 00:05:06
No, not quite. I don't remember actually doing that exercise, but my parents always remind me that when I was a child, I already wanted to be a bus driver. So I set my ambitions high at an early age, but unfortunately never got to drive a bus and instead discovered I was a lot better at sitting behind a keyboard and typing. So it wasn't until my teenage years, though, that I did actually figure out kind of like what I'm doing now almost. So there is a little bit of a story to that. But yeah, it took until I was about 14 or 15 to figure that out.

CLAIRE: 00:05:39
Okay, so what happened when you were 14 or 15?

PETER: 00:05:43
So I pretty much programmed my entire life, probably from when I was around like three or four, just because my father was really heavily, heavily, like geekily into microcomputers in the 80s. So he had them all over the place. He had all sorts of different ones and Spectrums and BBC Micros and things that were popular in the UK. Machines that didn't necessarily extend to the US. We had some rather odd machines. And of course, you know, I'd start to play with them and he encouraged it. And I was just into, you know, just messing around. Obviously, you know, it wasn't building, you know, databases or enterprise software at that age. But just, you know, messing around and doing loops and all this type of thing. So programming has always been a part of my life, you know, before I can even remember. So it's kind of been a natural thing for me but when I was a teenager I kind of got a lot more into it with like demo coding so that's where you like make you know graphical demos. This is back in the DOS days so you know it was it was all quite rustic let's say, but I got onto the yes I got onto the internet in 1990 I'd say early 1995 and I was basically addicted to [Okay.] a part of the internet called Usenet which is kind of around now but it's used for pirating movies and stuff but back in the day it was kind of like the equivalent of let's say like Reddit or forum software but it kind of ran a very low level on the internet you know it's just text but there were all these different groups where you could discuss with people you know whatever your favorite topics were and so I was actually on a lot of groups like I was on the JavaScript group back in 1996 so JavaScript had just come out and you know people are on there asking questions and you know we were chatting and everything but I was very heavily into BASIC programming so you know kind of a slight Microsoft connection there. I was using like QBasic and professional basic that you know Microsoft was producing at the time QuickBASIC 4.5 and I started what's what I called a fanzine where I would get submissions from people on the news group of you know cool programs that or tips that they had and then I'd package them together into like a you know plain text kind of like newsletter format and then once every couple of months or so I would then you know put that onto the group as a published thing and the funny thing is you can actually go back and there are still archives of these things so I can go back and look at you know the things that I was publishing when I was 14 and 15 it's quite fun to see.

CLAIRE: 00:08:19
Have you showed your family? Would they get a kick out of this?

PETER: 00:08:25
Not my current close family but my you know my parents were aware of it and everything [Okay.] but my kids wouldn't find it very entertaining at this point. It doesn't involve Peppa Pig or anything like that so, yeah, my children are not into technology whatsoever which is rather bizarre given my background but yeah I think it's their rebellious streak.

CLAIRE: 00:08:45
Well, maybe there's still time. You never know. They surprise you.

PETER: 00:08:47
You never know.

CLAIRE: 00:08:49
They pivot. They change. Who knows? [Sadly.] Okay, so you kind of found your calling at 14 or 15. But did you know it was your calling or no?

PETER: 00:08:57
Well kind of, [You just...] no, I mean it was just my hobby so when I eventually you know went through education and I was going to go to college I was going to go and study law so my kind of you know plan was I wanted to become a lawyer essentially, that didn't actually end up happening because my parents plan backfired on them so my plan after I finished high school was to you know just sit around and do what I wanted all summer and then go to college and my parents you know I grew up in a very working class family and they were like well no you're not gonna laze around all summer you're gonna get a part-time job or you know some kind of job, so you know you're gonna pay your way so you know I pretty much paid rent since I was 16 and you know which is fair enough that's just the way I was brought up so I went and got a job and because this was like the late 90s the whole you know we're kind of early on in that dot-com era very early on and I managed to get a job at a what was then called a new media company in London who were building websites that you know some of the earliest you know big budget websites and I just kind of got hooked on working so I was okay, I'm not going to go to college. I'm just going to keep working in this industry. So yeah, that kind of backfired. So I never actually became a lawyer. I just kind of fell into a job that naturally meshed with my personal interests. And it just happened to be there at the time, but that wasn't the plan.

CLAIRE: 00:10:22
I can't imagine your parents' conversations, like talking about how it backfired and how, oh, we shouldn't have said that. We should have done something different. But of course, I'm sure that they were quite pleased with how you landed, even though it wasn't the plan.

PETER: 00:10:36
Yeah, but I still didn't get to goof off that summer. I basically spent, you know, I was doing nine hour work days with a three hour, you know, commute each day as well. So it was like, yeah, ridiculous. But I get dropped into the, you know, the work world and, you know, really at the deep end. And, you know, just went from there.

CLAIRE: 00:10:52
Okay. Now, for anybody listening to this podcast who is unfamiliar with Postgres Weekly, you should definitely go check it out. postgresweekly.com is the right URL.

PETER: 00:11:02
Correct. Yeah.

CLAIRE: 00:11:04
And I remember when I first started working at Citus Data, it was maybe my first day at work, my second day at work. It was absolutely on the to-do list from my boss saying that you have to subscribe to this newsletter. It'll be really useful.

PETER: 00:11:20
What year would that have been?

CLAIRE: 00:11:21
What? What year? 2017.

PETER: 00:11:25
Okay. The reason I point that out is just because for the first couple of years of the newsletter, we had a different person was curating and editing it. So our first editor was someone called Craig Kerstiens, who a lot of people in the Postgres world will be familiar with. He essentially phased out over time. But yeah, by 2017 or whatever, it was pretty much me doing everything at that point.

CLAIRE: 00:11:50
Okay. Yeah. And Craig worked at Citus Data as well. So I worked very closely with Craig.

PETER: 00:11:53
He did. Yeah.

CLAIRE: 00:11:56
And so maybe that's part of why Umur suggested, oh, you have to go subscribe to Postgres Weekly because he knew about it from Craig. Or maybe it was just, you know, pervasive and considered generally useful. I do know I got a note from someone who is a senior leader in one of the Postgres, one of the several, many Postgres startups in the world. And he said that, quote unquote, I'm reading his DM to me right now. He said, Postgres Weekly is one of the only newsletters I'm subscribed to. And so I do think a lot of us read it. And it's super useful. So anyway, small plug there that if people are new to it, it's worth subscribing to. And I don't just say that because you're a guest on the show. I'm not just being nice. It's really, really quite useful.

PETER: 00:12:46
Yep, no, I'm happy to get as many plugs in as possible. So yeah, that's a nice big tick.

CLAIRE: 00:12:51
So what caused you to create a newsletter business? Let's fast forward from your 14 or 15, and then you didn't go to college and become a lawyer. And then you were working at a new media company. You were building websites. And then there must have been something that happened between that and starting to write these developer newsletters and founding Cooper Press.

PETER: 00:13:16
Yes, so I rapidly got sick of having a full-time job, you know, quite amazing for someone that's still a teenager. But I kind of realized after about nine months of that, that that was much more than I ever wanted to go through. And I was done with it. So I began working for myself, kind of as a contractor, as a, you know, just someone who had digital skills. And, you know, if you got in touch with me, I could put them to use for you, and you gave me money. You know, it's kind of a thing now that, you know, so many people do. But My father had always been, well, not always been self-employed, but had been self-employed for at least 10 or 20 years by that point. So he was familiar with how it all worked. And so it didn't seem like a scary leap to me to do that type of thing because I had contacts and people hitting me up to do work. So I did that for some years. And eventually, you know, I got into blogging and was blogging about Ruby quite a lot, Ruby programming language and Ruby on Rails in particular. And I had a publisher called Apress reach out to me and say, oh, we've seen your blog. You know, would you be interested in writing a Rails book for us? And I said, no, I'm not interested in doing that. I said, I'd rather write a Ruby book for you. Just because I liked Rails, but I was much more interested in the language behind it. So I said, you know, there's a massive hole here for, you know, a book that could teach people Ruby from the ground up. You know, there's some fantastic reference books and things like that. But there wasn't much for like kind of in the style of the 1980s programming books. I was used to, you know, hundreds of pages and, you know, take you through every single tiny little baby step. So I said, let's build let's do that. And they were like, great, let's do it. So I started a blog to help promote the book. Now, back in sort of the noughties, as I call them. So this is like 2005, that kind of area. It was very common to launch a blog to promote something. So whereas nowadays you might launch a podcast or you might launch a newsletter, back then it was launch a blog. So I launched a blog and just started posting updates about the book and things about that were going on in the Ruby world. And eventually the book completely eclipsed any success that the book, sorry, the blog eclipsed the success that the book had and became basically the most popular Ruby blog out there. And that was basically my almost my full time business for, you know, a handful of years and, you know, did very, very well. This was the era when, you know, blogs like TechCrunch and so on were coming up and they were making, you know, tens of thousands of dollars just having tiny little 125 by 125 pixel ads on the side of the site. I did something very similar. I didn't make tens of thousands of dollars, but I did, you know, more than well enough to pay my way. And eventually I got to a point where I saw there was lots of things going on with email. There were companies like Groupon and LivingSocial and Help A Reporter Out. And there were all these types of email projects. And I just like thought, well, let's try and just turn this blog into a newsletter on the side. And again, like the blog took off as a side project, the newsletters took off as a project. And that was the very first newsletter. It was rubyweekly.com. That was in about 2010, but kind of 2011 as well. it spread over this period where it kind of got formalized into a company. But yeah, by 2011, we were very formally established. And that was our one newsletter at that time.

CLAIRE: 00:16:47
Okay. Now, what was the name of the Ruby blog?

PETER: 00:16:51
Ruby Inside. And I can never remember if it's actually still up or not, because it's literally on an old server that's just a HTML mirror image of what used to be there. In fact, yeah, it's still there, rubyinside.com, in fact, it's so old [It's still there.] it doesn't even have HTTPS and actually it looks like someone has put loads of spam over the top two posts so I need to sort that out but that shows you how often I go to this website which is like never so yeah

CLAIRE: 00:17:16
Yeah. It's kind of got the look of, you know, how you walk into an old house and it's full of old furniture and it doesn't look modern at all. This doesn't look modern. I'm not, I'm not insulting it. I'm just like, wow, I feel like I'm going back in time when I look at it.

PETER: 00:17:33
So yeah I mean I haven't posted to it for 11 years so it's probably running an antique WordPress I did kind of delete all of the parts of WordPress that are usually the most hackable parts but it looks like someone's managed to uh, figure it out slightly, so uh yeah I've at least I found a new project to work on now which is cleaning up the site.

CLAIRE: 00:17:53
Okay. And the book that you published, Beginning Ruby: From Novice to Professional, Aaron just posted it in the chat. [Oh, cool] Is that still relevant? Do you still recommend it to people? I'm imagining it has stood the test of time.

PETER: 00:18:11
Yes and no. So I have you know plenty of funny stories to do with the book as well but it's now on its fourth edition so I was involved in the first three and I had kind of a few run-ins with the publisher let's say, and a slightly testy relationship with them, and so when they came to me for the fourth edition I was like I just can't be bothered with this anymore and they said well that's fine but we reserve the right to get someone else in to you know basically write the fourth edition your name will still be on the cover but their name you know they will become the lead author of the book and I said that's totally fine like I think the book's run its course by this point you know it's written in a slightly old-fashioned style for now you know not many people buy 600 page programming books at the moment, and so I was happy to do that so the book is out there in a fourth edition but yeah, it's it's gone it's gone through some history but uh you know it's done well over its time.

CLAIRE: 00:19:11
Okay. I love hearing the backstories. I also have to tell you, maybe the reason you and I got along, I was supposed to be a lawyer too, and then I didn't go to law school. But it was the plan. Everybody knew I was going to be a patent lawyer, and I started working at Sun and just never looked back.

PETER: 00:19:34
So yeah it never happened for either of us and now I look back and I think I really wish I'd become a doctor so you know it's we have all these kind of like you know thoughts about what our career could have been and should have been but yeah here we are.

CLAIRE: 00:19:48
So you create Postgres Weekly. I'm looking at the very first issue, which is actually easy to find. Because if I go to postgresweekly.com and then I look at the latest edition, you have those previous and next links at the top left and the top right.

PETER: 00:20:07
Yeah.

CLAIRE: 00:20:08
But I noticed in the URL, you just have the edition or I don't know, do you call them editions or versions? I'm not sure. And so I just replaced the 603, which is the latest one with the number one. And I can see it from March 13th, 2013.

PETER: 00:20:27
Yes. Yep, it's super easy for all those AI bots to crawl. It's very easy to navigate around.

CLAIRE: 00:20:34
And that was, that was on purpose, I'm assuming. Well, you didn't do anything on purpose for AI bots back in 2013, but I'm assuming it was for the crawlers, right?

PETER: 00:20:44
Pretty much yeah I mean it doesn't hurt us whatsoever for you know anyone to crawl our stuff and to link to our stuff so yeah there's literally no downside to make you know making there's no point in us making it difficult.

CLAIRE: 00:20:58
See at the top, it says that Heroku's Craig Kerstiens was the chief curator in the beginning. [Yes.] But what's the story on how this first one happened? Were you already keen on Postgres yourself? Did you have the vision that Postgres was going to grow like gangbusters? Or did Craig come to you? Or how did it all happen?

PETER: 00:21:19
This is going to be extremely boring because I actually cannot remember, I don't remember how it came together but uh I am going to guess just from a bunch of contexts that I have that I believe Craig probably did approach me to do it, and the reason that I say that is because although I was reasonably pro Postgres at that time I wasn't an active day-to-day user of it and that did follow on, but in everything that I was using up until that point I, you know, had pretty much been using MySQL or whatever you know just to that point and I think part of that becomes you know comes from my Ruby background so the Ruby world was very very heavily into MySQL rather than Postgres, it's changed a little bit now, although 37signals I believe you know are still heavily invested in that uh rather than Postgres but uh yeah so just because of my background, I wasn't in the Postgres space at the time. So I'm going to predict that Craig did come to me. And especially because at the time he would have been working for Heroku when they were building out their big, you know, Postgres hosting thing. So, yeah.

CLAIRE: 00:22:34
Well obviously the space has grown the investments in Postgres have grown the number of hyperscaler cloud offerings has grown, like the Postgres landscape has developed since you started this newsletter. That's my perception. What's your perception as a newsletter publisher? Like, did it exceed expectations or has it not yet met your expectations?

PETER: 00:23:03
It is well the newsletter itself is actually still our smallest of all our newsletters But I'd say it punches above its weight in terms of, you know, impact and community and so on. But in terms of raw numbers, you know, it's like a tenth of the size of our JavaScript newsletter. So, you know, is the Postgres community 10 times smaller than the JavaScript one? It's kind of hard to say. Like, I haven't, you know, got good numbers on that. But, yeah, it's worked out totally fine. And, you know, it has grown alongside the industry. And we've got some great readers who reach out to me and people tell me, you know, mention the newsletter in various places and so on. So, yeah, it could always be doing a lot better. Let's put it that way. But, you know, there's obviously part of the thing about plugging it on shows like this as well. So, yeah, everyone sign up and it will do even better.

CLAIRE: 00:23:57
So how do you find the Postgres links that you include each week? And what's your, I'm gonna ask two questions at the same time, which is probably a really bad best practice, but what's your guiding principle for deciding what to include?

PETER: 00:24:13
So, you asked multiple questions, I could take the political way of just saying what I want to say and just ignore the questions but I'll try and go through them in order, so you started by asking like you know where do I find things and obviously you know initially it was Craig doing it so he had all these contacts and you know he was on Twitter all the time and you know kind of just had the inside track on everything uh over time as I began to you know take over more of that role part of it came through the fact that you know I was actually beginning to use Postgres increasingly more so that really helped me out and so i was beginning to engage with people on Twitter that you know were well known in the space so you know they would tweet about something I'd be oh there's some news I can include that, you know that's still relevant, probably less relevant now uh than it was up until a few years ago but uh you know that's still one way of finding things but then actually Postgres is super well represented on sites like Hacker News for example I mean that's just one of many. There's other ones you know lobste.rs and Reddit and there's a whole bunch of you know different kind of community aggregators and planet sites and things like that but just picking on Hacker News, there are just so many items that get added to Hacker News that are Postgres related. It just seems to be a natural place that people in this space gravitate towards. Even if they're not commenting on there, even if they don't like Hacker News, they submit their posts to it or they're like, oh, can you vote up my post or this type of thing. Please don't send me those types of messages. So yeah, I find a lot of stuff on there. And part of it is because I've been on Hacker News actually since the very very early days and I also am glued to the newest page on there so for people that are not familiar with Hacker News you've got the front page which has like the top 30 items as of right now but then there's also a page called newest that you can click on at the top and it's like the raw firehose of items that have been submitted to the site. Now you know I would guesstimate that perhaps only about one percent, probably less actually, of the items added to Hacker News actually get upvoted and put onto the front page but that doesn't mean the other 99 are just you know trash uh there's a lot of good stuff in there so people will often post all their blog posts and you know corporate news and things like that and if you stay glued to that newest page or you subscribe to it through RSS like I do, you might end up with like a thousand things to look at each day, but you can quickly skim through and be like, oh yeah, there's something interesting. So that is a massive source of kind of items from just disparate places for me. But then you can add to that, you know, we've built up obviously over time, we've linked to hundreds of different blogs. And each time we find a new blog that's talking about Postgres, it gets added to my feed reader. So I'm then following it forever more. And so I find stuff that way. That probably makes up about, you know, so let's say like Hacker News and Twitter and Reddit and so on, make up about a third. The feeds I subscribe to probably make up another third. And then the other third roughly is people that then send us emails. So you can reply to any issue. And, you know, we do get this every single week. People reply and say, oh, I've just posted this or I've just released this extension. Can you check it out? And then, you know, we don't post everything that we're sent, but quite a lot of it makes it in. So we kind of have this nice split of different ways that we find things, you know, roughly by thirds, I would say.

CLAIRE: 00:27:48
Okay, so I've got to ask you, what do you use for a feed reader?

PETER: 00:27:53
I am currently using two different ones for very bizarre, complex reasons, but I use one called Feedbin which is at feedbin.com which just as it happens happens to be implemented on top of Ruby on Rails so I like to support that and I think it's open source as well so you can actually download your own copy and run your own version of it. And then I'm also using a system called feeder which is at feeder.co and the reason for that is that it has this kind of cool teams interface so the other people that I work with uh can look at the same feeds I'm looking at so if there's any kind of like you know blind spots or whatever they can catch things that you know I catch or they can catch things before I see them. So that quite often happens as well. So yeah, I use two different feed readers.

CLAIRE: 00:28:36
When you said it had a cool teams interface, I was thinking, huh, Microsoft Teams is integrated with that. [No, not Microsoft Teams. No.] I didn't know. That's not what you meant. Got it. Okay. Yeah, I've not found a good feed reader ever since Google shut down my favorite one, which, as you know, was quite a number of years ago.

PETER: 00:28:58
Well, I would say that Feedbin, feedbin.com, is definitely the closest to the reader experience that I've found. [Okay.] And it has all the keyboard shortcuts that work in exactly the same way. And, you know, j, k, and l, all that sort of stuff jumping up and down.

CLAIRE: 00:29:12
Okay, so like vi based or something okay vi based then j, k, and l, I don't know okay.

PETER: 00:29:13
It's my favorite. Yeah. Sorry, say that again? Yes. Yeah, it has the vi style, you know, keyboard shortcuts and things, which actually I think Feeder does as well. But I'm not quite so keen on Feeder but it does have some extra features that we need in the company but yeah, feedbin.com is worth looking at especially as it's open source and it might even use Postgres for the database as well. I'm not quite sure, I haven't looked into that, but it wouldn't surprise me.

CLAIRE: 00:29:40
Okay. Yeah, I wonder if anyone on the chat knows the answer to that question. So we'll see. We'll see if anybody chimes in. I think one of the things that I've talked about with people about your Postgres Weekly newsletter in particular is how it, on the one hand, it's very fair in the sense that it feels like you share links from all the different companies, all the different countries. Like you really do canvas a lot of sources. And I'm always finding things that were not in my feeds, right? That I was not going to catch on my own on Twitter or LinkedIn or Bluesky or Mastodon or all the places I pay attention to. So I love that kind of breadth of perspective. And yet, in addition to being fair, it's also opinionated. Like when I read your blurbs, you don't just factually pass on an article. This is an article about logical replication published by so-and-so from this company. Like it's not like that. It's not boring. [Hmm.] Your blurbs have opinions about what is being presented in that article or link or video or whatever it is. So is that like, do you write them all? And is it always your opinion? Or do you have a style guide and have a team of people writing those blurbs? And therefore you've each tried to adopt a similar tone of voice? Like, how does it how does it happen?

PETER: 00:31:12
In the in the case of Postgres it is 99 percent me. I do work with someone who is in the company and you know works on the editorial side and he will occasionally see an item that's really really like you know straightforward and he'll write it up and I'll often tweak it or sometimes I'll post if it's you know spot on. But in terms of most of the opinionated and technical stuff yeah with Postgres is 99% me. And, you know, just to kind of give some background as to why I have that style, a lot of it comes from being English. And you might think, well, how does that work? And the reason is, you know, I grew up with the BBC. And for all of its kind of warts and problems, especially nowadays, just over the years, the BBC kind of has this reputation of kind of being a big tent. So it can't just say, you know, we're going to go along with whatever the politics of the day are. You know, whoever's in charge, like we'll just, you know, cuddle up to them. They're always a big tent.

CLAIRE: 00:32:12
Being a big what? Being a big, what was the word he used? A big tent.

PETER: 00:32:17
Tent. Yeah, I guess I don't know if that's a word that's an idiom that's used anywhere else. [Okay, got it.] But in England, it's kind of means to adopt, you know, a broad set of viewpoints. So you don't want to leave people outside the tent, as it were. You want everyone to be within the tent and then, yeah, you then you get to have the largest coverage. So there's no like companies, you know, in the Postgres space that I want to be outside that tent. I want everyone to be inside the tent. And then we write about everyone, you know, good or bad, you know, however we see it at the time. And so the BBC does this culturally. You know, they actually have as part of their conditions for being able to collect TV license in the UK, which is basically a tax we pay to watch the TV. As ridiculous as that sounds, part of the condition behind that is that they provide programming for underrepresented communities, whether that's ethnic communities, religious communities, just all different types of thing of that nature, because they have to basically be there for the entire United Kingdom. And so in a very similar way, some of that, I picked up some of those kind of ideas for osmosis to use in my newsletters where, you know, maybe I'm not particularly interested in, let's say, I don't know what, you know, a particular company is doing. Let's say I'm not interested. Let's just say, for example, that I was really anti-Microsoft, which I'm not. But it's, you know, an easy thing to pick on, as I remember all the kind of the background of, you know, people into Linux didn't like Microsoft, those old kind of wars people used to have. you know let's say I was in that kind of camp, well I would still believe that I still need to you know faithfully represent everything that Microsoft's doing if it's doing important work in the Postgres space which it does. So I'm not tribal about anything I kind of look at all the different things and even though I have opinions about oh this company's doing something better than this company I'm not just going to say well we're going to kind of ignore this whole section of you know the ecosystem just because I don't like it. And so that's why it's a very broad tent and you know I try and keep everything included.

CLAIRE: 00:34:23
I think maybe that's why people are so supportive of the newsletter and will recommend it to others and will recommend it to new hires, right? Because you're not tribal and because it seems fair. And I've never heard this idiom of the big tent before, but I'm going to run with it. I love it. What a great image that brings to mind. Cool. And just for the record, let me repeat that I would like to believe that Microsoft is doing good things for Postgres. And hopefully you believe that as well. And you just decided to nitpick on us because I'm here and I get it.

PETER: 00:35:00
I think it's just because I was picking up on just something very historical. So because I was, you know, around 20 years ago, you know, kind of in these spaces still and, you know, the Linux ecosystem and everything, there was obviously a period of time where, you know, Microsoft by certain people was seen as the enemy company. all this type of nonsense, which has actually been kind of like withdrawn to a large extent. We don't see that hostility now. I mean, they obviously exist in some camps, but broadly, Microsoft has made so many great steps forward into communities like the Postgres community and releasing things like Visual Studio Code. And it's just so many different things, even in .NET, bringing in the Mono project and all these types of things and WSL. And I guess it's is completely countless the number of things that Microsoft does now but it's a totally different era I mean to what it was 20 years ago.

CLAIRE: 00:35:51
It is. Well, you know, I started in Citus. That's, well, I didn't start my career in Citus, but I started my Postgres work at Citus Data. And then about two years after joining, we were acquired by Microsoft. And it was really interesting because friends of mine, former colleagues and teammates who were like, especially the Sun Microsystems friends, they were all over the industry at this point, right? Because most of them were no longer at Sun. And I would get these notes and these DMs and these calls saying, you know, Claire, if this had happened to you 5, 10 years ago, I would be offering to help you find a new job and commiserating with you. But you know what? Congratulations. This is pretty cool. And I heard that same theme in multiple conversations and it was kind of cool. It really was kind of cool. And yeah, thank goodness. It's nice. If you're going to get acquired by another company, you want to like that culture, right? And like that team that you're joining. So anyway, yeah.

PETER: 00:36:51
Yeah, I was acquaintances with the GitHub founders just because they came from the Ruby on Rails world because GitHub was originally developed entirely in Rails. And so I used to hang out with them on IRC and chat rooms and stuff like that. And, you know, I know they had very similar things happen there because obviously, you know, GitHub was acquired by Microsoft and, you know, it was kind of, that was at the period, I think, where the attitudes were just starting to shift. But people were still concerned like, oh, Microsoft's going to take all that source code and all this type of thing. So, you know, but it has turned out to be an extremely good steward of GitHub. So, yeah, I think that helped a lot.

CLAIRE: 00:37:23
Yeah. Well, Git and GitHub are near and dear to me because I started my career at Sun in developer tools. So I worked on a predecessor to a predecessor to a predecessor of Git and GitHub. So anyway.

PETER: 00:37:40
Yeah, all these worlds are colliding. It's very nice.

CLAIRE: 00:37:44
Okay, so I'm curious. Obviously, not everybody who reads a book is curious about the act of writing that book and what the author went through. But I am, and I get to ask the questions. So I'm curious about the process for putting together a newsletter. And I fully respect that there might be aspects about the process that are like your secret sauce that you don't want to tell us about because you don't want anyone to be able to replicate what you do, the magic in your newsletters. But seriously, tell us what you can about creating a new edition of Postgres Weekly besides the sources that you already covered.

PETER: 00:38:25
So we have an internal system that is kind of like the core of all of this in terms of us collecting together links and writing things up. And we have a few external curators, not for Postgres, but on some of our other titles. So we can all use this system that we have internally, which was built wow sort of about 12 years ago and it's it's barely changed over that time it's essentially you know a bookmarking system but kind of like turned up to 11 that allows numerous people to use it and you know allows us to collaborate on you know writing things up and stuff like that, but it also acts as kind of like a queue of things for us to include so there you know there are basically tens of thousands of links in this thing, and so I can when we link to something in the past and I can see information about how often we link to something. There's always different things I can look up. So I can really get a good feel for the rhythm of including certain things or certain topics in newsletters. So that always really, really helps. But then it turns on a weekly basis, what I do is, I mean, I'm addicted to Hacker News and Reddit and all these different sites anyway. I'm always in my email inbox, even though I take forever to reply to emails, I am always reading my emails. And so things are getting added into this essentially queue for future is all the time. So I can literally go on there right now. You know, I'm going to be doing Postgres Weekly tomorrow. And there will already be, you know, 20, 30 different things that have been added over the past week that are, you know, of some interest.

CLAIRE: 00:40:03
I sent you a few.

PETER: 00:40:03
And then there's hundreds of things. So, yeah, well, exactly. You sent me some things, but there's hundreds of things as well from previous weeks that just never made it in. So sometimes, you know, if we're really desperate for something or we want to try to build up a connection between two stories, sometimes I'll go back and look for things that join the dots together. So that's kind of probably like the most journalistic part of the job. I mean, it's not a journalistic job whatsoever, but that's the part where I'm trying to flesh together a story or a chain of items that kind of relate to each other. Sometimes we have this system that allows me to do that. But then basically the way it really works is that when we get to the day of publication is I will then hit up all my different feeds I'm looking at. I'll go and do last minute checks of, you know, like Postgres website and some of the prominent people in the space just to make sure nothing has just dropped at the last minute, which it quite often does. And yeah, I get those things into the system and written up. And then what happens in that system is there's kind of a bunch of stages that items go through. So we can just push raw links into it and say, you know, these need to be reviewed or this is like something we definitely want to do. Or this is something that's kind of borderline or it's a submission or whatever. I will literally kind of take a prioritized look at that. So if there's anything marked as to do, this is really urgent. Then obviously I will tackle that first. And then we work our way down until essentially we have, you know, the right number of items to work with the issue. And we also have a real time collaborative editor for the newsletter. So we have a slightly odd proprietary system because. Back when we began doing this, we used MailChimp and MailChimp's editor was really, really slow. And it was really designed for marketing emails. It wasn't designed for the type of thing that we do. So because I had an API, I built numerous different editors over the years that allowed me to write issues in our own kind of XML style language, which is actually a little bit like MJML. If anyone's familiar with MJML, it's like a format for HTML email that then gets compiled down to the HTML and plain text and CSS in a way that works across different email clients. Well, before that existed, you know, I built our own system, which we still use. It allows us to mix things like Markdown, HTML. It allows us to put things like mastheads in, you know, with a special masthead tag. And we have all these different tags that we use for things. And then it shows on, so on the left-hand side, you've got the code. A little bit like a modern Markdown editor. You can on the left-hand side, you've got the code. On the right-hand side, you've got the live preview of the issue. So I can go to the issue right now for tomorrow, and it will be pretty much empty except the sponsors. But I can then gradually over time just add in the links. And then eventually we click a button and then it uploads it to the website. It puts it into our email system. And I can do things like start sending tests. And, you know, eventually we can move on to actually sending the final thing once it's passed forward to reviews. But the process is, you know, it's not a complex process. It's just lots of different proprietary bits and pieces that we put together over the years to make it really streamlined, so that we can focus on our job being the curation and reading things. And, you know, often we test things that we link to, you know, if it's a new project and we're like, well, this doesn't look quite right, we'll give it a test to make sure it all works. So we can focus on that part of the job, which is the interesting part, not the, oh, how am I going to make this render in the, you know, different email clients and things. That's the part that we've sorted out and we outsourced all these other systems.

CLAIRE: 00:43:44
But the rendering and not getting caught up in spam filters and making sure that you know the the emails get across like that's the part that so many people can screw up on right even if they beautifully written, really interesting newsletter. If you're not paying attention to those technical bits under the covers, people won't necessarily get your mail or when they do, it won't look right. It won't be so readable.

PETER: 00:44:13
One advantage that I've got is that I've been doing this for a long time. So I know where to go, you know, when there's bugs in the email sending process, I know roughly where to look. Like if you're a developer, then, you know, you know where to look for bugs once you're, you know, sufficiently experienced. We also use some other systems. So litmus.com is one that we use, which is kind of expensive now. It used to be quite cheap, but you can send your email to them and they will render it in 50 different, it's about 50 different email clients and you can visually see it on the screen and it will let you see what your email is going to look like. So we use that when we introduce new units to an issue just to make sure that everything renders fine. But we do have some funny stories over the years of things that have broken email systems, things that you wouldn't expect. So the one story that I often bring up is we had an issue of, I can't remember what newsletter it was, but it was constantly going to the spam in Gmail. And every single test we ran, spam, spam, spam, spam, spam. Now what we did was we actually have this system that allows us to kind of run almost like a binary search over our emails. So it will send out emails with like half the items in and then the other half of the items. And then you pick the one that worked and then it will put in half of the broken items and it will basically work its way down to working out what item was it that broke the email. And we found this website that broke the email and it was just an innocent developer blog. But the thing about their domain name is it began with 0x, which anyone who's been developing for some time or uses C, for example, will know that if you use 0x, you use that to notate hexadecimal numbers in a language like C. And I was like, well, why would that trigger it? Well, it turned out that you can also create URLs that are called long IP URLs. Now, you almost never see these. So if you have a URL like https://1.2.3.4, it will go to the IP address 1.2.3.4 You can, however, write IP addresses in what's called a long IP form, which is basically if you take the 32-bit number, this is getting very technical now, [I love it.] If you take the 32-bit number that 1.2.3.4 represents, you can then turn that into the 32-bit integer, put 0x in front of it, notate it in hexadecimal, give it to the web browser, and it will still go to that IP address. That, I believe, still works in web browsers. I'm not sure if it does, but it did 10 years ago. Now, the problem is, is that Gmail added in its infinite wisdom a filter so that if anyone sent an email that involved a long IP URL, it would immediately put into spam because it was being used very heavily by, you know, people trying to scam people and, you know, get around, you know, all these types of things. But they got the filter wrong. So they just looked at any URL that began 0x and they didn't notice that the fact this was like 0x something, something, something dot-com or whatever the TLD was and they shouldn't have done that. So I reported the bug, they eventually fixed it, and happy days, but yeah for some period of time people that had you know links going to 0x anything was completely broken in Gmail so that's just one example of some silly tiny thing that made a massive difference and you know we've run into so many of those over the years but that's just one I mean and you can see how deep that one got.

CLAIRE: 00:47:44
Right, right. So a lot of details that have to be sweated in order to make this work. [Mmm.] Speaking of, that's a memory about the technical implementation of the newsletter and staying out of the spam filters, are there any other memories that are worth kind of going down [Yeah.] the lane and talking about about I don't know shifts you've seen in the Postgres ecosystem or particular articles or news that came out, I know we've had a lot of big news in the last couple of weeks with two big acquisitions of Postgres startups, right? [Yeah.] Both Neon and I don't know if you'd call Crunchy Data a startup, but, uh, you know, both of them acquired by much bigger players. But over the last 12 years, does anything stand out to you? Because you've seen the patterns as they've evolved.

PETER: 00:48:37
Yeah I can't pick individual blog posts because I mean literally you think about the amount that I read, so many things just go through my brain out onto the page and then they're gone. So there's very few blog posts that kind of like really stick with me for years and years, but the broad stories stick with me a lot better. So, you know, in terms of overall trends and things like that, you know, one massive difference that we've seen just from 12 years ago to now is that we do write about the business side of things a lot more. Like, I mean, 12 years ago, Postgres was seen in a very different way from a business perspective. You know, if you go back 12 years ago, that's actually when Amazon RDS first introduced its Postgres support was in 2013, just after we launched. And, you know, people weren't really that heavily talking about it at the time. You know, it was news, but it wasn't news in the way that it would be if, you know, Amazon launched a new Postgres thing now like they have with DSQL. You know, corporate news in the Postgres space just wasn't that big a deal back then, whereas now it's a massive deal, whether it's Supabase, you know, getting funding or even things like this week, Timescale is, you know, renamed to TigerData. Like, that's a story now, whereas it wouldn't have been 12 years ago. So I've seen that occur, the growing role of companies in the ecosystem. But, you know, there are also things that haven't changed over the years that when I was kind of doing my research for this, you know, I realized just how important extensions are to Postgres's success, because the very first thing that we linked to in Postgres Weekly all those years back was an extension that allowed you to do stuff with the HyperLogLog algorithm in Postgres, which I don't think the extension is relevant now. But the fact is we've covered extensions all the way through and it hasn't changed. It's still very, very important and still an ongoing debate and discussion. And it's also a massive difference to MySQL and lots of other databases where there isn't that extension ecosystem. It's been so important to the Postgres ecosystem, something that I really enjoy seeing go on. So that hasn't changed at all. Then, you know the things that stand out to me are just when new things are introduced that kind of kick Postgres up to the next level so as I said you know we started in the Postgres 9.2 era and 9.2 was a notable Postgres release just because it introduced JSON so Postgres went from this kind of idea of it just being a you know a row store essentially to being a multi-model store you know you could do some, you know, document kind of basic document stuff with it. And now that's just moved on and on and on over the years. We've had JSONB, we've had, you know, things like pgvector come in. There's other, like, there's the Apache product that adds graph support to Postgres, there's PostGIS. And things have just grown and grown and grown. And now you can look at Postgres as being a true kind of multi-model database at some level, thanks to its, you know, extension support and thanks to all of these different things being added in at the base level. And so it's really been fun being on that journey and seeing Postgres evolve, but having like such fertile ground to grow on that a lot of other database systems just haven't had. So that's been really exciting to see all of that growth.

CLAIRE: 00:52:10
Sometimes when I give talks that focus on the history of Postgres, I will take a screenshot of the original paper that was published by Michael Stonebraker and Lawrence Rowe called The Design of Postgres. And I just love the fact that, like, number two, I'm looking at it in front of me so I can read it exactly. Number two on that design doc is provide user extendability for data types, operators, and access methods. And it just wows me that that was part of the design from the very, very beginning. Now, obviously, some aspects of extension support didn't come in until later, but it was the intention from the beginning. So I'd be happy.

PETER: 00:52:50
Yeah, whereas that wasn't the case with MySQL. Now, I just want to say, I don't know whether you say My "S-Q-L" or My "Sequel". I tend to say MySQL, but obviously you can say it either way. But I mean, that was one difference with that is that even though MySQL can be extended, it's harder, a little bit riskier, less kind of like rewarding, essentially, than in Postgres, where it's so simple. Like Postgres has, you know, much more kind of API stability, and it has more hooks to like, you know, hook into it. Like, you know, this is one thing that we've covered over the years as well, is people releasing toolkits and libraries for building Postgres extensions with languages, you know, like Rust and Zig, and, you know, even with things like JavaScript and Perl and I've not seen it in Ruby I must admit but all these different languages you can build things in and it's actually not that difficult like almost anyone that is a software developer if they really wanted to right now they could sit down and they could figure out how to build a Postgres extension and distribute it and get people to install it in a way that you just can't do very easily with other database systems and I think that has had a massive effect on you know, Postgres's current status.

CLAIRE: 00:54:02
I want to circle back for one second to a thread from earlier when you were talking about your feed readers. And Rob Treat on the chat did look it up. And I think, well, he looked it up and I had it in front of me and now I don't. Oh yeah, he says it looks like it uses a combination of Postgres, Redis and ElastiCache. So, yeah.

PETER: 00:54:22
Oh, there we go. That's good news.

CLAIRE: 00:54:24
So it has a nice little Postgres connection.

PETER: 00:54:26
Yeah, and I do love Redis as well. So that's another favorite database of mine. Although I tend to call it a data structure store whenever I link to it rather than a database. But yeah, it's all semantics. [Because why?] Just because you are working at a much more low level, it's more granular, you're dealing with just, I'm storing a string or I'm working with a hash or I'm working with a list of something. Like it's, it feels like another dimension down from what something like Postgres or even, you know, SQLite or whatever, you know, SQL server is dealing with. So I tend to think it is a data structure server, but you could argue that, you know, things like Postgres are also storing data structures. But I just need to, I feel like to make the distinction, it kind of belongs in the same family as things like Memcached and, you know, Berkeley database and things like that. It's a different thing to me.

CLAIRE: 00:55:23
Okay. When you talked about your process a few minutes ago, both at the technical level and at the curation level and how your team collaborates and everything, one of the things that's new, like when I first started reading Postgres Weekly in 2017, I don't think you had this feature yet, but you now include a beautiful graphic banner at the top of the newsletters, at least the ones I receive. Maybe if you send a text-only version, it does not have the graphic. And it's really interesting. I can always tell there's – I don't feel like you just threw it up there. It feels like there's some thought and some intentionality to it, although maybe I'm giving you more credit. I don't know. So can you talk to us about how those graphics came to be and why and are they working for you? And is it a delight to create them or a bit of a hassle?

PETER: 00:56:17
So, when I was running the Ruby Inside blog, one of my kind of things I made sure to always do was include an image in every single post just because it seemed on blogs that that kind of helped your engagement especially with feed readers at the time you know people see an image and they'd be more likely to come through and read the whole post I don't know why but there's just like some psychological angle to it I guess and I kind of forgot about that when I was starting doing email I was just like let's keep it really really plain, really really simple, I don't want images in my emails I just want to keep it really really plain after some years I kind of got addicted to another newsletter called Podnews which is a podnews.net and that's a kind of a daily email about the podcasting industry which I definitely recommend subscribing to it's very very good. [Is it still relevant?] It's it's still relevant yeah it's you know it's interesting, it's interesting to people that are interested in podcasts as a business I think probably more than you know just average listeners.

CLAIRE: 00:57:13
Okay. So they're probably not talking about Talking Postgres a lot, at least not yet.

PETER: 00:57:18
No although you know if it was if they might mention it that type of thing like you know obviously oh Microsoft have released this or whatever but uh yeah it's more about the industry but it was one of the few newsletters that I kind of took direct inspiration from in having some new features, and one of the things that they always did and they still do is that they have a image at the top of their email and it's about you know whatever the top item in the newsletter is and I was like I really like the way that works I like the way they've implemented it so I was like I'm gonna just steal the idea from them, and I kind of knew the guy that runs it and everything he was very cool, so you know we began if the top item in a newsletter had an obvious screenshot or something that I could just take uh you know I'd take a screenshot and we'd you know put it in and uh you know people seem to like it people click on it and great uh over time obviously you know we started to get a habit of having these in but there wasn't always a screenshot to include so I was like well what can we do for these and I kind of had a weird background from many many years ago of being very briefly in involved in graphic design and so I had you know I've always been addicted to Photoshop, I use Photoshop nearly every day for doing random things or you know coming up with joke images or memes or whatever and all this type of thing and so I was well, you know, can I use some of my skills there to produce these images? And so we described to, I can't remember who it was now, like Shutterstock or someone like that, one of the stock libraries. And I would basically put together like composite images of things. So one of the worst ones that I remember in Postgres is that we had an article about auto vacuuming. And it was something about auto vacuuming getting stuck. And just I came up with this horrible montage of like a robot vacuum cleaner getting stuck under a table or something. And it's a terrible, terrible image. Very, very cheesy. But it made me laugh. And I was like, I actually quite enjoy making these silly images. And so it just went from there over time. You know, we've introduced all sorts of various different tools into it. You know, I will use AI to produce elements of images and stuff like that. But at the end of the day, I do a lot of composite compositing work to bring things together. You know, we don't just generate an image somewhere and just slap it in. We're like, you know, I need to make sure there's a story kind of told in the image. And so I'll bring together icons and, you know, bits and pieces and screenshots and put them together into something that kind of vaguely makes sense. So it's actually one of the most fun part of my jobs. I actually quite like doing the image for things and it makes a change from, you know, doing just purely editorial stuff. So yeah, I enjoy that.

CLAIRE: 00:59:56
So maybe your writing and the blurbs that you write are fresher because you've had that distraction using another part of your brain, I don't know.

PETER: 01:00:06
Yeah it's kind of a multimedia aspect to it I guess but uh yeah in terms of actually drawing things from scratch I have absolutely zero talent to do that so you know putting together montages and things like that and kind of newsy kind of pictures is is definitely more my style and actually I need to shout out someone in that regard because there's someone that's put together like a public domain uh group of images all to do with like gophers so whenever we do stuff like the Go programming language uh we use that public domain stuff quite a lot and we shout them out in the newsletter from time to time so I just wanted to mention it, just because you know there are groups of people out there illustrators and so on that produce images that we do use and we credit them and everything. But it's very important still to have that human aspect in there.

CLAIRE: 01:00:54
So, on the chat Aaron seems to have found an edition of Postgres Weekly from August 2022 and I don't know if this is the right one but it's a very sad looking blue elephant [Okay.] looking at a robot vacuum cleaner that's on fire. It's one of those circular ones that kind of rolls around on the floor by itself. Is that what you were thinking of? [All right.] Or is it different?

PETER: 01:01:17
Hang on, let me just quickly look at this link. No, it's not that one.

CLAIRE: 01:01:20
[Okay.]

PETER: 01:01:21
But no, there is one that is very obviously Photoshop where I've put this table and a vacuum cleaner. It's just absolutely terrible. No, it's not that one. That is a very related one. But yeah.

CLAIRE: 01:01:30
Well, now I want to see. We're going to go hunting until we can find the one that's absolutely terrible.

PETER: 01:01:35
Yeah, I've shared it on our work Slack numerous times because it's like the nadir of my image creation efforts.

CLAIRE: 01:01:42
I really enjoy the graphics. I don't know why. They pull me in and they kind of anchor the post as well and they tell you what the top article is about without even having to read a word which is kind of cool.

PETER: 01:01:56
Yeah, it can be good.

CLAIRE: 01:01:56
Okay so You don't just do this for free It's a business [Yeah.] And I'm assuming the bulk of the income is from advertising. [Yes.] Just full disclosure, my team at Microsoft and prior to this, my team at Citus Data, we have advertised with you. I wouldn't say always in the last seven or eight years because, you know, budgets come and go and projects come and go. But like most recently, we have placed some POSETTE ads to make sure that people knew about the POSETTE: An Event for Postgres virtual conference

PETER: 01:02:32
[Yeah.]

CLAIRE: 01:02:33
that we organize and trying to get people to attend or submit talks to the CFP or, you know, now, now that the videos are all published, to watch them. So, and we're not the only ones, like lots of Postgres companies spend dollars advertising in Postgres Weekly. And I'm sure the same is true across the other, across the other newsletters also. Has building up that, the business side, of your newsletter business been surprising in any way?

PETER: 01:03:05
No, I mean, it's basically just fallen into our lap in some sense. You know, when I began doing the newsletters, it wasn't like the current business model wasn't what I planned to do whatsoever. So my original plan was I was going to have the Ruby newsletter and I was going to basically sell tickets to an online training course I was running at the time each month, which was an advanced kind of Ruby, like online webinar type thing. You know, I've worked for various Ruby concepts with people and people would pay, you know, three, four hundred dollars. And I would get, you know, like 20, 30 people to join on that. And I'd help them all out and all that type of thing. You know, we had a forum and stuff like that. And so, you know, that was good money. If you run that once a month, then, you know, you have a full time income. And so that's what I did in the early days. But then I very quickly realized that I'm just not my brain just isn't in teaching. I'm not, you know, really good at the real time thing. So I had these people originally reaching out saying, oh, we've got this thing we want to sponsor the newsletter. And it's like, well, we don't really do that. But then I realized, well, let's just do it temporarily and then we'll see what happens from there. Well, whenever I say let's just do something temporarily, it tends to become permanent. I'm still doing it 10 years later. So that is very much the case here. It's another one of these things where it was a side idea and just because there was the demand, it carried on. And that's literally how we got to where we are now. And so over time, you know, I've hired people because I got to a point where we just had too much inbound business. So I had to hire someone to take charge of the whole sales side of things. And she is still with us today. So she's actually just celebrated 10 years with us. So, yeah, it's it's worked out quite well.

CLAIRE: 01:04:51
Well, and the ads in Postgres Weekly, at least, are tasteful. Does that make sense? I don't know how much thought you put into that. How much thought you put into the placements or whatever the guidelines are of what you accept or what you don't accept. But they seem, I don't know, they seem to fit the tone that the newsletter has of being useful.

PETER: 01:05:17
Yeah, we try to make sure that's the case. I mean, partly because that performs better. So if there is some kind of relevance or even if it's like content, you know, we do have sponsors that will literally link to a blog post, which can sometimes get a little bit weird because it might have been a blog post we were going to feature anyway. And so I try to maintain a really good divide between sales editorial. I try not to get involved in sales because I don't want to be unduly influenced by it and say, oh, this company is great because, you know, they sponsor us. I try to avoid even having that potential by staying out the sales side of things. But sometimes it does have an influence on us because it's like, well, actually, I was going to feature this post anyway, and they're already paying us to promote it. So I can't have it as the feature and the sponsor. So sometimes it actually acts to a detriment. [Yeah.] But most times, you know, like the people I'm working with have a really good nose for, you know, what is something that's legit and kind of earnest and interesting. It's not always perfect, but it's usually pretty good. We don't do outbound. And so that's quite important because outbound can really lead to low quality kind of sales.

CLAIRE: 01:06:21
What do you mean by that?

PETER: 01:06:21
You know, what I mean is like most companies that are in media, they will do outbound sales, which is where you look at companies that you might want to have or that you think have got lots of money. And you reach out to their marketing departments and say, oh, we've got X thousand subscribers. You know, you could come in our newsletter and blah, blah, blah, like do that kind of sales pitch, which we don't do. Like we, I think we've done it like once or twice for some job ads, like 10 years ago or something. But like, it's not something that is a part of the business at this point. All of our sales are inbound. It's where companies have seen us. They've seen the newsletter. They know what it is. And quite often it's because employees at the company, so like engineers have seen, you know, been subscribed to Postgres Weekly. They get asked by the marketing department of the company, you know, what's somewhere that's relevant to what we're doing. And they'll say, oh, well, here's this newsletter that I've seen ads in it before. Like, you know, contact them. And that's how we get the contact. So people that work at companies are actually probably our best sales agents in a way because they recommend us to their marketing departments and then it goes from there. So that is our most common way of getting customers.

CLAIRE: 01:07:31
And then there's a question from Kaiting on the chat about how your business team decides who gets to sponsor. Do you ever have people competing for the same newsletters? Because it's not overflowing with ads. Presumably you have a limit. We're only going to have this many ads or these many sponsors in a given newsletter.

PETER: 01:07:53
Yeah, I mean, you know, we have different ratios that I'd like to maintain. You know, the fewer, the better, but obviously to a point, because obviously we're a business. but you know I try to make sure that you know we don't just have like ad, ad, ad, ad, ad, ad, like you know it's not like a fashion magazine where it's just you know 20 pages of ads before you even get to the you know the content. So you know I've kind of just got this basic feel of what if I was a reader what I would tolerate and deal with but then obviously the other aspect of it that kind of goes to Kaiting's question which you know is how do we get decide who gets to sponsor you know it's not just like if anyone turns up and anyone has money then you're in uh you know our team usually is good at sniffing out things that are relevant I have kind of directives I've given them things I'm not interested in covering whatsoever or having involved with us. I'll give you one example of that which is, it's less important now but it was quite big a few years back which was companies that were very heavily into cryptocurrency and it's not because I'm against cryptocurrency per se it's just that when they advertise it tends to be quite overly commercial and overly I don't know this is the right word but kind of exploitative the thought the weight the things that they're trying to promote like it's not there's no kind of like organic edge to it whatsoever so they're the types of companies that we're not interested in and like NFTs things like that just kind of turn me off, but then there's companies you know that really make sense like you know when it's someone that's providing Postgres managed services or something like that then obviously for Postgres Weekly it's perfect fit, and if they can link to content or libraries that they produced or you know things of that nature like, you know, Timescale does with like their AI libraries and things like that and you know [You mean TigerData, right?] yeah oh sorry TigerData yeah they've renamed or you know or Microsoft with POSETTE, when it's things that you know that are going to get an editorial mention in some way even though it's not the ideal way that the sponsor would want that's a good fit because then it allows them to tell their story in their way if you know they want to add to what we're saying so, you know it works out quite well in those situations and yeah my team seems to really you know get it and they understand what I'm aiming for and yeah we don't tend to we you know we occasionally have the odd weird sponsor and I'm like why did you sign them up and you know we figure it all out in the end it gets figured out we talk to them and see what the deal is and what they're going to run and we'll sometimes help them finesse and hone what they're going to run but nine times out of ten we don't have any problems at all but you know it's uh it's worked out pretty well.

CLAIRE: 01:10:29
Okay, so now I have a couple random questions for you before we wrap. And the first is, I don't know if you ever listened to the episode with David Rowley. It was on Talking Postgres last August. And it was one of the canonical How I got started as a developer (& in Postgres) themes. That was the title of his episode. And he started in a really unusual way. It involved a cheese factory and a forklift. And for some reason then, a whole bunch of the guests brought a cheese story to the podcast. and somehow cheese was involved in their journey into Postgres or their work on Postgres. So I just have to ask you, do you have a cheese story to share with us?

PETER: 01:11:12
No, I mean I absolutely hated cheese for probably the first 20-25 years of my life I couldn't stand cheese. I actually do like it now but it's pretty much something I only eat at Christmas so I associate it with Christmas and holidays, because if I eat any other time of the year I'm just going to get very very fat so yeah I'm not a big cheese person but when I do eat cheese I eat it very seriously.

CLAIRE: 01:11:35
Okay, then. Books you're listening to, anything you want to recommend. [No, I mean I was addicted to Audible.] Or books you're reading. You don't have to be listening to it. You could be reading it.

PETER: 01:11:47
Well so in terms of listening you know I was addicted to Audible I had the subscription I used to get stuff all the time I'll be listening to like true crime and murders and all these horrible different things, I never listened to anything technical like in an audiobook just because it doesn't work for me like it's I need to read that type of stuff on paper [Okay.] But then it tends to be things like about the Linux kernel and things of that kind of nature which is just out of personal interest so most of the books I read have no direct implication on my job you know unless it's something that I'm reviewing because I'm going to mention it in an issue or something like that but that's not very often, so no.

CLAIRE: 01:12:25
So would you say you read to escape? You read to kind of get away from your work?

PETER: 01:12:31
Yeah I don't read a lot of books now I used to read a lot more in the past because I got three kids I don't really get a chance to do that now, so yeah I read a lot of technical books just to kind of scratch the intellectual itch, let's say, but in terms of novels and things you know I used to read tons and tons when I was younger I used to read a lot horror novels like Dean Koontz and Stephen King and that type of thing and then later on sort of like very sort of highbrow sci-fi like J.G. Ballard I was very heavily into for a while but yeah lately i've not read any novels oh in quite a few years so yeah it's very sad but uh you know.

CLAIRE: 01:13:11
Okay.

PETER: 01:13:12
These things come and go as come and go with age I think you know eventually once the kids have moved on and well if that ever happens uh you know I'll be sitting there with a glass of wine the garden, reading novels again, I'm sure, at some point, but this is not the stage of life for that.

CLAIRE: 01:13:26
I'm sure you're going to be hesitating when they're in that last summer before potentially going to college and you're trying to decide whether to encourage them to get a job or not or whether they can laze around. You'll be sitting there thinking, hmm, okay, if I do this, if I do that, what are my choices?

PETER: 01:13:41
Yeah, and yet the thing I know just from age and experience is that no generation is like the previous generation. Like, you know, my parents couldn't identify with things that my generation was going through. And I already know I will not be able to identify with things that my children's generation is going through. So to a large extent, I leave it up to them. Like, I know how it, you know, how it worked for me. Like, they're going to figure it out if they're smart enough to, you know. I am the safety net for my children. I am not their guide in telling them what to do.

CLAIRE: 01:14:13
Parenting advice from Peter Cooper.

PETER: 01:14:16
Yeah, so maybe not a cheese story, but you get a cheesy story. So, uh, yeah.

CLAIRE: 01:14:21
I love it. All right. Well, thank you so much, Peter Cooper, for joining us today.

PETER: 01:14:27
No, you've been fantastic.

CLAIRE: 01:14:28
It's, it's been really a lot of fun.

PETER: 01:14:29
The time has flown by. [It has flown by.]

CLAIRE: 01:14:34
All right, wrapping up. If you liked today's episode and you want to hear more of these Talking Postgres episodes, you should subscribe on Apple, Spotify, and YouTube, or wherever you get your podcasts. Subscriptions are a great way to help the algorithms promote the show to new listeners. And please tell your friends too. And if you leave a review, that helps even more people discover the show. You can always get to past episodes and links to subscribe at TalkingPostgres.com. And transcripts are included on the episode pages on TalkingPostgres.com too. We actually put a little bit of effort to make sure the transcripts are pretty good. So, yeah, let us know if we're succeeding in that. And a big thank you to everybody who joined the live recording here on Discord and participated in the chat. Thank you again, Peter.

PETER: 01:15:24
Thank you very much.