It's actually pretty interesting and amusing to see all the ideas here. I don't think i'd have ever guessed it started life as a fan controller. Still, sign me up for that gasoline powered indoor fan. Hells yeah!
Thank you. Hope you don't mind the different topic with this video, it's not exactly SGI but I think it's interesting and the stats on the original Pegasus X video seem to agree.
The stories behind what spawns our passions are well worth telling, as much as they are worth hearing, so don't worry :D. I'm excited for the next parts.
Yes, Amity Automotive is discussed in Episode 2 but that's not out yet. That's not even the most ridiculous thing he did. He wanted to build a balloon to go to space, and a sentient AI.
Dodoid Whoa. What the fuck? I mean I understand this and I could make a smart electric fan that heats (cooling would either too inefficent or too complicated and big) but CARS AND A SENTIENT AI!?!? Actually it'd be pretty easy to make an electric vehicle out of 2 old bikes and strap on a electric motor and a few car batteries to run it.
Jake was a strange guy. I left him be and worked on the projects that I thought had a chance. His car plans weren't electric though. He wanted to build gasoline engines. Had he ever so much as looked at an exploded diagram of one? No. He certainly hadn't.
Google Is Fucking SJW Scum Guy: "Hey I made a paper fan" Jake: "hold my beer..." 2 months later Jake: "so the teleportation system didn't work out..." guy: "what..."
Jake did not have a huge degree of respect for intellectual property law. The Pegasus X logo was somewhat of a ripoff of the Xbox logo, and we were never totally sure of the licensing of music on the Pegasus Shop.
Yes, Azathoth plays "John Smith", and is an IRL friend of mine. "Jake" has a RUclips channel, but I don't believe he watches mine. I don't talk to him very often at all (maybe only a couple of times in the two or so years), but he now laughs at Pegasus just as much as I do.
Dodoid it's nice that you can look back and laugh. It seems like a really good idea, and if it was still a thing, it could be marketed as a smart home dashboard. Especially if you integrated Amazon Alexa or Mycroft.
Rittybox was a name I had held onto for a long time for a number of computer-related things (I have evidence of myself using the name when I was five and possibly four years old). By the time of the Pegasus X's start of development in early to mid 2014, it had been reasonably serious for at least a few years, and had a decent few members, though most were almost entirely uninvolved (we even had an official term, "inactive member"). It was highly bureaucratic, and spent most of its time reorganizing itself and debating whether to give up projects. Rittybox tended to get bogged down in this, but we did still have a few reasonably capable members, and a few ongoing but ultimately pointless projects. Pegasus was a lot more nimble, and handled this significantly better, actually managing to release their products even if they weren't any good. Pegasus was much better at "failing fast", whereas Rittybox lumbered along complaining about itself for years before and years after. I think perhaps the most indicative thing of the problems Rittybox had is that, on the present website, the only remaining page is for the organizational member area. If that doesn't scream "bureaucratic and ineffective to the very end", I don't know what does.
Earlier on, we did not have an official stance on licensing (it doesn't exactly matter when you don't have any products to license). Once it started to look like we might actually release something (never happened, but it could have), I believe we planned to use the GPL. The idea was not bad at all, and some of what we worked on was almost sane, but we just ended up destroying any chance of success the second we prioritized internal infrastructure over products. I will repeat this again: Pegasus did this better, and had Rittybox been run more like it had, we might see one or even both companies/groups surviving to this day (Rittybox acquired what remained of Pegasus). As it stands, I've chosen the most promising Rittybox project (Kronos) and the two core members (myself and Shroomy) and spun it off into its own thing. Thankfully, I'm reasonably sure we've learned from our (Rittybox's) mistakes. kronosproject.org.
We had a lot of ideas, some potentially good, some absolutely dumb. Among the better were Protop/Project Omnipod, Project Kronos, and, towards the very end, Project Type. Omnipod/Protop was among our earlier "serious" projects, and the only hardware project I would deem at least half-sane. It was to be a hemispherical desktop computer, using a very small (likely sub-mini-ITX) motherboard and featuring a circular display on the front known as the Rittybox Eye. Plans got reasonably far, and I am reasonably convinced that, given enough funding, we could actually have built it, but unfortunately, the case design was really the only novel part of the design, and it would prove too expensive to continue development, especially before the first makerspaces with 3D printers opened in Ottawa. Kronos was our only "released" product, used on the Pegasus X. It was a GNU/Linux distribution intended to make heavy use of a web application layer based on WebKit and Apache, with applications intended to use PHP for their "local" needs (eg. data storage). While earlier versions (such as the very early test version mentioned in this video, and even, to an only slightly lesser extent, the latest Pegasus X software, version 1.4.1) are full of security holes and bugs, some small pieces of its code, as well as some of the general concepts, are actually rather workable, and this is the reason why it is the only Rittybox project to see continued development today. I have already replied to one of your comments to mention this, but Kronos is now developed independently of the mostly-dead Rittybox by a new organization I have set up known as the Kronos Project. We are planning a reworked 2.0 release for some time in 2018, with the first public beta launching extremely soon. Type is by far the simplest and newest project I'm talking about here, but it was intended to be a sort of modern command line, a text-based method of input with visual output and some degree of "intelligence". It could perhaps be considered as conceptually similar to Google's "cards" you get at the side of a search, but the focus was slightly different, more intended to perform traditional personal computing tasks than to obtain information from the web. Not exactly AI, but it was intended to be able to work out what you wanted it to do with an input slightly less structured than a traditional CLI command. This never got very far, as while it was absolutely feasible, Rittybox was basically dead by the time it was even conceptualized.
I was going to give a big long response on why attempting to duplicate, or simply using Kronos 1.0 is a bad idea in a number of ways, but perhaps it's better to ask, what are you trying to do? The Pegasus X is not going to be how you get there, but I hope that at least it can serve as a lesson in what not to do. What is your project?
Dodoid i am attempting to construct a pda like system with a custom lightweight os. All i find on the internet is showing you how to make a linux distro. I just want something basic built for the form factor that me (and a couple friends) could build and modify. You are the first person i found who made something similar to what i wanted to make. If you would like to know more i will link you to my schematics once i polish them up. I could also send them to you via email or any other means of digital contact.
Kronos is built on GNU/Linux (both versions 1.x and 2.x), so if you're worried that GNU/Linux is too heavy, I can't really help you. I presume when you say "custom" you mean built from scratch? There is a lot of complexity there, and if you would like a glimpse into just how much complexity that is (even for a "lightweight" OS), it might be worth thinking about just how much you want your OS to do. You say lightweight, but (and this is assuming it's on PC, though other systems would likely present equal difficulty) you probably still want memory protection, preemptive multitasking, the ability to draw pixels using, at minimum, VESA, some method for IPC, drivers for the devices you plan your hardware to have, support for networking (that means drivers, but also a working TCP/IP stack and the utilities to use it), support for a mouse or touch input, support for USB or at least a serial port, and the SMP support for the ability to scale to multiple processors (this includes multiple processor cores). This is before you even add any actual, user-facing features. These are just a few of the parts that have to be there before you can even think about that stuff. Also, while the examples I give are for a PC, no, you can't avoid them by "going to Pi". ARM SBCs still need to do this stuff, and in many cases you'll find them even less well documented, particularly at the low-level, due to proprietary firmware and a lack of information. Want graphics? Have fun, VESA doesn't exist and you're dealing with some awful proprietary Broadcom graphics chipset whose developers won't give you the time of day. Network? Well, that's behind a USB ethernet controller (no, you don't have a datasheet, what are you crazy), on a USB switch (which, chances are, will turn out to be from some Chinese company that hasn't existed in 11 years and only published docs under NDA), connected to the SoC's USB port, which, oh wait, that's a blob in the Linux kernel, better go try to convince the SoC maker to write a blob for YOUR OS. I am not even an expert in this stuff. I have never written my own kernel, this is just the very base-level stuff that springs to mind. I'm sure that, were you to go talk to some Linux kernel developer, they would be able to raise a lot more problems with trying to "just write your own" than I have here. If you want to build a mobile computing device, and see the Pegasus X as an example, then you might consider Kronos 2.0, the successor to the Kronos 1.x run by the Pegasus X. It's rewritten, almost completely from the ground-up, with a focus on fixing what the Pegasus X's software got so wrong. There are no releases yet, but you can check out the website at kronosproject.org. If you have any more questions, feel free to ask.
Dodoid thank you for th7e reality flash i will stick to just making a gui and then once i have more experience i will go at this from a new angle. Thank you for all of the help you have provided today. Have a happy new year.
If you just want to make a GUI on top of a reasonably lightweight OS that will try to figure out the rest for you, Kronos 2.0 allows pretty much exactly that. Technically, Kronos 1.0 does too, but as I've mentioned before, it's unfortunately a terrible pile of garbage, so use the new Kronos. We plan to put out our first public developer release, 2.0-GDR4P, in only a few weeks, so if you wanted to get started with GUI development on the Kronos platform, then would be a great time to get going.
Nice. Looking forward to part 2!
It's actually pretty interesting and amusing to see all the ideas here. I don't think i'd have ever guessed it started life as a fan controller.
Still, sign me up for that gasoline powered indoor fan. Hells yeah!
GloomyJD Gasoline smells fucking great. Fuck yeah!
Yo dude. I'm loving your videos, keep it up.
But is it true that making a cooling fan requires an understanding of aerodynamics?
Recently stumbled onto your channel from your SGI history video. Very excellent stuff Dodoid. Quality content and insightful presentation.
Thank you. Hope you don't mind the different topic with this video, it's not exactly SGI but I think it's interesting and the stats on the original Pegasus X video seem to agree.
The stories behind what spawns our passions are well worth telling, as much as they are worth hearing, so don't worry :D. I'm excited for the next parts.
Jake also wanted to branch out into the auto industry. no joke.
Yes, Amity Automotive is discussed in Episode 2 but that's not out yet. That's not even the most ridiculous thing he did. He wanted to build a balloon to go to space, and a sentient AI.
Dodoid he wanted to reproduce his kind and use the balloon to return to his home planet
Dodoid Whoa. What the fuck? I mean I understand this and I could make a smart electric fan that heats (cooling would either too inefficent or too complicated and big) but CARS AND A SENTIENT AI!?!? Actually it'd be pretty easy to make an electric vehicle out of 2 old bikes and strap on a electric motor and a few car batteries to run it.
Jake was a strange guy. I left him be and worked on the projects that I thought had a chance. His car plans weren't electric though. He wanted to build gasoline engines. Had he ever so much as looked at an exploded diagram of one? No. He certainly hadn't.
Google Is Fucking SJW Scum
Guy: "Hey I made a paper fan"
Jake: "hold my beer..."
2 months later
Jake: "so the teleportation system didn't work out..."
guy: "what..."
Actually what confused me most is that pegasus was actually an earlier name for that Asus company. Was kinda hoping for an asus history.
Jake did not have a huge degree of respect for intellectual property law. The Pegasus X logo was somewhat of a ripoff of the Xbox logo, and we were never totally sure of the licensing of music on the Pegasus Shop.
Easy breeze is trash. pegusus is a real man's smart fan! Jake is a true hero and he needs all the money!
Azathoth I am guessing you are "Jake"?
Matt Green No but I know the man. I was on this channel before in an old video parodying the Nintendo switch launch
Azathoth John Smith?
Yes, Azathoth plays "John Smith", and is an IRL friend of mine. "Jake" has a RUclips channel, but I don't believe he watches mine. I don't talk to him very often at all (maybe only a couple of times in the two or so years), but he now laughs at Pegasus just as much as I do.
Dodoid it's nice that you can look back and laugh. It seems like a really good idea, and if it was still a thing, it could be marketed as a smart home dashboard. Especially if you integrated Amazon Alexa or Mycroft.
I really want the Pegasus-R now.
Hey, lemme turn on the fan...
(turns key) BEEEEEEP-
V R O O O O O O O O M
@Dodoid
And we can't forget about the supercharged Pegasus-V.
Browningate Did not expect you here.
@Google-is
Actually, I kind of did expect him, even if it was more of an anticipation than an expectation. Nobody is obligated to reply to comments.
what the hell did you just do to my beautiful and 100% correct username!??!?!?!?
Where's part 2 ?
What the hell, Dragos Criste is working on this too? Wow.
Nice PFP.
I would like to know more about Rittybox. I found the website, it didn't tell me much.
Rittybox was a name I had held onto for a long time for a number of computer-related things (I have evidence of myself using the name when I was five and possibly four years old). By the time of the Pegasus X's start of development in early to mid 2014, it had been reasonably serious for at least a few years, and had a decent few members, though most were almost entirely uninvolved (we even had an official term, "inactive member"). It was highly bureaucratic, and spent most of its time reorganizing itself and debating whether to give up projects. Rittybox tended to get bogged down in this, but we did still have a few reasonably capable members, and a few ongoing but ultimately pointless projects. Pegasus was a lot more nimble, and handled this significantly better, actually managing to release their products even if they weren't any good. Pegasus was much better at "failing fast", whereas Rittybox lumbered along complaining about itself for years before and years after.
I think perhaps the most indicative thing of the problems Rittybox had is that, on the present website, the only remaining page is for the organizational member area. If that doesn't scream "bureaucratic and ineffective to the very end", I don't know what does.
Dodoid That sounds like what I could gather. Was it like open source stuff? It sounds like a cool idea, but I know what you mean about organization.
Earlier on, we did not have an official stance on licensing (it doesn't exactly matter when you don't have any products to license). Once it started to look like we might actually release something (never happened, but it could have), I believe we planned to use the GPL. The idea was not bad at all, and some of what we worked on was almost sane, but we just ended up destroying any chance of success the second we prioritized internal infrastructure over products. I will repeat this again: Pegasus did this better, and had Rittybox been run more like it had, we might see one or even both companies/groups surviving to this day (Rittybox acquired what remained of Pegasus). As it stands, I've chosen the most promising Rittybox project (Kronos) and the two core members (myself and Shroomy) and spun it off into its own thing. Thankfully, I'm reasonably sure we've learned from our (Rittybox's) mistakes. kronosproject.org.
Dodoid What had you planned to release? I am glad you all learned from your mistakes, it is very helpful.
We had a lot of ideas, some potentially good, some absolutely dumb. Among the better were Protop/Project Omnipod, Project Kronos, and, towards the very end, Project Type.
Omnipod/Protop was among our earlier "serious" projects, and the only hardware project I would deem at least half-sane. It was to be a hemispherical desktop computer, using a very small (likely sub-mini-ITX) motherboard and featuring a circular display on the front known as the Rittybox Eye. Plans got reasonably far, and I am reasonably convinced that, given enough funding, we could actually have built it, but unfortunately, the case design was really the only novel part of the design, and it would prove too expensive to continue development, especially before the first makerspaces with 3D printers opened in Ottawa.
Kronos was our only "released" product, used on the Pegasus X. It was a GNU/Linux distribution intended to make heavy use of a web application layer based on WebKit and Apache, with applications intended to use PHP for their "local" needs (eg. data storage). While earlier versions (such as the very early test version mentioned in this video, and even, to an only slightly lesser extent, the latest Pegasus X software, version 1.4.1) are full of security holes and bugs, some small pieces of its code, as well as some of the general concepts, are actually rather workable, and this is the reason why it is the only Rittybox project to see continued development today. I have already replied to one of your comments to mention this, but Kronos is now developed independently of the mostly-dead Rittybox by a new organization I have set up known as the Kronos Project. We are planning a reworked 2.0 release for some time in 2018, with the first public beta launching extremely soon.
Type is by far the simplest and newest project I'm talking about here, but it was intended to be a sort of modern command line, a text-based method of input with visual output and some degree of "intelligence". It could perhaps be considered as conceptually similar to Google's "cards" you get at the side of a search, but the focus was slightly different, more intended to perform traditional personal computing tasks than to obtain information from the web. Not exactly AI, but it was intended to be able to work out what you wanted it to do with an input slightly less structured than a traditional CLI command. This never got very far, as while it was absolutely feasible, Rittybox was basically dead by the time it was even conceptualized.
I just really want to bring him to a barber to clean his whole face up lol
How would someone develop a os similar to the one youve shown
I was going to give a big long response on why attempting to duplicate, or simply using Kronos 1.0 is a bad idea in a number of ways, but perhaps it's better to ask, what are you trying to do? The Pegasus X is not going to be how you get there, but I hope that at least it can serve as a lesson in what not to do. What is your project?
Dodoid i am attempting to construct a pda like system with a custom lightweight os. All i find on the internet is showing you how to make a linux distro. I just want something basic built for the form factor that me (and a couple friends) could build and modify. You are the first person i found who made something similar to what i wanted to make. If you would like to know more i will link you to my schematics once i polish them up. I could also send them to you via email or any other means of digital contact.
Kronos is built on GNU/Linux (both versions 1.x and 2.x), so if you're worried that GNU/Linux is too heavy, I can't really help you. I presume when you say "custom" you mean built from scratch? There is a lot of complexity there, and if you would like a glimpse into just how much complexity that is (even for a "lightweight" OS), it might be worth thinking about just how much you want your OS to do.
You say lightweight, but (and this is assuming it's on PC, though other systems would likely present equal difficulty) you probably still want memory protection, preemptive multitasking, the ability to draw pixels using, at minimum, VESA, some method for IPC, drivers for the devices you plan your hardware to have, support for networking (that means drivers, but also a working TCP/IP stack and the utilities to use it), support for a mouse or touch input, support for USB or at least a serial port, and the SMP support for the ability to scale to multiple processors (this includes multiple processor cores).
This is before you even add any actual, user-facing features. These are just a few of the parts that have to be there before you can even think about that stuff. Also, while the examples I give are for a PC, no, you can't avoid them by "going to Pi". ARM SBCs still need to do this stuff, and in many cases you'll find them even less well documented, particularly at the low-level, due to proprietary firmware and a lack of information. Want graphics? Have fun, VESA doesn't exist and you're dealing with some awful proprietary Broadcom graphics chipset whose developers won't give you the time of day. Network? Well, that's behind a USB ethernet controller (no, you don't have a datasheet, what are you crazy), on a USB switch (which, chances are, will turn out to be from some Chinese company that hasn't existed in 11 years and only published docs under NDA), connected to the SoC's USB port, which, oh wait, that's a blob in the Linux kernel, better go try to convince the SoC maker to write a blob for YOUR OS.
I am not even an expert in this stuff. I have never written my own kernel, this is just the very base-level stuff that springs to mind. I'm sure that, were you to go talk to some Linux kernel developer, they would be able to raise a lot more problems with trying to "just write your own" than I have here.
If you want to build a mobile computing device, and see the Pegasus X as an example, then you might consider Kronos 2.0, the successor to the Kronos 1.x run by the Pegasus X. It's rewritten, almost completely from the ground-up, with a focus on fixing what the Pegasus X's software got so wrong. There are no releases yet, but you can check out the website at kronosproject.org.
If you have any more questions, feel free to ask.
Dodoid thank you for th7e reality flash i will stick to just making a gui and then once i have more experience i will go at this from a new angle. Thank you for all of the help you have provided today. Have a happy new year.
If you just want to make a GUI on top of a reasonably lightweight OS that will try to figure out the rest for you, Kronos 2.0 allows pretty much exactly that. Technically, Kronos 1.0 does too, but as I've mentioned before, it's unfortunately a terrible pile of garbage, so use the new Kronos. We plan to put out our first public developer release, 2.0-GDR4P, in only a few weeks, so if you wanted to get started with GUI development on the Kronos platform, then would be a great time to get going.
50th like.
Can I donate Dogecoin to you
Cut your hair.