05nelsonm
Matt Nelson

Navy -> NASA -> Sphinx Chat -> Strike -> FLOSS
PGP: 0x61471B8AB3890961
Kotlin Multiplatform and C are the way

Public Key

npub1j3fs72nlngpdgy2njn0rycd0f45lyv8r0j45sjvqcdmszzc0agmqvh7lq8

NIP-05 Address

Profile Code

nprofile1qqsfg5c09fle5qk5z9fefh3jvxh5660jxr3he26gfxqvxacppv875dsprfmhxue69uhhyetvv9ujuumwdae8gtnnda3kjctv9ukqzxrhwden5te0v4ehqetvdphjuemfwf5kumewdaexwqgcwaehxw309aex2mrp0yh8xmn0wf6zuum0vd5kzmqpzemhxue69uhhyetvv9ujuurjd9kkzmpwdejhgqg5waehxw309aex2mrp0yhxgctdw4eju6t0qyxhwumn8ghj7mn0wvhxcmmvjlqrru

Publishing to

Author Public Key

npub1j3fs72nlngpdgy2njn0rycd0f45lyv8r0j45sjvqcdmszzc0agmqvh7lq8

Show more details

Last Notes

2025-12-22 17:52:22 UTC

- reply

I've been going back and forth on making one, but it looks like quite the undertaking.
What were some of the learned lessons you had from iteration to iteration?

2025-12-18 19:10:38 UTC

- reply

Case in point: craigslist

2025-12-18 17:59:43 UTC

- reply

Only idea I have had is to use templates for PRs and Issues with a required check box to tick "AI was not utilized in the generation of this PR/Issue".
If it's not checked, auto-close that shit. If they lied, block them.

2025-12-17 19:06:49 UTC

- reply

They said the same thing about social security...
SCAM

2025-12-17 18:48:54 UTC

- reply

https://blossom.primal.net/4893d8b05c80cea062330c813ed537799a63f2b127687ea1a8fd356b6e8536f6.png

2025-12-17 18:24:23 UTC

I never knew how much I would need something... Buffered streaming of UTF-8 encoded bytes is OP.
https://encoding.matthewnelson.io/library/core/io.matthewnelson.encoding.core/-decoder/-companion/decode-buffered.html

2025-12-15 19:41:25 UTC

- reply

All for show.

2025-12-15 10:11:15 UTC

- reply

The fog will wear off eventually, but you'll come back to us eventually.

2025-12-15 10:10:04 UTC

- reply

Nothing beats black dark roast coffee and a cigarette for locking in to conquer the day. Nothing.

2025-12-15 10:02:40 UTC

I think today is going to be release day... Tons of new features, lots of bug fixes, several deprecations, and many many performance improvements.
Has been a fruitful couple of months, but now I must #shipcode

2025-12-15 09:58:40 UTC

- reply

How would you work around platform limitations on number of file descriptors that can be open at once? Or the incredible amount of overhead that would require? Or the fact that symbolic links are not a thing on Windows?
IMO, a filesystem is not the right tool for the job, a database is.

2025-12-15 09:09:01 UTC

- reply

I assume this is to open a file?
Personally, I took the approach of splitting into 3 separate functions; openRead, openWrite, openReadWrite.
This allows for different return types which inherit from a single common one, and makes clear the contract for which that thing (interface) can and cannot do. Eliminates nearly all API confusion and/or misuse.
openRead doesn't need to take parameters as it must always exist.
openWrite and openReadWrite take an exclusivity enum MaybeCreate, MustCreate, MustExist. This also allows a mode to be expressed for applying appropriate permissions immediately on creation when MaybyCreate and MustCreate are expressed.
openWrite also takes an append boolean parameter, constraining appending/truncation behavior to that function only.
I spent a lot of time playing with my API, bumping into and figuring out workarounds for several platform specific limitations (such as pwrite with O_APPEND being broken on Linux/FreeBSD).
https://github.com/05nelsonm/kmp-file/blob/master/library/file/src/commonMain/kotlin/io/matthewnelson/kmp/file/OpenExcl.kt
https://github.com/05nelsonm/kmp-file/blob/master/library/file/src/commonMain/kotlin/io/matthewnelson/kmp/file/FileStream.kt

2025-12-14 18:21:52 UTC

- reply

Vivaldi is a solid Chromium based browser.

2025-12-13 08:11:55 UTC

Anyone else find that writing documentation is more mentally draining than writing the actual code 😂?

2025-12-13 01:14:25 UTC

- reply

Oh, and "democracy".

2025-12-13 01:14:00 UTC

- reply

That oil

2025-12-11 03:59:57 UTC

They are gonna print so much money over the next 5 years. Buckle up.

2025-12-06 22:04:54 UTC

Anyone know of a *native desktop Nostr client for Linux?
*native: as in NOT an electron app

2025-12-06 04:35:18 UTC

- reply

Vivaldi has been the most stable non-Google Chromium based browser for me thus far; have been using it as my primary for 3+ years now. I use Brave and Vanilla Chromium, as well as Mullvad Browser, for some things, Brave being the most finicky.

2025-12-03 04:34:05 UTC

- reply

Is this supposed to be a slight?
Full retardation is the ultimate superpower; you're unstoppable with it...

2025-12-01 23:51:39 UTC

- reply

The link I provided is for adapting a PiHole which does in fact provide those features.
Yes, there is added latency because of the tor network (a feature, not a bug), but I have been running this setup for about 5 years now behind an OPNsense router with little to no issue.

2025-12-01 21:39:05 UTC

- reply

Yes. Also proton "desktop" apps. It all makes me laugh.

2025-12-01 21:36:59 UTC

- reply

Couldn't agree more. It is also probably the best bang for your buck in terms of changes you can make to improve privacy.
My issue is that neither of those solutions provide sufficient anonymity. At the end of the day, all that is being done is transferring ownership of your DNS query data to someone else.
Use DNS-over-TLS-over-TOR
https://github.com/piskyscan/dns_over_tls_over_tor

2025-12-01 21:23:46 UTC

- reply

The $4T COVID "stimulus" will look like peanuts compared to this upcoming one.

2025-11-17 00:08:59 UTC

- reply

API design is an incredibly underrated skill. I'd even call it an art.

2025-11-05 22:29:57 UTC

- reply

😂

2025-11-03 12:01:27 UTC

I'm no fan of The State, but I could get behind a new law requiring any and all data collection to be opt-in only, off by default.

2025-11-03 10:59:53 UTC

- reply

Will you be renaming your product to OverCompleX Chat, too?

2025-11-03 10:54:41 UTC

- reply

Sorry, the string value to declare in your manifest is actually `PACKAGE_REPLACED` not `ACTION_PACKAGE_REPLACED

2025-11-03 10:51:59 UTC

- reply

> Wow no I wasn't aware that's crazy. Does it need migrations?
Unsure what you mean by migrations. The OS will install the new version over the old one. If you have defined in your manifest a broadcast receiver for `android.intent.action.ACTION_PACKAGE_REPLACED`, you'll receive the Intent upon completion and can restart your services and what not.
> Do you wanna help dev with bitchat btw? I could use a pro to show me the way.
I'm no pro 😂, just have been around the block once or twice. I'd love to, but have my plate full with a grant from @nprofile…grcu working on kmp-tor for the next 3 months, then back to my other project.
See: https://developer.android.com/reference/android/content/Intent#ACTION_PACKAGE_REPLACED

2025-11-03 01:39:00 UTC

UTF-8 encoding/decoding is wild. I never appreciated how elegant it is, until having to implement it. Ken Thompson is an absolute beast.

2025-11-03 01:28:57 UTC

- reply

Did you also handle the needed restart on new version being installed?
When user updates the app, Android will kill the process, then dispatch a system Intent to it after installation completion. It will receive a warm start whereby the Saved Instance State is recoverable.

2025-10-27 17:18:54 UTC

- reply

😂

2025-10-24 17:45:52 UTC

- reply

*by the government, at the behest of jews

2025-10-24 17:34:58 UTC

- reply

You realize how big of an ask that is, right?

2025-10-23 17:12:41 UTC

- reply

👀

2025-10-19 23:38:54 UTC

- reply

I feel like Luke is the living embodiment of "I told you so". Does anyone else feel this way?

2025-10-19 19:11:36 UTC

- reply

Just cut the first release of kmp-log. It's F'n awesome; check it out.
#shipcode
https://github.com/05nelsonm/kmp-log
#nevent1q…asyl

2025-10-19 17:07:37 UTC

- reply

The meme's are gonna be so good. Am excited.

2025-10-18 22:14:40 UTC

- reply

🇮🇱

2025-10-17 13:31:37 UTC

- reply

Kind of crazy it even needs to be said. Are people really that dumb?
Like, if it was bitcoin, it'd be called bitcoin and not cashu...............

2025-10-15 16:42:28 UTC

- reply

This has been quite the undertaking, but will be the only logging library you'll want to use for #kotlin Multiplatform going forward.
Very lean (no dependencies for core modules), highly configurable, sane defaults, fluent API.
#nevent1q…d2j0

2025-10-13 02:04:02 UTC

- reply

I legit just updated to v24.0.1 last weekend. Barring anything critical, won't be needing an update for at least 5 years.
Some people NGMI getting their panties all in a bunch.

2025-10-11 10:21:21 UTC

@nprofile…tmpg are you alive? I miss you.

2025-10-08 18:54:13 UTC

- reply

I could understand if it was in alpha, was a major version update and the breaking changes were due to something unavoidable, or the APIs you were using were experimental.
If none of the above, that's an immediate remove and replace with something else for me. Otherwise, you're setting yourself up for failure going forward.

2025-10-08 18:40:41 UTC

- reply

Lmao. Very innovative way to destroy your product.

2025-10-06 21:42:02 UTC

- reply

Mynymbox, Namecheap, Porkbun

2025-10-03 19:55:53 UTC

- reply

Why would you pull out of the EU market, instead of forcing them to ban you?
I, for one, would like to compare them to the CCP.

2025-10-03 19:50:06 UTC

- reply

Wen "This is my politics" commercials?

2025-10-03 15:16:23 UTC

- reply

Saying it without saying it. I love it.

2025-10-02 17:43:15 UTC

- reply

Should be merged to master tomorrow (needs documentation and more tests). Will cut a -SNAPSHOT release.

2025-10-02 17:36:20 UTC

- reply

If they haven't left NY by now, they deserve it, lol.

2025-10-02 07:43:02 UTC

The latest Android Developers Backstage podcast episode gives insight into the coming Developer Verification system.
The hubris of these people is incredible.
https://adbackstage.libsyn.com/android-developer-verification

2025-10-02 00:11:05 UTC

Have been working on adding #kotlin coroutine support to kmp-file via extension module. By and large, Js & WasmJs have been the most work figuring out how best to integrate it. Pretty happy with how the API is shaping up though.
#shipcode

2025-09-29 18:57:57 UTC

- reply

I don't even think the British Govt has those 😂

2025-09-28 15:58:25 UTC

The state of #kotlin Multiplatform logging libraries is disastrous. I'm not pulling 10 transitive dependencies into my project, man...
Looks like I'm about to embark on another side quest 😓

2025-09-28 10:47:05 UTC

- reply

The benevolent dictator model for OSS is the best model. It either fails fast, succeeds, or if the software is good enough, will splinter into many forks until a capable benevolent dictator is found.

2025-09-23 20:19:53 UTC

- reply

😂

2025-09-22 09:37:41 UTC

- reply

Love them for who they are, and who they are not. Be there for them.

2025-09-19 21:40:37 UTC

I do not understand how someone can claim to be "risk-averse", while simultaneously being anything less than 75% bitcoin. It makes no sense to me...

2025-09-19 21:23:36 UTC

- reply

I cannot even imagine being less than 75% net worth in bitcoin with all the global financial tyranny afoot. The risk associated with Fiat is way too high for my taste.

2025-09-19 13:24:48 UTC

- reply

💯
Attention to detail matters. @nprofile…fndp is a top-notch noticer.

2025-09-19 12:23:58 UTC

- reply

Release day (yes I'm shipping on a Friday)!
#shipcode
#nevent1q…nwd2

2025-09-18 00:17:29 UTC

With Kotlin 2.2.20 released last week, I've been updating all the libraries I maintain.
Finally able to implement support for use with a lower KotlinVersion so dependency consumers can exclude the transitive stdlib dependency and remain on their version of Kotlin (instead of always having to upgrade).
Aiming for next week:
• All KotlinCrypto libraries
• error
• bitops
• sponges
• random
• core
• hash
• MACs
• encoding
• immutable
• kmp-file
• kmp-process
• kmp-tor-common
• kmp-tor-resource
• kmp-tor
#shipcode
https://github.com/05nelsonm

2025-09-17 23:48:49 UTC

- reply

Logic based problem solving with a focus on minimizing complexity is feminine?

2025-09-10 17:06:21 UTC

- reply

System 76 all day every day
1/2 the price
10x the performance
Wonderful customer service
Right to repair maximalist ran company
US based company
Not an Apple product

2025-09-10 03:24:06 UTC

- reply

Have also added support for Linux Musl aarch64, i686, and x86_64.
Should be available in build-env 0.4.1, as well as the next release of kmp-tor-resource.
#shipcode
#nevent1q…m3t4

2025-09-09 01:33:47 UTC

- reply

Good. More SATs for the plebs.

2025-09-06 19:39:02 UTC

- reply

👆👆👆

2025-09-06 19:33:46 UTC

- reply

Never. kmp-tor's Android service is configured to export=false. Developers can lock down notification action Intents even further by declaring a signature permission in their manifest, and configuring the service with their custom string. Giving any app on the device access to a Service or BroadcastReceiver requires careful work given the security implications.
Most app's only need access to the tor network while they're running. kmp-tor provides that for them. Archiving kmp-tor and doing something like an Orbot app would be asinine.
The application I am working on uses kmp-tor, but yes, I have baked into the SDK's IPC protocol InetSocketAddress and UnixDomainSocketAddress information for immediate access to a tor socks/http proxy, either by requesting it via Intent, and always being sent along with the forwarded http calls.

2025-09-06 17:04:03 UTC

- reply

Why are you running tor through a VPN? That is not good and hurts the anonymity protections it provides (and also slows down the connection even more). If your VPN app allows exclusions, exclude your tor only apps.

2025-09-06 16:55:04 UTC

- reply

I would highly advise against such a setup, where every app imports the server. Logistically, that is a nightmare due to how aggressive Android is in killing processes.
Instead, I'd create a dedicated companion application users can install, and a client SDK with a thin API/protocol for developers to hook into and communicate locally via IPC.
The separation is a huge benefit for everybody (especially users) because the less code in that SDK, the more you can huck into the app and manage from 1 place. Security wise, this is beneficial too.

2025-09-05 19:02:06 UTC

- reply

Oh. Like microG?
https://github.com/microg

2025-09-05 16:53:34 UTC

- reply

It can be done in a decentralized manner; that is exactly what I'm building now... Device to device. No centralized server. Can ping any app on my phone (as long as it's registered with my app) from my laptop using terminal, curl and tor.
Server is embedded in the application that other apps on the device can register with. Http calls hit the server and get routed to the registered app.
It's a one click install for users. All overhead is pushed to developers to setup the client library for IPC, and then me to maintain the app & client library. UX is dead simple consumer wise, unlike other products that may require users to run a server at home or something.

2025-09-05 16:36:40 UTC

- reply

The beast is getting scared 😂

2025-09-05 16:32:54 UTC

- reply

Why a whole separate user? Wouldn't a work profile be better?
https://gitea.angry.im/PeterCxy/Shelter

2025-09-05 16:14:02 UTC

- reply

The coordination needed to do that is tremendous, especially when you get into versioning. A dedicated companion app is better. "Hey user, if you want to use this feature XYZ in my app, install this other app so mine can register with it".
Makes the UX dead simple for both users and developers by moving the overhead to a single application & thin IPC client library for other apps to import and configure (just like registering with Play Services Framework).

2025-09-05 15:54:43 UTC

- reply

None that would supplement google/apple services like OP desires.

2025-09-05 15:41:58 UTC

- reply

I'd love too, but am booked for the next couple of months under a sponsorship from @nprofile…grcu
to work on kmp-tor. After that, I resume work on that project.

2025-09-05 13:19:52 UTC

- reply

Signal uses the socket-io library to maintain an open socket with their backend; it's incredibly resource intensive.
Tutanota uses the SSE API for their push noticication service and does a pretty good job on their end as it relates to battery consumption.
That said, if every app were to implement their own push notification service, 5+ apps doing on a device that turns a "mobile" phone into something that always needs to be plugged in. It does not scale.
I ran a test on Android using kmp-tor and a ktor server exposed as a hidden service, pinging it every 5 min via cronjob from my server at home. Over 24h, battery consumption was about 5%-7% depending on the test device.
Working on a product right now for drop in replacement to google/apple services, but it's far from ready.

2025-09-05 12:47:51 UTC

- reply

On Android you can use a bound foreground service and keep it running into perpetuity.
Must bind/start with applicationContext
No need for requesting user opt out of battery optimizations
This is what kmp-tor does
https://kmp-tor.matthewnelson.io/library/runtime-service/io.matthewnelson.kmp.tor.runtime.service/-tor-service-config/-builder-scope/stop-service-on-task-removed.html

2025-09-05 12:19:31 UTC

- reply

Lots of slop and lost skills.

2025-09-05 12:18:30 UTC

- reply

And the new architecture is added to kmp-tor-resource
https://github.com/05nelsonm/kmp-tor-resource/pull/165

2025-09-05 11:39:18 UTC

- reply

That was quite the undertaking.
build-env 0.4.0 is out with added support for Linux Libc riscv64.
tor support for riscv64 via kmp-tor-resource will land next release.
https://github.com/05nelsonm/build-env
#shipcode
#nevent1q…3mkw

2025-09-02 03:44:22 UTC

- reply

Beautiful. Right on cue

2025-08-28 21:50:02 UTC

- reply

Story of my life.
Supporting Android is never a simple task. There's always some whacky bug on certain API levels in anything you do. I actually started going straight to using android.system.Os whenever possible to avoid the java code.
https://github.com/05nelsonm/kmp-file/blob/master/library/file/src/jvmMain/kotlin/io/matthewnelson/kmp/file/internal/fs/-FsJvmAndroid.kt

2025-08-27 18:27:44 UTC

I hate that to build C projects, you must become an expert in compiler toolchains.

2025-08-26 19:04:02 UTC

Looks like Android API 35+ has support for riscv64 coming. I wonder if the EAP clang/clang++ they shipped with NDK 27 will go stable in NDK 29?

2025-08-26 16:50:27 UTC

- reply

Never going to work based on the sole fact of it being manufactured in Spain 😂.

2025-08-26 01:02:52 UTC

Maintenance releases for all KotlinCrypto libs!
#shipcode
https://github.com/KotlinCrypto

2025-08-25 21:43:17 UTC

- reply

Gay and retarded

2025-08-24 21:50:16 UTC

- reply

I find it more likely that the redcoats will keep taking it in their awful teeth, and continue to cry about it with 0 action.
#noballs

2025-08-23 16:03:51 UTC

- reply

Many hurdles to using alternative OSs, not limited to just GrapheneOS.
One of the major ones are push notifications. Most apps are developed expecting Google APIs to be available on device. If they aren't, those apps do not work properly (sometimes even crashing). That in turn severely impacts the UX of an alternative OS.
That problem alone is a big one, and not easily solvable. The startup costs to reimplement such infrastructure and then maintain those APIs as an alternative OS is tremendous. Some workarounds do exist (such as microG, or GrapheneOS's sandboxed Play Store app), but it still takes a lot of development bandwidth.
Until there is a solid alternative to Google push notifications, app developers must roll their own custom in-app solutions (See Signal, or Tutanota). Not only does that NOT scale well, but most developers won't take that task on for a small subset of users of their product/service; and who would blame them?
And when I say "solid alternative to Google push notifications", I mean a super simple UX for users; extremely low bar, as in "Hey user of my app XZY, if you would like to receive push notifications for our app, install this other app here that handles it on-device which we can hook into." Requiring users to run some backend server at home kills any adoption of an alternative because 99.5% of people cannot, or will not, do that.

2025-08-22 21:25:32 UTC

gradle-kmp-configuration-plugin 0.5.1 is out
#shipcode
https://github.com/05nelsonm/gradle-kmp-configuration-plugin

2025-08-21 18:10:56 UTC

- reply

I'd never burn one because that's gay and retarded, but making it a crime is even more gay and retarded and against the 1A.

2025-08-21 17:54:56 UTC

- reply

That's a huge red flag for a library (also it being archived). Would avoid it like the plague 😂.

2025-08-20 18:52:12 UTC

- reply

Was finally able to push out a release of kmp-file 0.4.0 today. Was arduous getting the API right, but am very happy with the result!
#nevent1q…385z

2025-08-10 07:19:18 UTC

- reply

Sounds like a game day (board games, not vidya)!

2025-08-05 01:43:18 UTC

- reply

The only path towards greener pastures, is through the hellscape.
Bullish.
#nevent1q…tl9r

2025-08-05 01:38:29 UTC

- reply

Why withdraw though? Make them ban you so they are on par with China.