@teajaygrey@rap.social
@teajaygrey@rap.social avatar

teajaygrey

@teajaygrey@rap.social

Incarnated as a human in the area of "Yay".
What others call the Bay.
Encountered networked computers before TCP.
Email? UUCP before SMTP.
I knew the late great Doug Engelbart, personally.
Helped patch an embargoed bug in BIND 2013-4854 by CVE.
Helped restore UNIX before C.
1 of 4 skratch deejays in ThudRumble's 33.3 Club as well.
Struggles amidst these Saṃsāric rings of hell.
My 2nd language is Japanese.
Default to English, if you please.
I'm a polyglot & read & write in multiple orthographies.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

jonny, to random
@jonny@neuromatch.social avatar

Reading the Swarm whitepaper, and I know this has already been said to death, but i freaking hate how the blockchain mfs absconded with the popular conception of #p2p and got it all filthy with their libertarian ancap stuff. Bittorrent culture is like "it is good to seed way more than you download for no reason," and despite the trivial neoliberal objection that "if there is no reason then nobody will do it" it freaking WORKS.

then the blockchain people come along and are like "any imbalance in bandwidth between peers needs to be settled by MONEY and you have to RENT the ability to share anything." and it's just a preposterously bleak world. like, because they can't imagine organizing anything together with other people they have to imagine a storage infrastructure that needs an economy to keep it running.

I have no interest in permanent immutable storage of anything because that is impossible and undesirable. I want to make a digital space that exists for as long as it is needed, and sure some domains in that like research data and whatever might need to last longer than others, but the preservation of that is always a social phenomenon, and I would way rather have that come from a place of shared belief in something that is important rather than being yoked to yet another zero-sum system of wealth and debt.

teajaygrey,
@teajaygrey@rap.social avatar

@jonny Bram Cohen got a lot of things incredibly right with BitTorrent, where many other p2p implementations didn't even seem to realize there would be challenges.

My favorite aspect is perhaps the "rarest first" principle in particular. It helps maintain homeostasis and health and is a brilliant strategy to improve even adversarial seeding dynamics.

I last crossed paths with Bram at a Skinny Puppy concert, so I know he's got decent taste in music too. ;)

teajaygrey,
@teajaygrey@rap.social avatar

@jonny

Yikes regarding Swarm.

I don't really know much about Chia, maybe it was a BitTorrent Inc. driven "trying to stay in the headlines" thing? I dunno.

I did have an interview at BitTorrent Inc. years ago, and it was not a very good one. I knew a couple of people in their ops team who seemed OK, but they weren't the ones to interview me. In general, the corporate BitTorrent Inc. just felt, fishy without getting into details, so if Chia was sketchy as well, it wouldn't surprise me.

I also have no idea who uses µtorrent anymore. At least personally, I went to rtorrent in a tmux/screen session well over a decade ago and never looked back; but I guess I am among the limited sorts who prefers text/(n)curses clients to GUI stuff?

Circa 2008ish my then employer (CNEMedia) were in discussions with BitTorrent about a p2p live streaming implementation; but IIRC, their prototype: 1. was proprietary not open source 2. required a browser plugin. Both of which were non starters for us.

Fast forward to 2023 PeerTube exists & more or less does what BitTorrent Inc. promised but better.

Reading those screenshots I am still having trouble wrapping my head around people who think that immutable storage is a tractable problem that needs solving.

The Egyptian pyramids are still standing, but data from even 15 years ago suffers from bitrot, often.

I don't think computer science is going to be able to solve immutable storage, at least not in an always-on network available method. I am not even convinced it is a worthy goal.

IPFS I guess is also making efforts to do something along those lines perhaps?

Maybe a more sensible thing to strive for is: avoiding outages like MySpace which lost 12 years worth of data and never recovered. They had hundreds of millions behind them IIRC.

ajroach42, to random
@ajroach42@retro.social avatar

what would the modern computer look like if it had not been designed by the military?

teajaygrey,
@teajaygrey@rap.social avatar

@ajroach42 NLS (oNLine System) was an outgrowth of Engelbart's Augment group at SRI (Stanford Research Institute) and also: not military.

I'm attaching an image from The Engelbart Hypothesis (2009) which shows the meditative full lotus/cross legged seating in use by some.

Bill English (also part of Engelbart's Augment group) not only invented the mouse, he basically pioneered the field of ergonomics. They partnered with Herman Miller to design a lot of their prototypes.

Trivia: (learned via John Daneen [sp?] at n CoLABoration 2010 Program for the Future held at the Computer History Museum with Engelbart et al present) SRI actually terminated Doug Engelbart after he gave the "Mother of all Demos" presentation in 1968.

Eventually, J.C.R. Licklider caught wind of the presentation, and was so excited that someone had built a computer network (which he had theorized about, though doubtlessly "Lick" was also privy to SAGE) that he contacted SRI to discuss providing them with funding (which is when NLS got subsumed into [D]ARPANet).

SRI, very quietly, hired Doug back.

There's unfortunately, a lot of confusion about a lot of this stuff as much of this technology was commercialized by companies which had no stake in creating it. e.g. Pirates of Silicon Valley (1999) gets more or less all of it wrong, implying that Micro$oft and Apple stole such ideas from Xerox PARC.

PARC had a cross licensing agreement with SRI and SAIL.

Additionally, it is my understanding that Apple paid SRI licensing fees when they implemented their own version of the mouse (but they only used one button, whereas SRI/Bill English had already experimented with many variations and determined through user studies that three buttons had the fewest trade offs).

@alcinnz

teajaygrey,
@teajaygrey@rap.social avatar

@alcinnz @ajroach42 (Dr Marshall) Kirk McKusick is on record from MeetBSD 2014 that if he had a time machine, he would go back and make IPv3/48-bit addressing the winner:

https://www.youtube.com/watch?v=DEEr6dT-4uQ&t=1996s

Also with context about how NCP had 8-bit addressing.

Berkeley apparently always advocated for 48bit IP addresses, but they lost that debate. ;(

teajaygrey,
@teajaygrey@rap.social avatar

@vertigo @ajroach42 @alcinnz

x.25 was late 1970s (1976 specifically) and even in the 1990s, in practice: much slower than Ethernet (which was already 3Mbps->10Mbps in the 1970s and by the 1990s SGI had 10Gbps interconnects).

"Telcos" which in the 1970s meant: just AT&T before they had been convicted as a monopoly (1982) worked on SS7 as their own "also ran" packetized network. (SS6 and earlier prototypes existed supposedly, yet were never deployed).

My first job after graduating from University in 1999 was working as Lead Network Technician for Aeris.Net (aka Aeris Communications) in San Jose, California. They did (presumably still do?) SS7 <-> web telemetry.

SS7, circa 1999 was still: running on 56K leased lines. The only way to get more bandwidth, was adding more leased lines.

It's an "out of band" control network though, so even Aeris which had roaming contracts with every cellular carrier in North America when I worked for them (circa 1999->2001) only needed something like 6 leased lines to handle all of their SS7 signaling traffic.

Even that went the way of the dodo though, SS7 for decades now, is typically tunneled over MPLS, which in turn: is typically trunked/tunneled over TCP/IP and Ethernet. SS7 never got faster, it didn't even get fully deployed in the telco world, because of AT&T's conviction in 1982, and SS7 being relatively new, vast swaths of traditional telephony infrastructure were stuck in "RBOCs don't want to pay to upgrade that, but are happy to keep thieving grandma's monthly telephone rental fees" mode.

RBOCs were definitely still selling horrific things like FRATM (Frame Relay over ATM) into the 2000s. The last time I dealt with anything ATM was circa 2006 with T1s that were IMA (Inverse Multiplexing over ATM) from PacBell/SBC. They were: overpriced, slow as hell, and so unreliable they would fritz out if it rained, so priority sites would have 3+ and still experience total outages.

teajaygrey,
@teajaygrey@rap.social avatar

@alcinnz @ajroach42 while the military had SAGE (Semi-Automatic Ground Environment) in the 1950s and as previously written, J.C.R. Licklider funded Engelbart's Augment group at SRI (Stanford Research Institute)'s NLS (oNLine System) thus sort of military-perverting what was novel research.

In my humble opinion, no one should overlook FidoNet though.

Tom Jennings essentially created FidoNet as least cost routing over the entrenched post-convicted monopoly of AT&T circa 1984 using anarchist principals.

It was the world's LARGEST computer network until the mid 1990s when the World Wide Web finally outgrew it yet at its peak had in excess of 39,000 echos/nodes circa 1996 (and I wouldn't even begin to know how to estimate how many users at its apogee). That was also: after Tom Jennings had abandoned it himself AFAIK?

When I first encountered the World Wide Web at systems at nps.navy.mil there were approximately 3000 websites, in total. So, it really didn't have a lot going for it initially, particularly since it more or less required an overpriced NeXT or Sun workstation to even use at first.

FidoNet ran happily on all sorts of low cost personal computers with far more modest specs.

teajaygrey,
@teajaygrey@rap.social avatar

@vertigo Yeah, we presumably have vastly different experiences.

I knew someone who was going to an x.25 hacker conference in .ru circa 2006; I was boggled that such things were still in use, but for me: maintaining legacy systems is a chore.

Many people's idea of "retro" is my version of "nostalgia" and quite often, the so-called "good old days", were not good and I really never want to revisit the abuses, even in memory.

@ajroach42 @alcinnz

teajaygrey,
@teajaygrey@rap.social avatar

@vertigo I get that. I am mostly indicating that even within our shared reality, there were alternatives in existence. SS7, is more generally referred to as ANSI-41 (previously: IS-41 [where IS= "Interim Standard"]).

@ajroach42 @alcinnz

teajaygrey,
@teajaygrey@rap.social avatar

@lori I'm mostly just paraphrasing what the I recall being described as results from user group studies at SRI were on pointing devices.

The mouse wasn't the only thing they tried. Light pens predated the mouse (e.g. in Ivan Sutherland's Sketchpad: https://www.youtube.com/watch?v=495nCzxM9PI )

My understanding is that Bill English/ARC/The Augment group at SRI (Stanford Research Institute) under Doug(las) Engelbart experimented with various pointing devices, before settling on a 3 button mouse.

Some iterations had fewer (perhaps even no? I don't recall) buttons, some had as many as five buttons I seem to recall?

They even purportedly experimented with a pointing that was driven by knee movements (presumably to allow the hands to be free for other things, though perhaps this may have also been useful for accessibility much in the way there are some alternative pointing devices based upon eye tracking or breathing in more recent decades)

In SRI's studies apparently 3 buttons was considered ideal by most users?

Admittedly, they experimented with a lot of other things when it came to user input too.

For example, instead of relying solely on a QWERTY keyboard layout, NLS used a "chorded" keyboard (image attached).

Similar to playing notes on piano keys, or stenographer keyboards, multiple keys could be held simultaneously, to produce different characters.

Some years ago, an app was made available for mobile touch screen devices, by Adam Kumpf from Teague Labs but that app did not keep up & isn't in app stores anymore. (remnant: https://www.fastcompany.com/1669042/a-famous-inventors-forgotten-idea-a-one-handed-touch-screen-keyboard).

Others made an interface for the original hardware to an iPad (e.g: https://valerielandau.wordpress.com/2010/08/10/engelbart-typing-on-the-ipad-with-the-chorded-keyset/ ).

Presumably due to the versatility of the chorded keyset (typically used by the left hand) excessive buttons on the mouse (typically used by the right hand) made it such that 3 buttons seemed sufficient?

@alcinnz @ajroach42

#NLS #Mouse #ThreeButtons #ChordedKeySet #Engelbart #SRI #ComputerHistory #oNLineSystem #DougEngelbart #TipTap

trashheap, to random

deleted_by_author

    •
    teajaygrey,
    @teajaygrey@rap.social avatar

    @trashheap Usenet was viable!

    Past tense seems to be correct in that. ;)

    teajaygrey, to random
    @teajaygrey@rap.social avatar

    Weird, I can't seem to find https://bikeshed.party/@feld from here yet @feld functions OK.

    ¯_(ツ)_/¯

    ramokromok, to random

    I've added a new bunch of #amiga sites to https://boing.directory

    Let me know if you want me to add anything else

    teajaygrey,
    @teajaygrey@rap.social avatar

    @ramokromok Present whereabouts (even online) of Amiga MOD composer: Deathjester aka Magic Firesheep aka Joey Jaime aka DENIZEN AUBERON would be welcome, but I won't be holding my breath.

    Long ago, I was blessed to be contacted by Deathjester via email, and later we chatted via IRC and I was turned onto Infected Mushroom (who were also tracker composers once upon a time apparently); but have long since lost that thread. ;(

    jensimmons, to random
    @jensimmons@front-end.social avatar

    What’s new in CSS at #wwdc23 today!

    Super useful info on how to use new CSS for

    • layout
    • wide gamut color
    • selectors
    • typography

    I’m so excited I got the opportunity to make this session for you. Please take a look.

    https://developer.apple.com/videos/play/wwdc2023/10121/

    teajaygrey,
    @teajaygrey@rap.social avatar

    @jensimmons "Be sure to stick around to the very end, when I do a quick review of exiting new changes in new developer tooling."

    I think you probably intended to write: "exciting new changes"?

    Good thing ActivityPub/FediVerse/snac/Mastodon at least allow for editing!

    Sorry to be the typo bearer.

    jerry, to random

    Computers were a really bad idea.

    teajaygrey,
    @teajaygrey@rap.social avatar

    @jerry Given that the Wolfram's 2,3 system has been proven: https://www.wolframscience.com/prizes/tm23/technicaldetails.html

    It seems as if it may be more difficult to find things in existence which are not Turing complete.

    For example, DNA's TAGC exceeds the proven minimum bound for a Turing complete system.

    In other words: maybe computers aren't an idea at all? Maybe they are an emergent property.

    Em0nM4stodon, to random

    “Why do you use Signal and all this Encryption!
    Do you have anything to hide? 😡“

    Yes! I do!

    • The color of my underwear
    • My friends’ cats photos
    • My failed gym class grades
    • My first attempt at "portrait"
    • The outcome of my last meal
    • The weird mole on my left toe
    • How much I cried watching Star Trek
    • How much cheese there is in my fridge
    • My failed knitting experiment
    • The horrible poem I just wrote
    • My bank card pin number
    • My social security number
    • My main password
    • The web search history for your birthday gift

    Privacy is a Human Right! ✊

    Not sharing publicly what you do not wish to share is your right! 🔒✨

    #Privacy #Encryption #E2EE #RootForE2EE 🎉

    teajaygrey,
    @teajaygrey@rap.social avatar

    @Em0nM4stodon I am all for end to end encryption and privacy as a fundamental right which should be provided by default.

    Unfortunately, in the USA, under CALEA and the PATRIOT Act, anything which has a phone number is legislatively subject to warrantless wiretapping, Signal is no exception. I can cite at least one court transcript with Signal logs provided as evidence.

    In other words: Signal is not providing you the privacy they are telling you they provide. I am not the only individual to call them out on their BS either.

    teajaygrey, to random
    @teajaygrey@rap.social avatar

    Note to self: repology.org is not authoritative.

    I checked it this morning and thought, "oh good, 100% nothing needs updating."

    Then I checked @grunfink' stream and saw that snac 2.34 is out!

    Repology still shows 100%, but I have some work ahead of me to update the MacPorts and I won't be able to get to it as quickly as I would like.

    reina, to random

    A friend of mine once said: "foxes are dogs on catOS"

    teajaygrey,
    @teajaygrey@rap.social avatar

    @reina I do not know what catOS is.

    However, I used to really enjoy CatOS (Catalyst OS).

    Alas, as far as I know, it was EOLed by Cisco. It had its beginnings in Crescendo Communications's Catalyst Switches. Cisco acquired Crescendo, kept CatOS around for a while (last time I used it was on a 6500 series) and the newer gear they sold was mostly all IOS. ;( Cisco's IOS I used to hate, until I started encountering IOS clones, which were so bad, it made Cisco's IOS seem almost OK.

    grunfink, to fediverse
    @grunfink@comam.es avatar

    It's #SoftwareReleaseSaturday again! I'm glad to announce the release of version 2.34 of #snac, the simple, minimalistic #ActivityPub instance server written in C. It includes the following changes:

    Polls can now be created from the web interface.

    New Mastodon API features: polls are shown and can be voted on.

    The user@host identifier is now shown next to the user avatar (contributed by Haijo7). A small tweak to the default CSS was made to suit this change; please consider updating your style.css.

    Clicking on an image attachment opens it in a new tab (contributed by Haijo7).

    Bots are marked as such using an emoji (contributed by Haijo7).

    https://comam.es/what-is-snac

    If you find #snac useful, please consider buying grunfink a coffee: https://ko-fi.com/grunfink

    teajaygrey,
    @teajaygrey@rap.social avatar

    @grunfink With my apologies for the delay, this PR has been submitted:

    https://github.com/macports/macports-ports/pull/18956

    CI/build bot checks passed!

    It's up to someone else to merge it.

    #snac #MacPorts #ActivityPub #FediVerse #Mastodon #ANSIC #NoDatabaseNeeded #TotallyJavascriptーfree #NoCookiesEither #NotMuchBullShit #ANSIC #OpenSource

    mmitchell_ai, to random
    @mmitchell_ai@mastodon.social avatar

    6 (!) years ago, @timnitGebru put forward a profound idea for AI research: document your training data. To say she was treated like shit would be putting it lightly. If she had been respected like her male peers were, imagine what a different path AI could have taken.

    teajaygrey,
    @teajaygrey@rap.social avatar

    @mmitchell_ai I already consider myself a survivor of one of the earlier AI winters. As far as I'm concerned, the next one can't come quickly enough. The fallout from my vantage is already horrible.
    @timnitGebru

    grunfink, to fediverse
    @grunfink@comam.es avatar

    I'm glad to announce the release of version 2.33 of , the simple, minimalistic instance server written in C. It includes the following changes:

    Polls (in lingo, Questions) are shown and can be voted on. Creating a poll is not yet implemented, though.

    If an URL written in a post links to an image, it's converted to an attachment.

    Fixed a bug in the semaphore code that caused total hangs on OpenBSD.

    https://comam.es/what-is-snac

    If you find useful, please consider buying grunfink a coffee: https://ko-fi.com/grunfink

    teajaygrey,
    @teajaygrey@rap.social avatar

    @grunfink I submitted a PR to update MacPorts' snac to 2.33 here:

    https://github.com/macports/macports-ports/pull/18843

    CI/build bot checks passed!

    It's up to someone else to merge it.

    teajaygrey, to fediverse
    @teajaygrey@rap.social avatar

    Excellent!

    Someone updated the MacPorts' Portfile for @grunfink's snack and it wasn't me!

    https://github.com/macports/macports-ports/blob/master/net/snac/Portfile

    It's a minor change, basically there is now a PortGroup for @Codeberg now and this change utilizes it.

    So, I guess this simplifies the Portfile a little bit.

    Neat! I love it when others improve things and I don't need to lift a finger!

    grunfink, to fediverse
    @grunfink@comam.es avatar

    I'm glad to announce version 2.31 of , the simple, minimalistic instance server written in ANSI C. This time it's mostly bugfixes contributed by fellow developers:

    Fixed webfinger and curl issues, which crashed snac while trying to follow Mitra users (contributed by Poesty).

    Use named semaphores instead of anonymous ones, to make it work under MacOS (contributed by Saagar Jha).

    Fixed semaphore name conflicts (contributed by Poesty).

    Fix warning in ugly pointer substraction (contributed by Saagar Jha).

    Added user-agent to webfinger queries (contributed by Poesty).

    Mastodon API: implemented notification type filters, fixed (partially) the issue with mentions without server.

    https://comam.es/what-is-snac

    If you find useful, consider buying grunfink a coffee: https://ko-fi.com/grunfink

    teajaygrey,
    @teajaygrey@rap.social avatar

    @grunfink Awesome!

    I submitted a PR to update MacPorts' snac to 2.31 as well, fun:

    https://github.com/macports/macports-ports/pull/18651

    MacPorts specific notes:

    • fixed livecheck.
    • Removed the "use_parallel_build no" line (thanks to @saagar for the reminder to see if that was still an issue) and it passed the CI/build bots check!

    It's up to someone else to merge it.

    #snac #MacPorts #ActivityPub #FediVerse #Mastodon #opensource #ANSIC #NoDatabaseNeeded #NoCookies #Javascriptーfree #NotMuchBullshit

    ish, to random

    What better way is there to bring iSH to the fediverse than by having it introduce itself? The post you're reading right now was written and served from iSH, running an instance of @grunfink's snac :)

    teajaygrey,
    @teajaygrey@rap.social avatar

    @saagar Oh, nice one!

    Hopefully this will be included in 2.31?

    As far as the disabled parallel builds question?

    If I recall correctly without it, the CI/build bots may have failed? That was a simple fix when conferring with other MacPorts contributors a while ago as far as their suggestion.

    @grunfink @DavidBlue @ish

    teajaygrey,
    @teajaygrey@rap.social avatar

    @saagar

    I just tested removing the "use_parallel_build no" line from a Portfile locally.
    Still seems to build for me cleanly!

    I know the MacPorts' CI/build bots have changed (e.g. they now test against macOS 13) since I first submitted snac for fun.
    So the next time I submit a PR, I'll see if that alteration will run?

    @grunfink @DavidBlue @ish

    teajaygrey,
    @teajaygrey@rap.social avatar

    @grunfink

    Groovy!

    Thank you for your diligent contributions, which reminds me, I want to donate to your Ko-Fi.

    @DavidBlue @ish @saagar

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • megavids
  • osvaldo12
  • DreamBathrooms
  • mdbf
  • InstantRegret
  • ngwrru68w68
  • magazineikmin
  • everett
  • thenastyranch
  • Youngstown
  • slotface
  • cisconetworking
  • kavyap
  • ethstaker
  • JUstTest
  • modclub
  • GTA5RPClips
  • khanakhh
  • tacticalgear
  • Durango
  • rosin
  • normalnudes
  • Leos
  • provamag3
  • tester
  • cubers
  • anitta
  • lostlight
  • All magazines
    •