Jeremy Cherfas: Posts

Jeremy Cherfas

Punter? Or professional?

Monochrome betting slips in someone's hand

Gonzo Aussie podcaster bloke Mike Williams put this up yesterday:

The average punter goes to the races with $5 and expects to win $1000.
The experienced punter goes with $1000 and hopes to win $5.
The average podcaster does a little bit of distribution work and expects to gain 1000 listeners.
The experienced podcaster does a stack of distribution and hopes to gain five listeners.

I wonder though, whether the person who goes to the races hoping for a $5 return on their $1000 can be considered a punter. Experienced, certainly, but that’s the kind of person I think of as a professional gambler. And that may be why I have such a poor record of doing what Mike calls distribution. I cannot think of myself as a professional podcaster as long as podcasting does not supply the major part of my income. And as long as I don’t think of myself as a professional podcaster, I don’t seem to be able to do the stack of distribution I need to do to gain a few listeners.

Some money does come in, thanks to generous donors, and it is enough to keep the lights on, more or less, by paying for hosting. I’m grateful for that, of course, because it makes the podcast more than a pure labour of love. It is validation, too. But I’m under no illusions; Eat This Podcast is never going to provide me with much of an income. In that sense, I’ll always be an amateur rather than a professional.

But there’s another, much deeper problem with the distinctions Mike draws. The horse race gambler wins by picking the right horse most of the time. Sometimes — whisper it — the horse may not win strictly on merit, and the gambler is in on the fix. But what is the podcasting equivalent of picking the right horse?

There isn’t one.

Not one of the myriad pages that tell you how to grow your audience actually tells you how to grow your audience. They’re like the tipsters that hang around the Tote windows, giving each punter a different tip and hoping to collect a little something from at least some of them, as long as they can remember which horse they offered to whom.

So, is there a secret to distribution work?

I’ve tried audiograms with squiggly waveforms of enticing snippets, and detected absolutely no additional interest. Those things take real work, and I gave them a fair shake, assiduously doing the work for five consecutive episodes before abandoning it as pointless. Should I have continued?

I plug past episodes when something in the news or on Twitter seems to justify it, with occasional results. Slowly, I’m overcoming my reticence too, so that may be something I can do more of.

I’ve run giveaways, and become disheartened when five random winners in a row failed to claim the reward. The thought of rewarding existing subscribers for twisting the arms of their friends into subscribing in order to win -- what? -- is slightly abhorrent. Maybe I need to get over that and make it happen when the new season launches. Or maybe I should just pay for an advert in, say Overcast, and see what happens.

Eat This Newsletter is, I’ve been told, interesting and informative, and I recently stopped sharing the links anywhere except in the actual newsletter, while still promoting the fact that I’ve sent out a new issue. Maybe I should promote those links one at a time too, through social media.

Then I wonder, why bother? If my podcast is not going to bring in any significant income, why put any effort into gaining a few more listeners? Probably because listener numbers are an even more direct source of validation. I pursue things that interest me and then share them, because that’s what I do. If more people agree that they are interesting, that makes me feel better about pursuing them. And if some of them end up donating, that would enable me to pursue even more interesting things.

Flickr photo by John Wardell

Fri, 10 Jul 2020 11:30:00 +0200

Monthly report: June 2020

Rushing this out, a little, because tomorrow we hit the road and I know that if I leave, it won’t happen. So, possibly sketchy, but done.

Highlights of the month:


Steps up 1000/day, weight creeping down, sleep 8:05; lower on the month and 30 minutes lower over the past 3 months.





Even more paid work this month than last, and it is too much. I need a break, and maybe to turn some stuff down. But there’s always that fear. What if they never ask me again?

Hours logged per month

Percent of logged hours: 2020

Previous years are on an archive page.


Only four new posts in June, and a couple of older ones, both bread recipes that I noticed people were looking for.


This remains unlooked at: It would be good to error check the script that pulls in the podcasts I’ve listened to. Why does it occasionally miss an episode when the data clearly show I listened to it?

Final remarks

Very ready for a rest.

Here's the table

Click the triangle to see or hide the table
Month Total Daily Admin % ETP % Other %
06 171 5.70 26 19 55
05 170 5.67 40 22 38
04 175 6.03 36 18 46
03 164 7.50 38 27 35
02 129.0 6.50 45 17 38
2020-01 89.25 5.25 48 19 43

Wed, 01 Jul 2020 17:20:00 +0200

Seed Stories in Brief

three cucumber seedlings

Just caught up with the podcast version of Dan Saladino's excellent Seed Stories from the Lockdown on Radio 4's The Food Programme, and of course it prompted a flood of emotions, reminiscences and recognitions. As soon as I got home, I walked through the terrace, thinking about which of the plants I had grown from seed and, more particularly, home-saved seed.

There's the dill, the Cupani sweet peas and Grandpa Ott's morning glory, the flower that launched Seed Savers Exchange. There are two kinds of tomato, Pugliese Stays Green and the strange brown-fruited variety that The Main Squeeze takes care of. Ancho peppers saved last year need planting out very soon, and the velvety sunflowers that look like they're coated with Vantablack are growing like crazy. The nasturtiums volunteer crazily all over the place, and then there are the wisteria and the yellow trumpetbush, grown from seed that I gathered but that have not yet produced seeds of their own. And that's not counting the things I have multiplied by vegetative propagation.

Each and every one of them gives me the same immeasurable joy every year. They're a far cry from food security, but they're pretty great for mental security.

And right after doing that, I made a thing over at Eat This Podcast.

Sun, 21 Jun 2020 12:20:00 +0200

Flaxseed currant ciabatta

Last week I was sorting through my recipe folder — the physical one, which is a mess of printed, scribbled and ripped bits of paper — and came across a bread that I had not made before and that for some reason called to me.

It was very good, so I wrote it up at the other place: flaxseed currant ciabatta.

I should note, too, that it made an excellent basis for a canonical hipster lunch.

avocado smooshed on toasted flaxseed currant ciabatta

Sat, 20 Jun 2020 14:30:00 +0200

Oh, The New Yorker, how could you?

I’ve ranted many times here about the wanton misuse of biological scientific names. Those are the things, generally in italics, that name a species in such a way that we can all agree what it is we are talking about; Rudbeckia hirta, for example, rather than black-eyed Susan.

People are forever mistakenly referring to the Rudbeckia hirta, which is just plain wrong. A scientific name is a strong proper name, which never takes a definite article.1

So it was with amazement that I saw The New Yorker, of all places, famed as it is for the breadth and depth of its fact checking and grammarian exactitude, refer to “the Candida auris” on page 66 of the 18 May, 2020 issue.

Photo of the offending section of the new yorker article

This was in a review of the Merlin Sheldrake’s2 book on fungi, which sounds like it might well be worth a read, as an old fan of Paul Stamets and fungi in general.

  1. Unless, of course, you had two specimens and wanted to say something like “the Rudbeckia hirta on this herbarium sheet is very different from the Rudbeckia hirta over there.” 

  2. See what I mean? 

Thu, 18 Jun 2020 19:00:00 +0200

Monthly report: May 2020

For us, another good month. Things are slowly becoming easier; making sure I have gloves and a mask in my pockets when I go out is becoming a habit. Washing hands when I get back in has been for a long time. It’s great that the park is open, even if it is overrun most of the time. Children being at home can’t be easy. Everything has really been rather good. I missed the travel I had booked, but the online Dublin Gastronomy Symposium was very enjoyable and not nearly as difficult as I feared it might be.

Highlights of the month:


Steps creeping up, weight creeping down, sleep a little lower.





A couple of times I just had too much to do. Also failed to appreciate just how exhausting concentrating during a >4 hour Zoom meeting might be. So, needed to get on top of that a couple of times. Otherwise, all is good. Maybe a little too much?

Hours logged per month

Percent of logged hours: 2020

Previous years are on an archive page.


Fourteen new posts in May, but no old ones. And though I did read on a fair number of days, I notice that when I’ve finished a book I enjoyed, I’m reluctant to start another for fear it won’t be as much of a pleasure.


It would be good to error check the script that pulls in the podcasts I’ve listened to. Why does it occasionally miss an episode when the data clearly show I listened to it?

Final remarks

I’ve ignored a couple of things I want to do, as a result of things I have to do. A perennial problem, how to carve out time for myself and not fritter it away here and there.

Here's the table

Click the triangle to see or hide the table
Month Total Daily Admin % ETP % Other %
05 170 5.67 40 22 38
04 175 6.03 36 18 46
03 164 7.50 38 27 35
02 129.0 6.50 45 17 38
2020-01 89.25 5.25 48 19 43

Sat, 06 Jun 2020 12:20:00 +0200

Code and living processes

My online chum Lewis Coles has, like everybody and their dog, been baking bread in these troubled times, and he’s not happy.

Lewis is a software engineer. I can’t be sure, but I guess that he thinks that if you follow a set of instructions, you should end up with the same result every time. So he’s understandably peeved.

“Why the hell are the instructions to make bread, so poor,” he asks, preceded, of course, by “Of course I’m not serious, but …”

I take his point. Judging by the questions that come up in the bread forums I frequent, many of the instructions going around clearly are pretty poor. And, Lord knows, I’ve had my share of disastrous errors in instructions going way back. But there are also some fundamental misconceptions in Lewis’ approach.1

For a start, I’m dead against the idea of muscle memory as a guide. Consistent output requires consistent input, in bread as in code. You need to weigh the ingredients, not eyeball them, just as I need to remember to tell my code that, despite appearances, ”this is a single word”.

Even then, there may well be differences in the ingredients that are essentially invisible; the protein content of the flour, humidity, temperature, the state of the leavening, all sorts of things. That’s why, even with weighed inputs, you watch the dough, not the clock.

I’m not a good enough coder to think of an obvious analogy. Maybe something like not assuming that if you try to do a certain thing to a file, it will always work, but testing to see whether it has worked before moving on.

In bread, as in code, that takes experience. For bread, it requires baking the same basic bread quite a few times in a row and adjusting one thing at a time.

Dark tomato on a piece of my bright yellow experimental turmeric bread

I realise this is all in fun, and that’s the spirit in which I am engaging, something I said I would do three weeks ago, to my shame. I applaud anyone who learns to bake bread, and try to help as a thing I can do in exchange for the help I have had with code. In that spirit, to anyone who wants to improve their skill:

Some instructions are written for people who already know what they’re doing, others try to suit complete beginners, and they’re the most difficult ones to get right, because each beginner’s problems are unique. If I assume I’m writing for someone who knows how to bake, as I was this morning, it is much, much easier to tell them what to do. Just as it is for Lewis when he assumes I know what I’m doing with code.

Happy to answer questions about bread. Happy to ask them about code.

  1. As there are with people who think of DNA as code, but that’s a giant subject I am unlikely ever to tackle. 

Sun, 31 May 2020 16:15:00 +0200

Wayback Engine for the win

The Wayback Machine, part of The Internet Archive, is an absolutely essential part of the open web and, as it happens, my work. I use it for all sorts of things, most visibly finding archived versions of web pages that have vanished for whatever reasons. I support them with an annual donation, but today I felt compelled to give more, after The Guardian reported that Dominic Cummings falsified the record of what he said when about coronaviruses. The prompt:

First spotted by Jens Wiechers, a data scientist, the edits are verifiable through periodic snapshots of the blog saved by the Internet Archive, which shows the change occurring between 9 April and 3 May this year. A hidden record on Cummings’ own site shows the post was edited at 8:55pm on 14 April, the day he has told the public he had returned from his trip to Durham.

Cummings’ weblog is hosted by WordPress.com, so I’m guessing that the “hidden record“ is one of their secrecy-by-obscurity URLs that shows revisions, but I don’t honestly care. The Wayback Machine’s smoking gun is good enough for me.

Wed, 27 May 2020 11:07:00 +0200

An excursion!

Portrait of Jeremy Cherfas wearing a homemade mask at the bus terminus in Roma centro

The sky was the bluest it has been in a long time. There was a bit of a breeze keeping things cool. We had a cappuccino at the bar, my second since mid-March, me keeping a respectful distance from The Main Squeeze, and then we ran for a bus.

So exciting, my first excursion off the Green Mountain since it all began.

A woman glared at me as I got on the bus, tapping her mask, because mine was still in my pocket. Trusty beat up old bit of T-shirt to the rescue, I clung on for dear life as the driver put his foot down and rattled down the hill. About ¾ of the seats have a sticker on, saying not to sit them to preserve a safe distance. TMS snagged one, while I tried to stay upright until the glaring woman got off. Bravely, I occupied her seat, confident that her mask and gloves would protect me. And then there we were, downtown.

Not that you would have known it. The streets were empty, in contrast to the big nearby park, which is absolutely heaving with people all day because lots of people are still not at work and all the kids are at home too.

Sant’Agostino was open, and mostly empty too, so we took the opportunity to admire the Caravaggio and the Rafaello and a lot of other beautiful stuff. Saint Rita is big there, as an Augustine nun, and yesterday was her day. I assumed that she might be the patron saint of beekeepers, being as how as a child she was once enveloped by a cloud of white bees who did not harm her, but she isn’t. That honour goes to Ambrose, Gobnalt (?) and Valentine.1

A few of the shops the via Swankissimo were already empty and vacant, but not much had changed overall. We wandered around, sat down for an expensive coffee, then returned to the bus stop and came back up the hill.

All is slowly returning to normal.

  1. What about St. Bernard of Clairvaux? I dunno; I’m searching online same as you. 

Sat, 23 May 2020 19:30:00 +0200

A tool to make knowledge gardening easier

It’s so nice when things just happen to come together and move me forward. A teeny thing, but mine, and I hope worth sharing.

A week back I wrote about tending my zettelkasten garden, and trying to become more of a curator and less of a hoarder. Two days later someone mentioned readwise.io, which I had a quick look at. That same day, a cyberchum posted something “from this morning’s Readwise email”. So I was well primed.

And I thought, if Readwise can surface a quote or a highlight or something, maybe I can surface a random zettel to tend? At first I focussed on PHP, and discovered glob(), which looked interesting. Then I looked more closely at The Archive, because that seems the logical place in which to tend the chosen note. The Archive is not yet Applescriptable, but it does have an external URL scheme. I’ve never knowingly used such a thing, but after a bit more reading around, found examples of it in use. On a whim, I opened the Terminal and typed.

open thearchive://match/201901111413

Blow me if it didn’t just work.

So I decided to focus on a bash script, did a bit more searching, and came up with this.

cd /path/to/zetelkasten/notes
a=( * )
filename=( "${a[RANDOM%${#a[@]}]"1"}" )

filename=$(basename -- "$filename")

filename=${filename// /%20}

open thearchive://match/$filename

Tied that to an Alfred trigger and it does exactly what I need. Any time I need a quick respite from whatever I’m doing, I fire it up and tend that one note.

Make it easy, and it happens.

I should explain that the script is not entirely cargo cult. I did have to understand and modify bits of it. Picking a random filename here, with the details explained here so that I could understand it well enough to return just a single value. The Archive does not want to see the file extension, which took me here, ensuring that it would deal with almost any silly filename I might throw at it. And finally, The Archive does not want to see spaces in the filename. That one I knew myself.

Sat, 16 May 2020 14:50:00 +0200

Page created: Thu, Jul 16, 2020 - 09:05 AM GMT