It can be so hard to debug IndieWeb problems when they arise in Known. For now, this should be a webmention of the linked post, because Chris Aldrich reported a problem. But even if this works, I may need to go outside this installation to test properly. Or, perhaps, try webmention.rocks.

The past 24 hours saw perhaps more activity in the #known IRC channel (yesterday and today) and than I have ever seen before. Near the end of it all, jgmac1106, having previously voluntold me to be the first rotating organiser, voluntold me to “call all of today a meeting of the Open Collective”. Obviously you can’t have a meeting without minutes,[1] so here they are.

It all started with jgmac1106’s heartfelt plea that he just wanted to publish his site, “not learn backend engineering” and contemplating starting afresh. LewisCowles raised the question of how to reward Open Source software developers and maintainers, and that started a discussion of what it would take to put Known on a commercial footing.

Jgmac1106 was of the opinion that easier install with auto-update was needed. Lewiscowles and jeremycherfas thought that better direction of the project was needed, with a model that offered installation, domain management and updates, for a fee.

“Make it Known would be such a great tagline if we could get Sir Patrick Stewart on board.” Lewiscowles

There followed further discussion of operational models, including micro.blog; pay for hosting, including updates, and some backfeed, with a free offering open to IndieWeb if you have a capable site elsewhere.

On funding, jeremycherfas related his early experience hosting through IndieHosters and jgmac1106 talked about applying for grants to fund specific pieces of Known development. We played around with numbers, concluding that nobody knew enough to build even an outline business plan. There did seem to be agreement that venture capital should be rejected from the outset, while collectives and cooperatives could provide a more desirable structure, and that any kind of structure needs direction.

After a gap, some other people joined the channel and mapkyca explained that right now, a bigger block than money was time as he is working flat out. He also said that the maths does not work out for SaaS.

Benatwork then rejoined the meeting and explained in some depth the history of Known, including funding decisions and his original vision.

The original intention was to build a community platform that could be hosted securely, with discussion not monitored by the likes of a Facebook. … [I]t was never built to be an indieweb platform or an individual blogging engine from the start. The core idea was: flexible, social feeds that one or more people could contribute to, with per-item access control and integrations both in and out. I still believe that it has most value as a multi-user platform.

Major problem: we gave our entire platform away as open source, and it turns out there was a strong correlation between people who wanted to use it and people who didn’t want to pay. Although they were happy to pay for an account on a shared host, which of course didn’t go to us. So it didn’t really work as a scalable business.

Benatwork then filled us in on recent developments and why his direct involvement has dwindled, all of which is very understandable, closing with his belief that SaaS is not the way forward.

Jgmac1106 then voluntold jeremycherfas to take the lead on setting up monthly meetings for the next three months, as the first rotating organiser.[2] He also shared his idea of having something like Known to offer local media as something they can sell to subscribers as a built in social platform.

In response to a question from Aaron_Klemm, Benatwork shared the Known roadmap on github. He also explained some of the past technical decisions and that maybe some of those should be revisited to improve the product as a whole.

People shared their different ideas of what Known could become for them, with the question of the current admin tax prominent. Cleverdevil said he would be happy to pay mapkyca to update his site, raising again the potential demand for SaaS.

Benatwork’s vision is Known not as a blog CMS exclusively, but rather:

What Known can do is create a stream of many different kinds of content, and present it differently based on context. Filtering is a similarly powerful idea. “Show me all posts that are sensor readings and photos tagged with bats, from January 1st.”

There was some discussion of other aspects of Known that need attention, including the templating engine, which mapkyca said he hopes to separate completely from the back end.

Chrisaldrich raised the possibility of working with Reclaim Hosting to devise a package similar to what Reclaim offers universities, i.e. Reclaim does the heavy lifting for turnkey Known installs while allowing a small group of others to support people who signed up. Aaron_Klemm supported this idea strongly.

There was a lot more discussion of various ways in which Known could contribute to community internet literacy and how it might be used alongside other web publishing tools.

This summary is an entirely personal capture of the discussion; corrections and comments welcome. (You know how to do that, right?) I’ll suggest some times for an online meeting through the channel.

Thanks to an unfollowable stream on Twitter, I came upon the website of Joshua Nudell, an historian with an interest in ancient Greek breads. A post of his, translating from Athenaeus’s Deipnosophistae, refers in passing to "the loaf from rye (or spelt)". That's strange. So I left a comment on the post, as follows:

I don't know Ancient Greek, but I do know some ancient and modern cereals, so I am hoping you can elaborate on this. Does the list mean two different loaves, one of rye and one of spelt? Or does it mean that rye is sometimes known as spelt, which would be a very interesting reading indeed.

Thanks.

This could be interesting.

Spammers say the sweetest things:

Jeremy cherfas is the well known writer of this century. He is famous for the suppleness that are still like by many of the people. We should also read the blogs about him to gain knowledge for our own self.

One of the developers of Sunlit, a photo-sharing app that is part of the Micro.blog ecosystem, contacted me to say that “the images on your site have a MIME type of application/data”. I’d like to say I understood immediately what the problem was and what it meant, but I had to do some learning first. It wasn’t as simple as the extension, the bit after the filename that indicates whether it is a JPEG or PNG kind of image. Rather, it was about what my server tells your browser about the image.

To backtrack, Known stores all files as blobs that contain the actual file data, the 1s and 0s. Your browser, when it receives a post from my server, can often sniff out what kind of thing (image, audio, text etc) that blob of data represents and do a good job of showing it to you. Normally, you wouldn’t even notice. One clue is that if you right-click on an image, and ask to open it in a new tab, it actually gets downloaded instead, I suppose because the new tab doesn’t know what else to do with it.

Anyway, I confirmed that the source file for most images did not have an extension (which would have told the browser directly how to deal with it). Most, but not all. Files I had uploaded to my site directly did have an extension and the correct MIME type. The “bad” files had come from OwnYourGram or Quill, both of which are part of the joyful #IndieWeb. They use a standard called Micropub to send things to a suitably equipped website.

It seemed unlikely that both Quill and OYG would fail to send the requisite information to identify a photo, so I went digging into the code that Known uses to decide what to do with a post sent by Micropub. I made a bit of progress but although I could see more or less what was happening, I couldn’t see how to make it right.

Fortunately Aaron Parecki, who built Quill and OwnYourGram (and so much else), was around and gave me the clue I needed to investigate: curl -I example.com/file.

One beautiful feature of Quill is that if it is sending a photo and if the receiving site has a media endpoint for receiving files (which Known does) it uploads the file, shows you a preview and tells you the location of the file. With that, the curl command shows that the temporary file has the correct description of Content-Type: image/jpeg. Once Known has processed the whole post from Quill, though, the file that contains the image shows as Content-Type: application/data.

Somewhere between receiving the temporary file from Quill and storing it permanently, Known fails to give it the proper MIME type.

I wish I knew enough to discover where the problem lies. Most likely Marcus Povey – who keeps the wheels spinning at Known – will be able to do the needful, now that I have submitted an issue. And Sunlit will be able to share my photos far and wide.

@cn suggested I use the date as the title of a Post in Known, to ensure that the contents of the post gets through to micro.blog intact, and, as so often, he is correct.

All I have to do now is remember to salute him next #micromonday

For all the joy of the #indieweb, and the pleasure of civil discourse, I am becoming incredibly confused by aspects of micro.blog. There’s the question of titleless posts, of which is this is one as an experiment, versus status updates. There are posts that appear to be contributions to an interesting conversation but aren’t because they have been cross-posted automatically from elsewhere. And there is the lack of a scroll back, which means that as I follow more people and choose not to check in the middle of my night, stuff vanishes irretrievably from my timeline. 

There are also issues with Known that are nothing to do with micro.blog.

None of this is insurmountable. For me, though, it does add friction. 

During the virtual IWC tonight, we were discussing third-party clients for publishing to websites, essentially Micropub clients and MarsEdit. And it occurred to us more or less simultaneously, that I do not use Micropub for the site that supports it out of the box, whereas I do use a client for the site that does not support Micropub out of the box. And that is because the post-creation UI is nice and simple for Known, and a right mess for WordPress.

So, just to be difficult, I'm using Quill for the first time in a long while to create a Post in Known.

#IndieWeb

More good help from Manton, cleverdevil and others, but alas no nearer (although I may have eliminated some possibilities).

1. I was allowing both HTTP and HTTPS. There could have been some kind of mismatch, I suppose, but after editing .htaccess to force SSL, it made no difference to micro.blog's OS X app.
2. Then thought that possibly a different endpoint would help (despite the fact that I know all the same details are on the home page). 
3. Pointed micro.blog first at /profile and then at /profile/jeremy; still no good.

At this stage, given that Manton managed to get everything working from a clean install of Known out of the box, I think I need to try the same. If that works, well, if nothing else, it works.

I had been fretting about losing data, but if I install into a new sub-domain and it works there, I can always edit the config.ini to point back at the old database. It will be a good opportunity to see how good the instructions are to install at Dreamhost. Last time I managed without any instructions, and I also didn't write up my experiences. This could be an opportunity to pay back.

#indieweb

I have never yet been able to post from my micro.blog to this stream, although the feed from here is reliably picked up there, and brid.gy reliably pulls replies from there to here. @manton suggested we move my complaints to help@micro.blog, but I can see no way of actually engaging with that account. So this afternoon, I decided to attempt to go back to the beginning.

It was a miserable failure.

Here's how it went:

1. Revoke all current authorisations for micro.blog
2. There were four of them, two from yesterday when I last tried.
3. Launch OS X app
4. OK!
5. “If you’re using WordPress or another server, first open the preferences window and enter your web site URL to set it up for posting within the app.”
6. Roger that. But the old website was still there. Is that going to be a problem?
7. Start a new post; the old website is there at the bottom. This is going to be a problem.
8. Same old Same old; Error sending post.
9. Post does not arrive at micro.blog
10. Post does not arrive at WithKnown
11. WithKnown Error log is empty.
12. Access log shows no sign of anything from recent attempt to post from micro.blog
13. Check to make sure I have up to date micro.blog app
14. “Micro.blog can’t be updated when it’s running from a read-only volume like a disk image or an optical drive.” Move Micro.blog to Applications folder using Finder, relaunch it from there, and try again.
15. Strange. Check path to micro.blog.
16. Path is “/Applications/Micro.blog.app”.
17. Check the App Store; disappointed but also content that there is not an update.
18. Delete web site URL from preferences; quit micro.blog, mostly for superstitious reasons; launch micro.blog.
19. Very strange; web site URL is still there. Or back? Maybe the app pulls it from micro.blog?
20. Repeat; same outcome. Superstition justified. Go to my account at micro.blog.
21. See I have three App tokens. The one for MarsEdit is definitely pointless, as I am not hosted at micro.blog. Remove it.
22. Throw caution to the winds; remove the tokens for IOS and OS X
23. OK, IOS now says “Internal Server Error” on attempting to connect. I think I ought to sign out now and then sign back in.
24. Phew. All is good. And I have a new app token.
25. On iOS, try to write a new post; insert my Known site; authorise micro.blog; write a test post. Post it.
26. “Error sending post”. Tear hair out, as now I do not seem to be able to post to micro.blog from iOS app.
27. Go back to 22; remove iOS app token and authorisation token at WithKnown.
28. Log back in. Can no longer post without adding WithKnown, and posting gives an error, as at 26.
29. Try again from OS X; same error as at 8.
30. Post to WithKnown; feed is picked up.

Any and all suggestions gratefully received.

#indieweb

Playing Lyle Lovett singing Texas River Song, even though I know it's a stretch, to celebrate having eaten a giant bowl of my own dogfood.

A couple of weeks ago I followed Chris Aldrich to reading.am, which is a neat little spot for just putting down a marker for something that you're reading. I wanted more. I wanted to be able to save links to the things I marked. And now, a little over two weeks later, I've done it.

I have a PHP file that fetches the RSS feeds of things I've marked in reading.am, looks for any that are new since the last time the program ran, and then POSTS the results to Known's micropub endpoint. #IndieWeb PESOS for my bookmarks!

There is no way I could have done it without amazing help from people on the IndieWeb IRC, and it isn't perfect by any means. There's more work to be done, for sure, before I even think about sharing the code.

But hey, it works.

And, as my main helper said, "Launch early and iterate often".

I'll be doing that.

In the past couple of days, prompted by Marty McGuire's write-up, I raved about the potential of Audiogram to help promote the podcast by making it easier to share audio clips to social media -- by turning them into video clips. This afternoon, having managed to get tomorrow's episode edited early, and having had to chop quite a few interesting digressions, I thought I would have a serious play.

Tl;dr: It worked. I'd show you here, but I haven't found an easy way to upload video to Known yet. If you want to see the result, you can go to Patreon right now.

Installing Audiogram was not entirely plain sailing. Marty used Docker, and so despite warnings from other IndieWeb friends, I tried the same. All went well out of the gate, but then fell at the first. Something to do with virtualbox. So I switched to Homebrew and that did the needfull. Even then, though, Audiogram wouldn't start, but the error message made it clear that I needed to update node.js and npm. That done, it still wouldn't start.

Turned out I already had a local server running, via MAMP, and that was getting in the way. Switching off that server, and all was good.

That was two days ago. Today, I tried to use it for serious, and although there were plenty of hiccups along the way, I got there.

The instructions for modifying the theme are very straightforward, and with a bit of trial and error I was able to create a background for any future clips.

Uploading the audio, inserting the caption, all that was dead simple thanks to Audiogram's editor. Actually generating the videos, though, generated error after error, and some of them scrolled through several screens. But I kept my nerve, turned to search engines and StackOverflow and eventually got there.

Some of the fixes seemed to be pure voodoo. There's an invisible file that one of the Audiogram developers suggested deleting. The first time I tried that, it worked beautifully. The second time, not so much. Nor the third. But then, it worked again, at which point I called a halt, for now; a wise decision in my opinion.

I'm looking forward to seeing whether clips will attract listeners to the podcasts in their entirety. I put the first one on Patreon because the episode is not yet public, although Patreons have received it. I'll probably use clips there as bait and see how it goes. Once episodes are public I'll send clips to Twitter and, maybe, Facebook which will, I think make it relatively easy to trace any impact. And if the whole process isn't too hard (getting to the first video uploaded took almost three hours today) then I can imagine it might be useful to promote older epsiodes too, when there is a news peg.

So, grateful thanks to Marty McGuire and WNYC.

I’ve been reminded by Chris Aldrich of something I think I knew before:

[M]ost major CMSes (including Known) strip out or severely limit (for security reasons) the html that is accepted in comment fields. … Many also will mark as spam comments that have one or more URLs in them. As a result doing fancy or even mildly complicated html or markdown in replies is something for which most platforms just don’t build.

That’s fair enough. As ever, spammers are spoiling things for everyone. I do have an objection, though. If I am legitimately signed into my own site which, in the #indieweb, is where I will be if replying to some other site, then I’m unlikely to inject malicious code. And if I’m a spammer, and signed in under a false flag, then I’m not likely to need such subterfuges.

A really helpful CMS would, surely, allow me to do all the formatting I want on something I am generating myself, regardless of the specific type of entry.

Chris makes another point:

The other issue in status updates and replies is that they’re often syndicated to other platforms and it’s a more difficult issue to properly do this with each snowflake social media silo depending on how they individually handle html/markdown (or not).

Well, yes. But that’s not my problem on my site. Let them strip all they want, frankly, as long as the leave the link to my reply alone. As Chris acknowledges …

Either way, the end result on the other person’s site isn’t something I can ever control for, so I try not to sweat it too much. :)

For now, I think I’ll sweat this just a little, and add the u-in-reply-to by hand, and hope that does the needful.

I've been moaning to anyone who'll listen that there seems to be something wrong with Known; Micropubs could not seem to find the syndication targets. And other people had the same problem, I believe. But after a really enjoyable virtual Homebrew Web Club meeting, the problem might after all be at my end.

@zegnat created a fresh install of Known as we watched, hooked it up to Twitter, and was instantly rewarded with Quill seeing his syndication target, which it resolutely refused to do on my instance of Known. (It failed actually to syndicate, but that's a separate issue.)

So, now I need to try a fresh install myself. And as @Jeena suggested, better to do that on a new and different subdomain than risk messing everything up.

Alas, there is no way on Earth I can do this until near the end of the month.

I can wait.

The question "does @WithKnown have a future?" is cropping up increasingly frequently of late. And the "official" answer is that it most definitely does, look at all the activity on github, nothing has changed. And it's true, there has been a lot of activity and things are moving, if you go and look. But for someone just looking in and trying to decide whether to use the software, the lack of outward facing activity must be a bit off-putting.

Or maybe it isn't.

I have no idea.

All this was [kicked back and forth on the WithKnown IRC channel yesterday](https://github.com/mapkyca/KnownchatLogs/blob/master/2017-03-23.md), with -- alas -- no input (yet?) from the developers.

I'm going to continue trying to understand Known because right now it seems to me the best place to continue pursuing #indieweb ideals.

Writing something nice again, but I have enabled the Markdown plugin on Known.

And maybe a headline too

Like this

Utterly bizarre; when I went to write a new post, the above was already there. In other words, the content of the post from Quill, without the block quote.

>This should be a markdown block quote.

I'm knackered. Will try some more tomorrow.

#withknown