provoost
Sjors Provoost

How do reply guys / gals avoid getting in loops? And could we trick them into one? https://image.nostr.build/7ea4f4b9d3ff1a83430fb9d5d31c0baf769913703c01c81249c5f741dd7f128e.jpg

2024-09-21 15:29:05

Make it so, please! #note149r…286f

2024-09-21 14:14:44

- reply

@npub1ye5…knpr some small tweaks you could make on noStrudel: update the timestamp in edits so it shows up below the original in a thread, and above in a timeline.
Also the yolo button doesn't give any indication that it succeeded. It should probably jump to the updated message.

2024-09-21 14:09:03

- reply

Completely separate from edits you could have a maybe-delete and write again workflow. This might make sense if you said something completely insane and want to make sure that all your followers see the newly crafted message.
This will like a duplicate post to any client / relay that can't handle deletes. But then hopefully it's clear from context that the original was unintended.

2024-09-21 14:09:03

- reply

Completely separate from edits you could have a maybe-delete and write again workflow. This might make sense if you said something completely insane and want to make sure that all your followers see the newly crafted message.
This will look like a duplicate post to any client / relay that can't handle deletes. But then hopefully it's clear from context that the original was unintended.

2024-09-21 14:05:09

My thinking on how replies should work with respect to edits, without knowing how it actually works...
The edit should point to the original message. Clients should render the edited version if they know about it. But a reply should always refer to original id - even if they never saw the original version.
That way threads stay sane for everyone involved, including for clients that don't understand replacements.
Furthermore replacements should not be kind 1, otherwise they look like duplicates to clients that support replacements.
To avoid ambiguity about what you actually responded to, a reply could *also* point to the edited version in a separate tag. https://image.nostr.build/5a0a6a9dd17751167640e324ec9e8e249823f34ce89618824c3aeac526db1e7c.jpg

2024-09-21 13:54:09

- reply

Oh the edit button did work! But it shows a new post.

2024-09-21 13:18:09

- reply

@npub1xts…kk5s any idea why Damus doesn't want to play this? Posted from noStrudel and hosted on my blossom server.
Oh I guess it's because I added .mp4 but that should have been .mov. cc @npub1ye5…knpr would be easier if the client added an extension. https://image.nostr.build/1a8a42e014010d76936638fa9dea95c096de182e03c80ef674501231350ae9cb.jpg

2024-09-21 13:14:31

Still waiting for a block... #BitAxe
https://img.sprovoost.nl/b4f7d6cd9a150174dc4912412a52264194aff2fc7312aadc659024fe4349264d.mov

2024-09-21 13:14:31

Still waiting for a block... #BitAxe
https://img.sprovoost.nl/b4f7d6cd9a150174dc4912412a52264194aff2fc7312aadc659024fe4349264d.mov

2024-09-21 13:14:31

Still waiting for a block... #BitAxe
https://img.sprovoost.nl/b4f7d6cd9a150174dc4912412a52264194aff2fc7312aadc659024fe4349264d.mp4

2024-09-18 14:17:30

- reply

Presumably they relied on security by obscurity and were not interested in mass adoption. That's what makes them a politically easy target. And probably not that technically sophisticated.

2024-09-18 13:51:08

- reply

What I fail to understand is why criminals keep using these niche tools.

2024-09-18 13:38:27

- reply

> Private companies that wish to ensure their services are used in compliance with the law also have an important role to play. They must ensure that their platforms are not safe havens for criminals and should provide mechanisms for lawful data access under judicial oversight and in full respect of fundamental rights.
Last six words are a lie. Or if you want to be generous: nothing more than paying lip service.

2024-09-18 13:36:36

- reply

They can't even hide their malicious motives: this is about encryption, not crime.
> The encrypted communication landscape has become increasingly fragmented as a result of recent law enforcement actions targeting platforms used by criminal networks. Following these operations, numerous once-popular encrypted services have been shut down or disrupted, leading to a splintering of the market.

2024-09-18 13:34:32

And once more a privacy tool is attacked by state actors who frame it as "crime tool". They would not hesitate to take down Signal if it was politically viable.
https://www.europol.europa.eu/media-press/newsroom/news/global-coalition-takes-down-new-criminal-communication-platform

2024-09-18 06:59:18

- reply

Could brand it "expansion pack"?

2024-09-18 06:57:37

@npub1az9…m8y8 for a non-travel coldcard it could be handy to have one pin for duress (when you need to type the number) and one to trigger an explosion (when the bad the bad guy is typing).
But I suppose customs wouldn't like that. And ordering that component separately might also invite the wrong kind of attention.

2024-09-17 19:18:35

- reply

Mmm, looks like everything after the first hash tag fell off the bridge... https://image.nostr.build/e270d6c73673793b2a3cd9efdab101f6d456b7663ac99eb97016dc8d860fb7bd.jpg https://image.nostr.build/62cd7fac12308061e9d046049645cf18fe038436aec113071aae7ed9bad10a0c.jpg

2024-09-17 19:10:13

- reply

Je zou nog kunnen toevoegen dat Signal aangekondigd heeft uit de EU te vertrekken als dit doorgaat (niet zonder het eerst aan te vechten).
Met als gevolg dat burgers gedwongen worden om minder veilige chat apps van ja-knikkers te gebruiken. Of zelfs Telegram :-)

2024-09-17 18:36:24

- reply

There's discussion on HackerNews too about whether these are:
1. Normal lithium batteries with some software exploit;
2. Batteries modified (e.g. intercepted during shipping) to explode
3. Added explosives
https://news.ycombinator.com/item?id=41567573

2024-09-17 18:31:49

- reply

If true that would indeed be wise. Also for everyone else using that particular pager.
The few smart phone battery fire videos I've seen reminded me more of fountain fireworks than the explosive sound Bloomberg published.
https://youtu.be/ywLOaUJvbwI

2024-09-17 18:05:41

Misschien kan @npub1n09…ryx2 via-via deze ietwat provocerende Tweede Kamer vraag inkoppen: kan het door onze veiligheidsdiensten gebruikte Pegasus, of een vergelijkbaar cyberwapen, smartphones van burgers laten ontploffen? Zo nee, kan de minister bewijzen dat dat niet kan?
Ik pak🍿voor het antwoord.

2024-09-17 17:04:28

- reply

Yeah this has been the case for decades. Probably a liability thing.

2024-09-17 16:03:17

Friendly reminder that with the right backdoor, your smartphone battery is probably also an IOT bomb.

2024-09-17 06:34:34

- reply

We'll still doing this. But haven't released an episode in recent weeks.

2024-09-16 21:42:50

- reply

When I understand it :-)

2024-09-16 21:15:25

- reply

ping @npub1493…vtzu again because I broke federation

2024-09-16 20:58:06

- reply

Illustrated here.
https://mastodon.nl/@sjors_1/113149212075658710 https://image.nostr.build/636e8fac85f54fc0fc0c5bf58396f7de844cc41e823ba998ff1e5a97ce4d5a5f.jpg

2024-09-16 20:56:54

- reply

Ok, so @npub1493…vtzu (which I use to host Bitcoin Explained) is doing something in their HTML that Mastodon parses in order to produce an inline player.
@npub1v5u…n0v5 should do that too, cc @npub1unm…d0j2.

2024-09-16 20:46:18

Just a random episode from last year.
https://fountain.fm/episode/VQAVfgmv3HLdKhaYEubf
#nevent1q…s5kq

2024-09-16 20:36:15

- reply

Turns out Mastodon doesn't like it when it asks about a user and the server doesn't reply for a minute and then hangs up with HTTP 499 :-) cc @npub1q3s…d26p
https://gitlab.com/soapbox-pub/mostr/-/merge_requests/118

2024-09-16 19:45:20

- reply

Added an issue: https://gitlab.com/soapbox-pub/mostr/-/issues/90
(not sure which ones I'll personally get around to implementing, but at least they're remembered)

2024-09-16 19:30:42

- reply

Searching for it on mastodon does hit my mostr instance, but nothing seems to get fetched.
https://f.sprovoost.nl/objects/1b061e39517163479162ef0a74ebf353bbb7857fb87cf538db7472524b274113

2024-09-16 19:27:50

- reply

I'm not seeing this last message on mastodon . nl either. https://image.nostr.build/9281b3e2eac9361dd072065de0a254e3912db3da4fc34b7fb2208c5f238cf3dd.jpg

2024-09-16 19:21:56

- reply

This is what should have been sent to your instance. Now if I like my alter ego @npub1m93…4fr3 it should show up on mastodon . nl too. https://image.nostr.build/21b2efd098bb00b11f3aad08547926cb57bbc4ca011041cf276e4240c2ee3b70.jpg

2024-09-16 19:10:27

- reply

Tagging @npub1xuh…nmqu

2024-09-16 19:09:35

- reply

Similarly in the other direction, when I follow you I don't get historical posts. It should proactively fetch them.

2024-09-16 19:06:21

- reply

The follow worked. And your previous follow might have been deleted / lost. The only way you get messages is when Mostr posts them to your inbox. Of if your instance cached it, or maybe received it from elsewhere in the fediverse.
That's why it should have an output. https://image.nostr.build/90d7eeaf02bd97592b71b098b6ded2203c8deca3cc0c53205b247ee84cdbb5d3.jpg

2024-09-16 18:35:10

- reply

Mmm, an inline URL would probably just render the social media preview. Looks the same, but isn't. Though it could be a nice workaround.

2024-09-16 18:21:10

- reply

Interesting! Though in this case Phanpy doesn't render the original for me either. Or do I have to make an account on their instance to make it work?
https://img.sprovoost.nl/9f3e470b365cb2e0c9661208b5bf0f72fad5a42412824803702f84c3c914eaa1.png

2024-09-16 18:17:54

- reply

Oh sorry this is not an issue on your side. I was debugging my personal Mostr instance.

2024-09-16 18:15:45

It seems Mostr is very picky about what constitutes a repost (without a quote, Announce in ActivityPub).
Is NIP-18 still the cool way to do this?
https://gitlab.com/soapbox-pub/mostr/-/issues/75#note_2113475736

2024-09-16 16:56:53

- reply

Or maybe Mastodon doesn't actually support quote toots, I vaguely recall some controversy. Let's try a straight boost then.

2024-09-16 16:46:12

Let's see if I can blast the quoted message and its image across my bridge.
#nevent1q…gjgv

2024-09-16 16:41:00

Note that if I quote post someone from the Fediverse, they will generally not see your replies.
This is because my bridge only allows my own stuff through towards the fediverse.
However, if I rebroadcast your reply to my own relay it should get through. And that probably happens automatically if I reply to your reply.

2024-09-16 16:28:58

- reply

You may have caught me during one of a million restarts. Try again?

2024-09-16 16:27:51

Fixed both issues.
My nginx was disconnecting clients after 60 seconds. That's because I ignored setting from the README.
I went for the NIP-42 approach. Made two upstream PRs for that, but at least it works locally now.
https://github.com/scsibug/nostr-rs-relay/pull/215
https://gitlab.com/soapbox-pub/mostr/-/merge_requests/117 #note1xmu…8rq2

2024-09-16 14:59:16

- reply

Though this reply isn't. cc @npub1m93…4fr3

2024-09-16 14:52:40

- reply

It should be visible now: https://mastodon.nl/@sjors@f.sprovoost.nl/113147699244386325

2024-09-16 14:46:47

- reply

Oh misconfigured something.

2024-09-16 14:27:55

Or repost of a reply? #note132d…7rfg

2024-09-16 14:24:53

- reply

Repost didn't get picked up by mostr bridge, but does a reply?

2024-09-15 15:18:07

I'm mostly down to issues (1) and (3).
I can't tell if my relay is unreliable, if clients don't always send events to it, or if the Mostr connection silently breaks. Will need to do more debugging.
For whitelisting (one freshly created npub per fediverse user) I have three ideas:
i) have the relay do nip-05 verification (but there's a chicken-egg where it rejects the kind 0 even which it needs to verify the pubkey)
ii) just add localhost / 127.0.0.1 to the whitelist (made a feature request)
iii) give mostr a random private key, whitelist that on the relay and then use nip42 (opened an issue to figure out how to implement that) #note1n75…74jz

2024-09-15 12:10:20

- reply

Inline tag @npub1m93…4fr3

2024-09-15 12:09:13

- reply

Reply with inline tag @npub1m93…4fr3

2024-09-15 12:07:39

- reply

2024-09-15 12:06:13

- reply

Guy

2024-09-15 11:58:27

Test

2024-09-15 07:41:38

- reply

Either own their own initiative or by regulating noise limits. Though the latter may be tricky: not all sounds are equally annoying, it's more complicated than just decibels.
#note1u64…t5nt

2024-09-15 07:38:08

- reply

Seems like their main problem was noise, which could have been fixed with immersion cooling.

2024-09-15 07:03:49

- reply

As the saying goes: diamonds are forever, atrocious architecture is just temporary!
Anyway, my guess is they need a lot of sales and marketing folks to keep the prices up?

2024-09-14 19:59:19

Alright, this personal Fediverse bridge with a patched* @npub1d03…pzfs is promising. But I'll need to debug and improve it more.
1. It seems Mostr stops listening to my relay, or fails to listen at all at startup.
2. Replies from Nostr to Mastodon doesn't seem to work unless you @ the person explicitly, but see *
3. I have to whitelist on my relay every account on Mastodon that I want to follow
4. If someone else on Nostr jumps into a thread it's a mess. They'll only get relayed accross the bridge if either I whitelist them, or I engage them - because then my own client will broadcast it to the relay. But if they then use Mostr proper they'll look like two different people on the mastodon side.
My goal is mainly to get it good enough so I can occasionally reply to people on fediverse instances that block the main mostr bridge.
cc @npub1m93…4fr3
* = attrociously, so most bugs are probably my fault

2024-09-14 19:49:22

- reply

... ... @npub1m93…4fr3

2024-09-14 19:47:45

- reply

.. .. .. @npub1m93…4fr3

2024-09-14 19:46:31

- reply

J @npub1m93…4fr3

2024-09-14 19:21:43

- reply

Tags

2024-09-14 19:20:47

- reply

But still...

2024-09-14 19:18:45

- reply

Yes it is.

2024-09-14 19:17:56

- reply

Is Mostr even listening to my relay?

2024-09-14 19:14:10

Yo @npub1m93…4fr3.

2024-09-14 19:05:37

- reply

@npub1m93…4fr3 and I need to you explicitly for my reply to come back?

2024-09-14 18:47:08

- reply

Ping 3

2024-09-14 17:14:38

Just a test.

2024-09-14 11:28:07

- reply

* west

2024-09-13 17:11:56

- reply

The last second is cut off from the audio: "... in certain communities."

2024-09-13 17:08:53

Learn from crackpots who to email your proposal to!
https://fountain.fm/clip/O8vOwOqv5VMMZmyzM4dv
#LifeHacks

2024-09-13 11:33:45

Looks like the German police is intimidating a non-profit that runs Tor exit nodes. Anyone know more about this?
https://artikel5ev.de/home/hausdurchsuchung-am-fr-16-08-2024/

2024-09-13 08:24:55

How many pull requests have you rebased this morning?

2024-09-13 06:48:49

- reply

You can suggest topics here: https://github.com/bitdevsamsterdam/bitdevsamsterdam.github.io/issues/26

2024-09-13 06:39:53

- reply

Do you know if those companies now also use that same mysterious German custodian?

2024-09-12 20:47:29

And a solid intro: https://youtu.be/7F5MKaJMxDc #note1ujw…dc74

2024-09-12 20:24:05

- reply

Afaik in The Netherlands they switched everyone over to an EU entity last year but nothing changed.

2024-09-12 20:09:55

Looks more likely that is a technical problem, not a privacy crackdown. For now, sleep soundly.
#note1xeh…74dr #note13mj…mvrj

2024-09-12 17:42:10

- reply

Ok, so in Germany it's not a third party using Kraken systems as a whitelabel, but a completely different system?

2024-09-12 13:41:32

- reply

Is it possible to eat dinner early for those going to the Socratic Seminar that evening?

2024-09-12 11:23:22

Next Amsterdam Socratic Seminar will be the evening before #BitcoinAmsterdam at the Bitonic (slayers of the Dutch Central Bank) office.
https://www.flockstr.com/event/naddr1qqyxydfjxyerydtzqgsgdp0taan9xwxadyc79nxl8svanu895yr8eyv0ytnss8p9tru047qrqsqqql9n03axug

2024-09-12 07:08:31

- reply

May invite him to Nostr!

2024-09-12 06:50:26

- reply

Duplicate posts are on my list of most annoying things too. You get used to it, but still.

2024-09-11 19:39:46

- reply

Joepie, verplichte malware voor je fiets!

2024-09-11 18:50:10

- reply

One thing that wasn't spelled out, IIUC, is that in order to know the depth of the transaction merkle tree in a block, you just need to know the depth of *any* one transaction. This is because the tree is flat. So the coinbase transaction is used for this, plus the requirement that it starts with 36 zeros.
Though this could be annoying if the coinbase is 4MB? (I should keep listening)

2024-09-11 18:43:23

Voskuil explains how an SPV client would validate tx inclusion without 64 byte soft fork.
Looking forward to how future versions of Fountain will handle clip sharing. It's a bit too tedious still. Ideally timestamps in the shownotes are some RSS annotation could automate the process.
https://fountain.fm/clip/6e6xMkP1ilq0JcoohJTI
#bitcoin #consensus

2024-09-11 15:28:07

- reply

It sounds like Switzerland did a better job at protecting privacy here. The police mobile app thing referred to Finland. Maybe someone there knows more details.

2024-09-11 10:52:52

- reply

It's based on income, not wealth, so the state might actually have to Elon :-)

2024-09-11 07:44:21

Swiss dude gets a 98.500 Frank fine for tailgating. Similarly someone in Finland got a €121.000 fine for speeding.
The problem with these income adjusted fines is privacy. Finland has a central database with everyone's income that cops can apparently just access. It's effectively a kidnap shortlist. Though to be fair, these guys were already flaunting wealth so for them the additional risk is marginal. But everyone is in that database.
https://www.ad.nl/auto/rijke-advocaat-krijgt-boete-van-105-000-euro-voor-bumperkleven-in-zwitserland~afefaaf9/

2024-09-11 06:46:22

Wow, is it that late already? Oh wait, I moved the clock forward to CPU mine a testnet4 block.