# Learn more about twtxt: # https://twtxt.readthedocs.io/en/stable/ # https://www.uninformativ.de/twtxt-mention.html # # nick = movq # url = https://www.uninformativ.de/twtxt.txt # avatar = https://www.uninformativ.de/avatar.png # # For out-of-band replies: `echo ed.vitamrofninu@txtwt | rev` # # This file is available via Gopher, too: # url = gopher://uninformativ.de/0/twtxt.txt # However, since Gopher has no concept of user agents or any kind of # headers, I will not be able to see that you're following me and, thus, # will not be aware of any replies or mentions that you send. Please # either use the HTTPS feed or drop me an email. :) # # Legacy, don't use: # nick_alias = vain # # prev = kvwn5oa twtxt-old_2021-10-23_kvwn5oa.txt 2021-10-23T11:05:09Z @ (#zipopla) Lol, alrighty then 😁 2021-10-24T04:21:12Z Just saw the new CSS on twtxt.net, looks good! 👍 2021-10-24T05:04:29Z @ (#kwptbpa) “One hour” makes me wonder: Is this a DST bug? According to [this](https://www.timeanddate.com/time/change/usa/seattle), DST ends on November 7 in Seattle (that’s where she lives, according to her web site), so I guess we’ll see in a couple of days. 😁 2021-10-24T05:57:07Z @ (#kwptbpa) I haven’t fully thought it through. 😁 I’ll try. Time zones seriously make my brain hurt, though. 2021-10-24T05:57:30Z @ (#zlgcnzq) Morning! ☕ 2021-10-24T06:07:54Z @ (#kwptbpa) I think it *could* be something like this:  Seattle has UTC-8 in winter and UTC-7 in summer (IIUC). Let’s say it’s 9:34 AM in her local time. I suspect her script might be adding 8 hours to convert her local time to UTC – throughout the entire year. So her script will generate “17:34-00:00”. This is correct in winter, but one hour off in summer, when DST should have been applied: With DST, the script should have generated “16:34-00:00”. This would make her twts appear to be one hour in the future.  What’s strange is that some of her twts do have a `-07:00` time zone. That would be the correct time zone at the moment! The bulk of them has `-00:00`, though, so there’s probably two different twtxt tools at work, one of them ignoring DST … 2021-10-24T06:11:28Z (#kwptbpa) Heh, her feed starts on 2021-03-19. That’s a few days after DST began in Seattle (2021-03-14). So we really don’t know if my hypothesis is correct, because we have not yet seen twts of her without DST in effect. 🤣 2021-10-24T06:31:17Z Pushed one last change to jenny’s repo (`url_alias` has been removed, it’s obsolete). Going for a release in the next few days. And then I’ll tackle what’s in the `TODO` file. 👷 2021-10-24T06:32:00Z @ (#kwptbpa) Yeah, no, I don’t see a way to fix this in clients. 😕 2021-10-24T08:08:15Z Malevolent product design: ![](https://movq.de/v/8a2b2590f0/s.png)  That’s a netbook sleeve/case. That bright long-ish thing in the middle? It’s plastic. It doesn’t serve any purpose whatsoever. At least not from my perspective, but maybe from the perspective of the manufacturer: It’s made from that kind of plastic that dissolves over time. And that’s exactly what’s happening here. That thing is all sticky and gooey. So the idea is probably that I’m now supposed to buy a new one.  Well, screw that. I’ll probably try to cut it out and sew something else in there. 🤔 2021-10-24T08:28:14Z (#dpxms2q) Pretty disgusting process.  https://movq.de/v/d1e6d0fe98/  And now there’s some yellow “residue” left. 🙄 If I’m able to clean this up some more, I might just leave it like that. 2021-10-24T09:49:29Z (#dpxms2q) To be fair, that thing is already more than 10 years old. So it’s not *that* bad. But then again, everything is fine *except* for that useless plastic thingy. That’s the *only* reason. 🤦  I think I can keep using it. It won’t look pretty, but I don’t care. 2021-10-25T14:52:49Z @ (#yp3fkba) On the Internet, no one knows you’re a dog. 🐶 2021-10-25T15:21:12Z @ (#u6uumbq) There is no support for deletion, really. `lastmods` only affects retrieval of feeds (i.e., it sets the `if-modified-since` HTTP header), but when a feed *is* retrieved, then it is always processed in full. So, eventually, you get all twts that are present in a feed.  I’m curious, what is your use case for deleting twts? 2021-10-25T16:16:31Z @ (#u6uumbq) Hmm. 🤔  At the moment, it’d be rather easy to implement this in jenny: We could simply store the time stamp of the newest twt we’ve seen (that’s different than `lastmods`) and then discard older twts. I think @ asked about that a while ago, too. I *think* I didn’t implement this yet, because I simply didn’t need it. The twtxt world was so tiny that it didn’t matter. But you’re right, it’s going to be a problem in the long run.  Nobody ever really thought about “long term” in the twtxt world, if you ask me. I mean, it’s the same thing when *writing* twts. My feed keeps growing and growing … forever? Do I truncate at some point? Will all clients be able to handle that truncation? 🤔  I want to implement HTTP range requests next. That’s going to be interesting. Also, range requests and truncation probably don’t mix very well. 🤪 But I haven’t put much thought into it, yet.  tl;dr: I’ll put this on the TODO list after all. 👍 2021-10-25T16:50:10Z @ (#vez43wa) Cheers! 🍷 2021-10-25T17:59:42Z @ (#u6uumbq) Ouch, thank you. That (“feeds don’t need to be ordered”) pretty much kills the idea of doing range requests. 💀 I also didn’t consider metadata changes. 🤦  FWIW, I know of at least one feed in the wild that *prepends* new twts: http://codemadness.org/twtxt.txt 2021-10-25T18:00:17Z @ (#4bs3vta) I think I like this idea. 🤔 2021-10-25T19:54:57Z @ (#u6uumbq) Ya that gets out of hand real quick. I don’t think it’s worth it. There’s too much that can go wrong. 🤔 2021-10-25T19:57:05Z Now my laundry dryer is acting up. That’s one of those items that I don’t really need and, in retrospect, these guys were a huge waste of money. Sure, it’s luxurious and neat and fancy, but … Nah. If that thing dies, then it dies. (I just hope it doesn’t catch fire in the process.) 2021-10-26T16:41:23Z @ (#7tmfmpq) Maybe. But *in theory*, it’s legal to append twts in the middle. Or, and I’ve done that myself a couple of times, delete twts somewhere in the middle of the feed.  I mean, doing that *probably* means that the partial feed I’m retrieving starts in the middle of a twt now, and that’s something that I can detect, so I can then fall back to just retrieving the full feed. But it’s not guaranteed. There will be subtle bugs, it feels very fragile. I’m not convinced that it’s worth the effort. 🤔 2021-10-26T16:46:01Z @ (#3ezvila) I’ll have a look at those specs, too. The gist of it really is just `next` and `prev` in metadata, IIUC. And that should work pretty well. 🤔 2021-10-26T16:58:06Z Since we’re all on the brink of becoming alcoholics here, I’ll have a sip of this now and go watch the season finale of “Midnight Mass” on Nyetflix: ![](https://movq.de/v/6a44614f4a/s.png) 2021-10-26T17:01:38Z (#htxyltq) (In all seriousness, I don’t really drink a lot. It’s been ages since I was actually *drunk*, must have been my early 20ies. It’s just a small glass, 2cl, of good whisky every now and then, because I like the taste of it.) 2021-10-27T12:13:29Z @ (#3ezvila) … does that mean that archived feeds are supposed to be “frozen in time”? Nothing ever added to them, either? (Sorry, haven’t read the actual RFC, yet.) 2021-10-27T12:13:45Z A WebKit update broke my browser. Gotta fix that now. Fun. 2021-10-27T12:48:41Z @ (#r33kvxa) WebKitGTK. No Windows or Android, sorry. 😅 https://uninformativ.de/git/lariza 2021-10-27T13:15:15Z @ (#r33kvxa) 🥴🥴🥴 To be perfectly honest, that slogan didn’t work out *that* well in the long run … 🥴😢 Keeping up with GTK/GNOME shenanigans is anstrengend (exhausting) as well. (Same thing with my VTE terminal btw.) 2021-10-27T15:00:43Z (#3ezvila) So it’d go like this?  - `twtxt-2021-10-25.txt`: Feed of twts starting at 2021-10-25. For now, this is the “current” feed. - `twtxt-2021-10-25.txt` keeps growing for a while by appending new twts at the end. - Once `twtxt-2021-10-25.txt` is “full”, we add a `next` field to its metadata which points to … - … `twtxt-2021-11-01.txt`, which is initially empty except for a `prev` field that points back to `twtxt-2021-10-25.txt`. From this point on, `twtxt-2021-11-01.txt` is the “current” feed. It keeps growing in the same way. Eventually, it’s full and superseded by the next (partial) feed. - `twtxt.txt` could then be a symlink to the “current” feed file. - Non-current feeds could now indeed be considered as “archived”: Nothing ever changes in them anymore (except for metadata maybe? What if I change my nickname or feed URL?)  Well, at least that’s how I’d go about implementing this in jenny. 🙂 2021-10-27T15:02:37Z @ (#r33kvxa) I used https://addons.mozilla.org/en-US/firefox/addon/vim-vixen/ for a while until I eventually gave up on managing add-ons in Firefox. (They kept changing the way how add-ons are installed, which is super annoying for me, since I *always* use https://uninformativ.de/git/clean-browsers.)  But really, all browsers suck. 😁 2021-10-27T15:05:05Z I hope I get around to upgrading my OpenBSD boxes at the weekend. 🤔 There’ll probably be a short downtime of this server. 2021-10-27T15:41:36Z @ (#3ezvila) Nah, that was just an example. 🙂 I’d probably go for “`$n` months or `$m` twts, whichever comes first”, both variables being customizable. I mean, there’s no point in “paginating” a feed if there’s very little traffic. 2021-10-28T14:19:32Z (#yyiauda) Something something “Tom Scott”, “that way lies *madness*!” 2021-10-28T14:21:36Z @ (#eek7qna) Is it really that much of a problem, though? (It would be really funny if *you* introduced a bug now which accidentally drops some twts because something’s “wrong” with their time stamp. 🤣) 2021-10-28T15:57:20Z @ (#fr6hsja) Ah. Now I got it. That makes a lot more sense. 🤣 2021-10-28T18:05:52Z I hardly use my car since this covid thingy started, so I hardly noticed the gas prices going up recently. https://movq.de/v/9ad8e9b11d/sprit.png But what’s actually *more* interesting is how the price appears to have been “capped” at 1.6 EUR/l in the past. 🤔 I wasn’t really aware of that, either. 2021-10-28T18:06:47Z (#cbk2uaq) Maybe it’s just an illusion, I don’t know the first thing about gas prices and economics. 🤣 2021-10-29T11:02:24Z @ (#4bajyya) I’ll probably be going back to public transport and/or bike. I’ll keep my car until it breaks down (or becomes way too expensive) and then that’ll be the end of the “car era” for me. Public transport isn’t *great* where I live but it’s good enough, especially when I only go to the office 1-2 days a week. Cars are just super expensive and require so much maintenance, ugh … 2021-10-29T11:02:45Z (#ksg3bqq) What a weird choice. 2021-10-29T11:10:29Z (#ksg3bqq) … not that “Facebook” was any better. 🤣 2021-10-29T13:53:46Z @ (#ksg3bqq) Oh. 🤦 2021-10-29T17:20:11Z Removed the HTTP-to-HTTPS redirect from my web server for better compatibility with old browsers: gopher://uninformativ.de/0/phlog/2021-10/2021-10-29--website-http.txt 2021-10-29T18:10:39Z @ (#o5cto2q) 😂  Maybe I should have written it like this? 2021-10-29T18:15:42Z (#o5cto2q) Well, not’s a link, but it still looks funny. 😅 2021-10-29T18:43:55Z (#opw257a) Yes, I’m also pretty confident that this is a good extension. If I have the time and energy (maybe I won’t), I’ll open a PR with a formal draft for this on the weekend. 👌 2021-10-29T18:57:22Z @ (#c7dnr7q) I think your proposal is actually pretty good. Especially step 2 – that should add a lot of robustness. I hadn’t thought about doing this. 🤔  Still, I’m not sure if I’d implement that (in my client). It adds quite a bit of complexity and I’d like to keep things simple(r). Granted, I probably have a bit of an “extreme” view here: Complexity is the devil. 😈 I’m not dismissing this idea in general, I’m just speaking for my client.  (Also, I’d first like to see the pagination thingy implemented. I think we can gain *a lot* if we get all the “main” feeds down to a few kilobyte, instead of megabyte. And actually, pagination is just a different form of “range requests” …) 2021-10-30T07:27:36Z I see twtxt.net supports `if-modified-since` now, yay! 🥳 2021-10-30T08:30:19Z An attempt to formalize the pagination extension we were talking about: https://git.mills.io/yarnsocial/yarn/pulls/494 2021-10-30T10:05:17Z @ (#g6vcweq) Great! I’ll have a look after lunch. 😊 2021-10-31T06:06:45Z @ (#vm5bptq) jenny defers this decision to the `requests` library:  https://docs.python-requests.org/en/latest/user/advanced/#encodings  Honestly, I’d rather not interfere with that.  They refer to RFC 2616, which indeed says ISO-8859-1 should be the default for `text/plain`. However, RFC 7231 says in appendix B that this has been removed and it’s now up to the media type. When we look at https://www.iana.org/assignments/media-types/media-types.xhtml#text, we see RFC 2046 listed for `text/plain`. RFC 2046 including its update RFC 6657 specify US-ASCII as a default (https://www.rfc-editor.org/rfc/rfc6657#section-4). So, uhm, which one is correct? ISO-8859-1 or US-ASCII? None of those things specify UTF-8 as a default for `text/plain`, though, this only applies to *new* text media type registrations.  It’s a rabbit hole. That’s why I’d like to defer this to `requests`. 2021-10-31T07:01:34Z @ (#vm5bptq) I haven’t opened that image, but I’m pretty sure I know what it is. 😁 😁 2021-10-31T12:25:57Z (#zvi7l7q) Done. Couldn’t use sysupgrade this time, because I had to re-do the partition layout, so this wasn’t a particularly smooth upgrade … 😩 2021-10-31T15:51:24Z @ (#vm5bptq) Hmmmmm, good point. I completely forgot that twtxt enforces UTF-8. 🤦 I think I'll change this, yes. 2021-10-31T16:09:32Z @ (#vm5bptq) I pushed commit cb02422 to the repo. 👌  (Yes, the change is super simple. I just wasn’t sure earlier if I *wanted* to do this. But you’re absolutely right, twtxt says feeds must be UTF-8, so there’s no point in caring about the `Content-Type` header at all.) 2021-10-31T16:10:58Z (#vm5bptq) Fun fact, I had to jump through some hoops to configure *my own web server* to serve my feed with `Content-Type: text/plain; charset=utf-8` instead of just `Content-Type: text/plain`. 🤣 Maybe I’ll remove that hack from my config now … 2021-10-31T16:58:53Z @ (#nekynfq) Ya, why not. (I never *really* understood why the subject wasn’t first. 🤷) 2021-10-31T17:03:01Z “Hello darkness, my old friend. Now you start at 5pm.” 🌚 #DST 2021-10-31T17:53:34Z @ On the branch `track-lasttwts`, I’ve started implementing that “don’t recreate deleted mail files” thingy. So when you delete/move/archive twts, they should no longer reappear when you run `jenny -f`. Feel free to give this branch a try. 👌 (Bugs may lurk, it’s very fresh.) 2021-10-31T17:55:36Z @ (#el7d3ja) Gah. 🤔 I’ll have to rework this. Thanks for reporting! Never seen `F` before. 2021-11-01T14:38:22Z (#o5cto2q) Netscape 4.0 and Opera 3.62 on Windows 3.11: https://movq.de/v/2371a069ec/ Not perfect, especially with all the UTF-8, CSS and PNGs, but I like a lot that it’s still *usable*. 2021-11-01T14:39:32Z @ (#w4qg37a) #31 immediately caught my eye. 👌 2021-11-01T15:14:34Z @ (#w4qg37a) Something like that? 🥴 https://movq.de/v/664e4240bf/s.png 2021-11-01T18:36:27Z @ (#o5cto2q) It’s ancient software, but those are all “new” installations. The one on those screenshots is from 2007, I think. None of the real installations from the 1990ies have survived. 😢 At some point, I just thought: “Hey, let’s install all that old stuff again!” 🥴 2021-11-01T18:40:14Z @ (#p522joq) Great. 🙂 Guess I’ll merge that into `main` soon, then. 2021-11-01T20:11:53Z @ (#p522joq) Ah, yeah. Stupid mistake. It gets overwritten. Will fix tomorrow. 2021-11-01T20:13:53Z @ (#p522joq) Yeah, something along these lines. I think there’s something missing, I’ll have a look tomorrow. 👌  (Haha, will you look at that. Now we’re exchanging diffs over twtxt. I guess the original authors didn’t see *that* coming.) 2021-11-02T18:04:37Z (#p522joq) I merged that branch into `main` with a fix. It’s almost the same diff that you posted, just a minor edit. Thanks! 2021-11-02T18:06:26Z (#el7d3ja) On `main`, all maildir flags should be ignored now, so hopefully no more duplicates. 2021-11-02T18:13:04Z @ (#tzlkn3q) 😲 2021-11-02T18:13:20Z (#tzlkn3q) Just realized, you probably don’t see that emoji. 🤣 2021-11-02T18:41:27Z @ (#tzlkn3q) IIRC, his client runs in urxvt, which isn’t particularly good at that fancy font stuff. I might be wrong, though. I just vaguely remember seen screenshots like that. 🤔 2021-11-03T04:07:29Z @ (#tzlkn3q) On Linux, you don’t have that many options, realistically. If you want good emoji support, you basically need a terminal that uses the VTE library (or maybe something from the KDE world, dunno). Something like sakura or lxterminal.  Or you can try my xiate, but I don’t know how much longer I can keep maintaining that … GTK4 is lurking and everything is pushing hard towards Wayland, which means removing X11 features from libraries like GTK (because … yeah, why?). I have not yet started porting xiate to GTK4, though. If shit hits the fan, I might just switch to suckless st or use urxvt myself. Oh, and xiate is not memory friendly (anymore), it uses ~44 MB per terminal. 2021-11-03T04:09:32Z @ (#4663qha) Glad you like it. 👍 Let’s just hope that I don’t screw up and introduce lots of bugs in the future. 🤣  (Seriously, with all the changes lately, I’d like to slow down a bit and spend more time on quality, refactoring, more tests.) 2021-11-03T17:22:44Z @ (#bk3qohq) These didn’t happen to be the twts by jlj? His pod was unavailable or something. 🤔 2021-11-03T17:25:57Z @ (#bk3qohq) One of the biiiig problems with this feature is that I don’t use it myself. I just let my mailbox grow and I honestly don’t see the need (yet) to archive or delete twts. So I don’t observe those bugs myself. 🙁 (Actually, implementing features that I don’t use myself is against one of my ground rules. If I don’t use it myself, it’s not “battle tested”. Bugs may lurk and I never find them. Hmm. I was arrogant enough this time to think: “Oh it’s such a simple feature, I can implement that, even though I don’t use it.” Maybe that was wrong.) 2021-11-03T17:26:49Z (#v4bt6ga) Yep. For now, you can just save an empty file and then exit your editor. That’ll properly discard it. It’s just not very obvious that you have to do it like this. 2021-11-03T18:33:48Z (#bk3qohq) Alright, for testing purposes, I deleted all twts in my box up until 2021-10-18. Let’s see how it goes. 2021-11-03T18:43:54Z @ (#tzlkn3q) Fun fact, good support for emojis often also means good support for Unicode in general and that means … well, see for yourself, run this in a couple of different terminals: `curl -s https://raw.githubusercontent.com/nickboucher/trojan-source/main/C/commenting-out.c` (I’m honestly surprised why it took so long for this to get popular.) 2021-11-03T18:47:07Z Speaking of terminals and emojis … The fact that jenny uses these little “📣” symbols means you can’t really use it in XTerm, or urxvt, or st, or Linux VT, or … 😒 It’s nice *if* your terminal supports it, but terrible otherwise. 2021-11-04T14:34:49Z @ (#bk3qohq) Hmm, not sure. Here’s one scenario:  - You fetched jlj’s feed before you started using jenny’s new `lasttwt` feature. - jlj’s pod went offline. - You experimented a bit, archived more twts. - jlj’s pod came back. You fetch the feed, but `lasttwt` is now unset, so you get all of its twts, including those that you already archived.  Dunno. It’s not *that* unlikely that it happened like this. But then of course, I have absolutely no idea what you did on your computer, so it’s just a *very* wild guess. 😅 2021-11-04T14:38:31Z @ (#tzlkn3q) Yeah, it’s the correct URL. It’s a show case of Unicode “right-to-left” control characters. Depending on which program/terminal you use to display that file, it looks very different. See also: https://www.lightbluetouchpaper.org/2021/11/01/trojan-source-invisible-vulnerabilities/ 2021-11-06T06:26:23Z @ (#4663qha) Heh, thanks, but that’s not really my kind of thing. 😊 I just publish that stuff and if someone likes it, great – if not, then 🤷. 2021-11-06T06:27:18Z @ (#bk3qohq) Strange. Well, I’ll commit that stuff to a repo now, too, it’s a good idea. 👌 2021-11-06T06:39:46Z @ (#mrodbxa) Oh my god! 🥰  We’re having lots of wild parrots here, too, but they’re way too scared to get *that* close to humans. 😢 The legend goes that these guys escaped from a bird breeder some decades ago – they’re not native to central europe. https://movq.de/v/b5cc712fb7/ 2021-11-06T08:36:41Z @ (#mrodbxa) Aren’t cockatoos parrots? I always mush them together. 😁 2021-11-06T08:38:11Z @ (#mrodbxa) I guess it happens a lot that they escape. 🤔 It always amazes me that they still manage to survive in the wild. I mean, -10°C in winter, sometimes less? That’s crazy. 🤯 2021-11-06T13:14:41Z TIL: https://en.wikipedia.org/wiki/Catgut was used for stringed instruments like guitars and violins. So, humans played songs on the intestines of animals. How barbaric. 🥴 2021-11-06T13:16:40Z (#jzwcgyq) To be fair, I never really thought about it … But I was expecting something made from hair, like this one: https://en.wikipedia.org/wiki/Morin_khuur 2021-11-06T17:42:13Z (#t3wvu7q) @ It’s gone. Forever. This happens every time you touch a clock. So, careful with that! 2021-11-06T17:51:49Z (#v4bt6ga) @ @ Let me ask you guys a question: What’s your workflow here? You hit the “reply” hotkey, the editor comes up, then what? You start writing and writing and then, at some point, you decide “nah” and just do `:q!`?  The reason why I’m asking: I think I sometimes hit `:w` in between. This is kind of muscle memory for me (I’m actually not 100% sure if I do it or not …). So a `:q!` wouldn’t work for me, because I would accidentally send a half-written twt.  Another thing to consider: Maybe users have configured their editors to do auto-save after a while. 🤔  Either way, I suspect the current way to cancel a twt (saving an empty file) has to stay (whether I implement an *additional* mechanism to detect “cancel” or not). It is very explicit and the user/editor won’t do it by accident.  So, yeah, what’s your workflow? How do you use it? 2021-11-06T17:52:56Z (#i3hppxa) @ 2021-11-06T18:56:34Z (#v4bt6ga) @ I see. 🤔 We could check the file’s mtime after the editor has exited. If it’s the same as before, then no write has been made. This breaks for editors with auto-save, but it’s probably *good enough* in most cases. 🤔 2021-11-07T04:48:13Z (#v4bt6ga) @ I agree, just extending the “snip” line with a short message is probably the best solution. 👌 2021-11-07T04:55:00Z (#g6vcweq) Sooooo, I was quick to create that PR, but I haven’t touched it for a week now. Hopefully I can refine it a little more today? We’ll see. 2021-11-07T04:55:42Z (#zmxk6pa) @ Yep, no more DST in Seattle. 2021-11-07T05:02:48Z (#5s356la) @ Whoa 😲 You’re on top of the world now! 😁 Almost looks like another planet. 2021-11-07T05:10:56Z (#i3hppxa) @ Pretty good, I think. 👌 The only thing I noticed: h1, h2 and h3 are all the same size, which makes it a little hard to read the page, e.g.: https://adi.onl/fl.html 2021-11-07T09:19:20Z (#v4bt6ga) Neither was I. 😳😳😳 2021-11-08T14:33:20Z (#v4bt6ga) @ Actually, Git was one of the “inspirations” for the current behaviour. It’s a little bit different in jenny (I could make it match Git’s behaviour, though 🤔), but similar nonetheless. For example, `git rebase -i` pre-fills the editor with some commands and when you just exit the editor, Git will execute those commands. If you want to *cancel*, you have to delete all commands and then exit.  This commit now adds a little message: http://www.uninformativ.de/git/jenny/commit/9be74280cced4c1e00734f02d2ae98aaedfaff02.html 2021-11-08T14:34:22Z (#v4bt6ga) (Nevermind, I’m stupid, jenny’s behaviour already matches that of Git.)