skip to main content

 Half-Baked Ideas

These are fleeting, half-baked thoughts, that may or may not get fully baked into fleshed out blog posts. The notes below may also be syndicated to other platforms and websites. Enjoy reading my ramblings!

Not finding what you’re looking for? Browse the archives.
Interested more substantial thoughts? Read the Freshly Baked blog.

2022

November

Als iemand die zowel Webmentions als Pingbacks heeft bestudeerd en geïmplementeerd, begrijp ik nog steeds niet waarom de IndieWeb community zo hevig doet over het eerste. Pingbacks werken 100% hetzelfde. En Wordpress heeft al jaren ondersteuning voor Pingbacks.

“Jamaar dat is lelijk en oud, en XML!” Inderdaad, het is een XML. Maar het bevat ook niet meer dan een source en target URL. Mijn endpoint converteert trouwens de XML moeiteloos naar een WM en parset dat op dezelfde manier. Webmentions bieden op geen enkele manier een modernere of makkelijkere manier—het is ook een POST van een formulier.

“Jamaar Pingbacks zijn spam endpoints”. En wat denk je dat er gebeurt met Webmentions als dit officiëel wordt ondersteund door Wordpress? Juist ja. Heel vreemd argument dat ik keer op keer hoor. Validatie van Pingbacks strenger maken heeft niets met het protocol te maken maar met de implementatie van de verwerking ervan.

“Jamaar Webmention is van IndieWeb, Pingbacks niet”. Wacht even, “IndieWeb” betekent toch een onafhankelijke site hosten die jezelf voorstelt, niet het wiel opnieuw uitvinden door ingewikkelde APIs te verzinnen die eigenlijk al bestaan?

Microformats voor machine-parsable HTML daarentegen, daar ben ik alleen maar voor. Alleen… “Pingback conversaties” kan ook met zoeken naar u-comment, maar veel van die artificiële classes zijn in mijn ogen overengineering—h-entry kan ook gewoon weg: pak gewoon de <html/> root. Klaar.

Anyway, wat kritische bedenkingen! Bij deze een Pingb-euh, Webmention.

 |   |  Reply to:

Interesting approach, providing both options! I personally think DuckDuckGo’s search results could be better, plus it’s not neatly embedded into my own site: there’s a chance people won’t come back.

I couldn’t find your search link 😅 Why not add it on top of the blog list page? Also, if I disable JS, the message “Please enable JavaScript to use the search functionality, or use the form below.” appears below your footer, at the very bottom of the page, instead of next to the “start typing to see the results” input form.

That most popular posts link is interesting though, do you frequently update it by hand?

Oh, and in your “more” page, a link to /notes is missing, but I presume that’s on purpose? Currently, visitors can’t discover that here. Luke squishes multiple links together with a drop-down.

 |   |  Reply to:

August

Gotcha, I wasn’t aware that resources were the problem. There’s a long discussion on the discourse page of Hugo regarding this, but the gist is: that also works! 😅

I just tried that with my site. If I simply move every resource from /static to the corresponding folder, e.g. /content/post/2022/07/image.jpg, then it works without any change. Keep in mind that I include my images using the path ../image.jpg if you want your content file to be in the 07 folder as well. That’s not needed if you create a separate dir for reach post, and can be circumvented by hacking render-image.html.

Anyway. I agree that the whole SSG system can be needlessly complex (and confusing). To me, a Django or Wordpress setup doesn’t alleviate that.

 |   |  Reply to:

Hugo only generates the root my-post/index.html from posts/my-post/index.md. (I had to run Hugo first to test this.)

No it doesn’t. That’s completely up to the theme and usage of your _index.md files.

For example, I organize my Brain Baking content like so: content/post/2022/07/whatevah-title.md. Works flawlessly. Notes like these reside in content/notes/2022/08/timestamp.md. On Jefklak’s Codex, it’s the folders content/games/switch/some-game.md or content/articles/long-article-stuff.md.

I maintain the same structure in my static/ folder, making including media super-easy: ![](done.jpg). Just make sure to convert it to an absolute path in your RSS—that’s what render-image.html is for.

If you really want to know, you can dig in the source of my Hugo themes. Again, this might be the fault of the sloppy docs and the arguably high learning curve.

 |   |  Reply to:

This part made me want to reply:

Listens and watches probably constitute a large part of the posts of this website. Are they useful however? Do they bring anything to you, or to me? Honestly, I don’t think so. However, I enjoy seeing the summary about what I have watched and my listening habits.

So they do bring something to you: that summary. However, in my view, many IndieWeb enthusiasts overdo this bit. I also really really love lists and summaries etc, but simply keep track of them elsewhere: in my (analog) journal. Do they need to be put online? Is that the question you want to pose?

I honestly never read that stuff if I see it in my RSS reader. Except if there’s context added! Same for bookmarks/links: I don’t care for a random link collection, but I do care for summarized ones with some context. Which ones do you recommend me as a reader of your blog, and why? If it’s just stuff you’re gathering, why not keep it private?

So no, I don’t think every update should be a post. It doesn’t make sense

Agreed!

 |   |  Reply to:

Fundor 333 asked:

In your opinion something like Gitea with a syndication like Mastodon will solve some of the problems and move more people on this “Gitea with Syndication”?

I’d answer: yes and no. Yes, it will solve some problems—hopefully more easy collaboration across different instances. With GitHub, that’s not a problem, provided that everyone uses GitHub. And No, I don’t think it will move more people towards Gitea, since syndication and self-hosting are usually two “complicated” solutions. Note that I didn’t say complex. Most people will still find it too troublesome to move. Just look at Mastodon VS Twitter. The @user@mastodoninstance thing already trips most people up.

 |   |  Reply to:

Excellent summary Roy, cheers! I dug around in your archives and discovered you were into MtG way back in 2001—that’s exactly the same year as I started playing! Since you regularly post updates on your digital Arena grinds, I was wondering how you migrated from analog to digital MtG. I only play with “the real stuff”, but as a consequence, I regularly have trouble finding buddies to play with.

Since discovering Commander, I much prefer playing it like that: more chaos and politics, more crazy cards, and it’s not always the player with the most expensive deck that wins. Most of my stuff is geared towards a budget anyway.

Perhaps we can do a few rounds using https://spelltable.wizards.com/ one day? 😀

 |   |  Reply to:

Awesome to read your IndieWeb adventure!

Also, I think my Webmention workflow is lacking. I need to implement a way to automatically rebuilt my site when new Webmentions are coming in.

I never understood the need for “instant appearance” of comments on the site. Since I publish at regular intervals, that’s not a problem for me, but if you really want, you could simply revert to client-side JS code. This is the Disqus and Webmention.io way. Do a fetch() to your Go-Jamming endpoint, and for each mention, build the HTML layout in JS, and replace it in the DOM.

That way, every page is automatically up-to-date and you sidestep the Hugo regeneration needed part.

 |   |  Reply to:

July

Thanks for that short write-up, I’m a big Race fan and had no idea Dice Realms existed. Now I do and need to dig in. It looks amazing! I’ve been playing Roll for the Galaxy on and off with a friend digitally but do prefer the original Race. Custom dice, however, seem like something different all-together…

 |   |  Reply to:

Interesting, didn’t know this existed. But: this seems useless if 99% of the mentions you receive don’t include a vouch parameter. Are we to auto-reject those? Probably not since most of those are still genuine reactions/mentions. Validity still needs to be checked, even without one.

Also: just scrolling through this page again makes it painfully obvious how convoluted and complicated IndieWeb things become…

 |   |  Reply to:

Yeah that sucks, I’ve also encountered spam like this, and it ain’t all sourced via a Pingback, see https://brainbaking.com/post/2022/04/fighting-webmention-and-pingback-spam/. I “solved” this by blacklisting/whitelisting and a moderation queue but I have my own webmention server implementation. I honestly always get a little upset when people say “just unplug pingbacks”—I’ve had a few genuinely good interactions through that, and every Wordpress user automatically has support for that as opposed to Webmentions. Disabling something does not fix the spammers.

 |   |  Reply to:

June

Thanks for that Bloodstained report, I love posts like this! As a hardcore Castlevania fan, I had high hopes for this one—that partially waned as the development process dragged on and on. I played it on the Nintendo Switch and hated it. The performance was awful, and after more than a year, a “patch” that fixed a few things still wasn’t good enough. To me, the damage was done. What platform did you play it on?

Also, gameplay-wise, I agree with you on the spells: you usually stick to a few good ones you find. That’s okay, and my favorite Castlevania, Aria of Sorrow, where this game steals the soul system from, has the same “problem”. However, there Soma doesn’t scream Rigor Stomaea! Rigor Stomaea! Rigor Stomaea! as you spam your spells. Movement is stiff and slow, and some scenery is just plain ugly. (I ranted more in my review over at Jefklaks' Codex)

To this day, I much prefer Bloodstained: Curse of the Moon. Better music and gameplay, albeit not in a metroidvania style. I was glad it was over. I replayed all Nintendo DS Castlevania games at least twice and Aria of Sorrow like 5 times. I guess my expectations were heavily misaligned.

 |   |  Reply to:

Ik lees altijd mijn eigen posts opnieuw “in productie” in de feed reader die verschijnen na een uur. Meestal verbeter ik dan hier en daar typfoutjes, maar vaker zijn het niet de typfouten maar de grotere gaten in logica of redenering waar ik niet meer aan dacht en dan tijdens het lezen plots wel terug opduiken, die ik dan nog toevoeg.

Dat gezegd zijnde, zo’n verbetering/aanpassing duurt ook weer gewoon een uur voordat het online geplaatst wordt. Mij kan het weinig schelen of lezers de eerste of volgende versies van een post gelezen hebben. De kans is heel klein dat ze het gat in de redenering ontdekken, aangezien het hier voor mij altijd over een intern filosofisch gesprek gaat. Vroege lezers missen misschien een alinea.

(Ik push dit bericht naar de repo op 14:01, d.w.z. dat het op 15:00 online staat en de Webmention server aan de gang zwiert)

 |   |  Reply to:

A much clearer page compared to my archives page, thanks for sharing. I modified my own /archives to better reflect yours and Frank’s layout. Not yet sure what the cut-off point should be, I’m not that big on tagging yet and my blog isn’t as old as your, we’ll see how it goes. I have removed one-offs now, and might retro-fit in more tags from older articles.

 |   |  Reply to:

May

I’m having a hard time figuring out what’s going on in this picture 😄 I always try to avoid bike racks where you have to put your front wheel in between an iron bracket—once, when I came back, the whole frame was bent because someone pushed against my bike.

But this seems like a sure way to save space in the station area! (Once you figure out how to work those red handles, that is)

 |   |  Reply to:

An alternative could be ImageMagick’s array of convert commands? I’ve had success doing just that to convert a PDF into an image (or the other way around). I presume something similar is used by the API service. I guess it’s pretty easy to automate that yourself and just run those tools locally, or do you really need yet another third party service?

 |   |  Reply to:

I wonder how you set your date frontmatter, which isn’t mentioned clearly? If I make a new note, I fire up a stupid JS script that creates the file accordingly—something akin to your method—but also formats the date (and pries out the active Firefox tab I want to “respond” to using Webmentions, see https://git.brainbaking.com/wgroeneveld/brainbaking/src/branch/master/hooks/prenote.js )

For me, I don’t want that many prompts and find those annoying, I just want to automatically make the file in the right location with as much metadata filled in as possible. It then opens the file in Sublime. Title, tags, and content is added by hand.

 |   |  Reply to:

To take with me probably rules out fiction for me. Instead, I’d go for classics that inspire: Meditations (Aurelius), Essays (Montaigne). And of course, always keep a copy of The Art of Fermentation by Sander Katz and Bread by Jeffrey Hamelman, my personal bread baking bible. Perhaps also Drawing at the Right Side of the Brain by Betty Edwards in case one gets bored.

In any case, avoid all books on programming or computer science—they only drag you down!

 |   |  Reply to:

April

2021

June

@rubenerd Do you have a set of scripts on rubenerd.com that converts a typical youtube embed in Hugo to your nice image-with-play-button thingie? I’m tired of iframes and would like to adopt some sort of automation for this. Thanks!

 | 

February

Has anyone else noticed posted links on Twitter get re-encoded into t.co URLs in order to gather even more 'insight' data? That said, I'm having trouble finding anyone on the Fediverse… Any hints?

 | 

I'm Wouter Groeneveld, a Brain Baker, and I love the smell of freshly baked thoughts (and bread) in the morning. I sometimes convince others to bake their brain (and bread) too.

If you found this article amusing and/or helpful, you can support me via PayPal or Ko-Fi. I also like to hear your feedback via Mastodon or e-mail. Thanks!