Back
[01:02:47] <rue_house> <rue_house>
http://ruemohr.org/~ircjunk/tutorials/elex/p1080336.jpg
[01:02:48] <rue_house> <rue_house>
http://ruemohr.org/~ircjunk/tutorials/elex/p1080337.jpg
[01:02:51] <rue_house> hehe
[01:05:46] <z64555> what you solder on there?
[01:27:06] <rue_shop4> a fet
[01:27:22] <rue_shop4> one that just stripped the gears in a standard servo running on 5V
[01:27:42] <rue_shop4> (operating the motor directly)
[01:28:06] <rue_shop4> SI2318
[01:41:40] <z64555> it stripped the gears? I think you made it a bit too powerful
[01:41:54] <z64555> but awesome, nevertheless
[01:42:36] <rue_shop4> it stripped the gears at 5V
[01:49:40] * z64555 wonders if he should do some soldering tonight
[02:10:34] <RR101> 2:41:17 AMRR101good night
[02:10:34] <RR101> 2:42:04 AMRR101I'm looking for an expert in procedural algorithms.
[02:12:45] <RR101> cameronfr: hi
[02:18:15] <RR101> I need help of a good creator of algorithms.
[02:18:22] <z64555> <.<
[02:18:24] <z64555> >.>
[02:18:37] <z64555> well, I think I'm good enough for some things
[02:18:51] <z64555> Something about procedural algorithms?
[02:24:32] <RR101> z64555: yes. I'm working hard on a new watch, and I have a crippling design work, so I require help to generate a procedural generation algorithm. 2D designs are geometric shapes.
[02:29:46] <RR101> z64555:
[02:29:46] <RR101> There are designs that are impossible to be performed by a computer, I'm sure. I have design principles I want to see if they are possible to emulate an algorithm. I require an expert who can clarify these doubts.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:11] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:12] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:12] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:13] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:13] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:14] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:14] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:15] <RR101> I recently saw a program created by cameron, which lets you create random shapes in a 2D plane, and shows me that it is possible to build a procedural generator SF.
[02:33:54] <RR101> I thought I was not hitting well the text.
[02:48:46] <z64555> sorry about that. what's with the wall of text?
[02:49:55] <z64555> you might have problems with connectivity on your end
[02:51:11] <z64555> the only "impossible" designs that I know of are those that rely on visual inspection. And even then, if you code the computer just right you can have it do the design anyway
[02:57:04] <RR101> z64555:
http://prntscr.com/c6arjn Such a design is possible?
[02:57:54] <z64555> oh yeah, most likely
[02:58:48] <z64555> you have graphic elements which are regular in shape, and are concentric
[02:58:59] <RR101> z64555: You talk seriously is just sarcasm?
[02:59:07] <z64555> I'm talking seriously
[02:59:46] <z64555> Have you looked into fractals? they're one of the early forms of procedural algorithms
[03:00:05] <RR101> z64555: Well, I'll show a more human design.
[03:01:18] <RR101> z64555:
http://prntscr.com/c6at9z
[03:02:04] <z64555> you don't have to call my name so often. I can see your messages just fine. :)
[03:02:28] <z64555> that one is still regular, and can be made by a machine
[03:02:57] <RR101> z64555: Yes, you require someone knows how to create designs based on a set of own rules.
[03:03:01] <z64555> it makes primary use of drawing circles, and then drawing lines to split them up
[03:03:22] <RR101> oh sorry
[03:04:27] <z64555> If you're at a college, hook up with an art major that is taking courses on 2d graphic design
[03:05:25] <z64555> they often have rules in their courses that outline the methods that are perceived as aesthetically appeasing, and "proven" in the commercial market as successful
[03:05:44] <z64555> most notably is the concept of a grid and alignment of elements
[03:06:30] <RR101> The design Follows a pattern, you can see by the color, and the quantity of forms.
[03:06:33] <z64555> the latest exampe, for instance, has at least 2 or 3 grids
[03:07:33] <z64555> right, and if it has a pattern, it is following some sort of rules. And since it is following rules, it can be done by a computer
[03:12:31] <RR101> z64555: Great! how much you can cost a generator designs 2D in vector format, which generates random designs following these rules
[03:13:34] <RR101> The cost of building.
[03:15:13] <RR101> I currently have 6 design rules that can serve to shape the algorithm.
[03:16:39] <RR101> Also, I have developed the algorithm the app running on Android.
[03:18:10] <z64555> cost in terms of what?
[03:18:22] <z64555> man hours? CPU clock cycles/memory?
[03:18:46] <RR101> http://prntscr.com/c6az9j So designs are possible?
[03:18:48] <z64555> man hours, I have no idea how much work it would take to create and tune
[03:19:13] <z64555> CPU clock cycles/memory, it would likely be fairly intesive, depending on how complex the algorithm turns out to be
[03:19:15] <RR101> hours
[03:19:23] <RR101> hours working
[03:19:52] <z64555> hm, I'm thinking at least a college semester's worth for a 2 to 3 man team
[03:20:13] <z64555> I'm not sure how fast a professional coding outfit would do
[03:20:20] <z64555> haven't seen on in action yet
[03:21:07] <z64555> that latest design might be possible, but you'd have to feed a seed image in order for the algorithm to fit the shapes into
[03:21:11] <z64555> *last
[03:21:40] <z64555> the shapes are regular enough, but the fact that it forms a tiger face means some special consideration is needed
[03:21:49] <RR101> You place the price per hour and honesty, to show me the hours they actually you work.
[03:23:04] <z64555> I didn't understand that last one
[03:24:18] <z64555> *that last message
[03:24:33] <RR101> They may be only abstract, but with beautiful curves, a balance between beauty and what the rules allow.
[03:26:19] <RR101> You can spend 30 hours at work, and then tell me I work 40, and I have no way to know?
[03:28:45] <RR101> I require a confidentiality agreement before giving the necessary information.
[03:29:23] <RR101> We can communicate by private?
[03:29:35] <z64555> sorry, not interested
[03:30:06] <z64555> you may try some of the other channels on this IRC network
[03:30:40] <z64555> I think there's at least one channel dedicated to programming, you could search for journeymen and coders-for-hire there
[03:30:49] <RR101> Can you suggest me another person of good level in the area? Please,
[03:32:04] <RR101> There is a continuous employment, consider a freelance job.
[03:32:23] <z64555> I know of nobody currently available for such a job. sorry
[03:33:04] <RR101> Even I do not get to the person who really likes complex challenges.
[03:33:49] <deshipu> recruiting random people on the internet via IRC?
[03:33:51] <z64555> challenges, yes. employment for said challenge. not always. :)
[03:34:22] <deshipu> what could possibly go wrong
[03:34:23] <z64555> deshipu: yeah, it's a longshot, but in this day and age, wierder agreements have happened
[03:35:38] <RR101> The best is achieved in these dark places. deshipu
[03:36:25] <z64555> I highly recommend finding somebody that can speak your native tongue to aide you, at the very least as a translator or mediator
[03:37:08] <RR101> It is certainly a challenge, even you do not know the context of the project.
[03:37:28] <z64555> :)
[03:37:47] <RR101> Thank you for your suggestion, thanks for your time. ;)
[03:38:06] <deshipu> communication problems can make two well-meaning people be very disappointed
[03:41:05] <RR101> deshipu: For a reason why you intervened, you tell me, I'll be gone soon.
[03:45:05] * z64555 frowns at the knife he was using to strip copper cable
[03:45:05] <deshipu> RR101: intervened? this is a public channel
[03:45:22] <z64555> perhaps he meant interjected
[03:45:36] <deshipu> still that suggests I'm not welcome
[03:45:54] <deshipu> oh well
[03:46:10] <veverak> lal
[03:46:36] <RR101> z64555: A that you mean by mention a seed, to achieve designs like the tiger?
[03:46:39] <deshipu> by the way, did I spam this channel already with this?
https://hackaday.io/contest/12641-esp8266-micropython-contest
[03:47:08] <veverak> RR101: learn english, hard to get job with it otherwise
[03:47:24] <z64555> RR101: Yes, designs that resemble any particular shape that is not a normal geometric shape (circles, polygons, etc.) you'll need a seed
[03:47:52] <z64555> er, the normal geometric shape being the circles, polygons, etc.
[03:47:56] <z64555> sorry
[03:48:38] <veverak> deshipu: coool
[03:48:40] <z64555> Designs such as the tiger face, a person's face, and possibly a few plants and flowers
[03:48:58] <deshipu> veverak: espressif just said they will give 5 esp32s as prizes too
[03:49:04] <RR101> A seed, referred?
[03:49:34] <z64555> will require a seed. They are too complex for an alogrithm to generate within a reasonable timeframe for commercial computers
[03:49:40] <veverak> deshipu: cool
[03:49:41] <z64555> Yes, a seed. a reference image
[03:51:45] <z64555> It is theoritically possible to achieve such designs even without a seed, but you would need a rather impressive genetic algorithm
[03:51:55] <RR101> Now if I understand, and the reference image can be either monochromatic or single vector and formats?
[03:52:34] <z64555> for the purpose of your algorithm, the reference image would be a monochomratic image or a greyscale image
[03:52:44] <RR101> I'll show you something.
[03:52:54] * z64555 sighs
[03:53:22] <veverak> have you ever avoided running your code?
[03:53:27] * veverak feels like it for last two days
[03:54:19] <z64555> newb. I have avoiding running a segment of code for years!
[03:54:28] <veverak> daaamn
[03:54:32] <z64555> a very /particular/ segment of code
[03:54:48] <deshipu> veverak: recently I wrote some code for hardware I didn't have again, and it worked when the hardware arrived
[03:55:08] <veverak> wow
[03:55:21] <z64555> deshipu: congrats!
[03:55:33] * veverak evven finished all docstring and refurbished the code to be perfect by pep8
[03:55:36] <z64555> that's how you know you're starting to think like a machine
[03:57:21] <RR101> http://prntscr.com/c6bdzd
[04:00:35] <z64555> All doable except for maybe ISU5, which would require a reference
[04:00:52] <RR101> deshipu: Sorry if it's not my subject, but it's amazing if it's true what you just said.
[04:01:27] <z64555> specifically those horns at not made from simple geometric shapes
[04:03:16] <RR101> It correct what you tell, if it's worth investing in a procedural algorithm.
[04:05:14] <RR101> Currently I have hundreds of assets made by humans. The cost is mainly time. If you say you can, it is because you can build it.
[04:06:22] <deshipu> the problem with procedurally generated designs is that you still have to pick the ones that look good, and that still takes time
[04:07:01] <RR101> The goal is to have 10,000 designs ready before the commercial release.
[04:08:16] <RR101> deshipu:
[04:08:16] <RR101> It is less than the combined effort all design.
[04:08:44] <z64555> nah.
[04:09:11] <z64555> to be honest, it would probably be more economical to go wtih a team of graphic designers in such a short time frame.
[04:09:40] <z64555> Even though you could get something "working" within a year, it will likely take at least another year to get a professional looking design
[04:10:02] <z64555> from a computer algorithm
[04:10:28] <z64555> Since you would have to fine tune things, make additional cases of which elements should be used, and where they should not be used
[04:10:47] <RR101> Currently each design required to create the final version 10-15 hours.
[04:15:23] <z64555> which would be the equivalent of about 50 graphic designers working 40 hour work weeks for one year
[04:15:56] <RR101> Wow, yes.
[04:16:27] <z64555> and that's assuming they're capable of making a unique enough design from their past works and unique enough from each other
[04:17:16] <deshipu> z64555: you could make a penis image generator :D
[04:17:19] <z64555> which is another factor in your algorithm's design. uniqueness
[04:17:48] <deshipu> I hit that problem with a name generator for an rpg
[04:17:53] <z64555> deshipu: now why would I do that. there's already plenty of cocks in this world
[04:18:01] <deshipu> it would generate words that are not suitable for this kind of game
[04:18:26] <deshipu> I think they had that problem with a password generator at some large website too
[04:18:35] <deshipu> people would get offended by the generated passwords
[04:18:39] <z64555> Assy McAssBlaster the Third
[04:18:54] <deshipu> the problem is, computers can't tell when something they generated is offensive
[04:19:04] <z64555> unless you tell em
[04:19:11] <z64555> which is an additional rule set
[04:19:26] <z64555> And then you get into the implied meaning, which is another level of complexity
[04:19:31] <deshipu> not really, it's cultural associatoons
[04:19:38] <deshipu> associations
[04:19:55] <z64555> at which point you'd probably have a good speach interpreter and would be better suited for making a robot that can tell funny jokes at parties
[04:19:59] <deshipu> you would have to literally bring up your computer in your society
[04:20:22] <deshipu> right
[04:20:26] <z64555> :)
[04:20:50] <deshipu> of course it's a very interesting problem and probably a good job opportunity
[04:20:58] <deshipu> not going to finish any time soon
[04:21:48] <z64555> oh yeah.
[04:22:05] <z64555> Possible. yes. Practical. no.
[04:22:19] <deshipu> wasn't that how GNU was created?
[04:22:53] <deshipu> they paid Stallman for working on AI, and he wrote an operating system instead?
[04:22:54] <z64555> I thought GNU was a collective hobby
[04:23:02] <z64555> oh, lol
[04:23:07] <RR101> I have seen that there have been procedural metamorphic 3D figures, it is safe simpler 2D.
[04:23:52] <z64555> RR101: That's not safe to assume. There are many tricks you can do in 2D that can be as complex as 3D
[04:24:15] <z64555> You really need to talk to a professional graphic designer about this
[04:26:16] <RR101> I'm a graphic designer, programmer would be programming being now the algorithm.
[04:26:34] <deshipu> hehe, touche
[04:27:56] <RR101> You saw that the designs have from 40-42 forms in total.
[04:29:35] <RR101> That's the primary rule.
[04:30:12] <RR101> Then the 5 colors, green, red, blue, cyan and yellow.
[04:30:57] <RR101> It is the second rule.
[04:31:31] <RR101> After the forms are grouped in the colors similarly.
[04:31:49] <RR101> That's the third rule.
[04:32:57] <RR101> After the forms are separated to X apart distance.
[04:33:15] <RR101> That's the fourth rule.
[04:34:13] <RR101> After the forms are grouped in a circle holding the largest possible space area.
[04:34:33] <RR101> It is the fifth rule.
[04:34:49] <z64555> don't forget that you can do additive as well as subtractive operations
[04:35:14] <z64555> in fact, most of your example images were a combination of adds and subtracts of geometric shapes
[04:35:23] <RR101> The sixth rule can not say.
[04:36:27] <z64555> you will also need an algorithm to check for uniqueness of a design
[04:37:18] <z64555> a very simple one would be to subtract one image from another, and then get the average pixel value of the result
[04:37:40] <RR101> http://prntscr.com/c6bu4q
[04:37:59] <z64555> that would check the image as a whole, more sophisticated algorithms might be able recognize patterns, but that's on the order of machine vision
[04:38:51] <z64555> that's an interesting clock design
[04:39:15] <z64555> although I think the us are a bit too much
[04:39:25] <z64555> it would look solid to the ordinary eye
[04:39:44] <RR101> It seems to be a challenge. Each design has a purpose for now can not reveal.
[04:40:02] <z64555> heh.
[04:40:09] <z64555> You're making graphics for a game
[04:40:30] <z64555> that ones a clock, and the others look like status for various things, such as hard drive or partition space
[04:40:38] <z64555> a game, or an operating system.
[04:41:11] <z64555> the tiger is a nice unique one, could be a background screen for something
[04:42:17] <z64555> and of course you are under an NDA, so you can neither confirm nor deny this guess of mine. :)
[04:43:11] <RR101> Not know as you intuit , but you guessed right, it is a watch, I challenge you to decipher operation without Requiring my help.!
[04:43:53] <z64555> that's alright, I won't jeopordize the secrecy of your project. :)
[04:45:00] <RR101> http://prntscr.com/c6bx5b
[04:48:37] <RR101> hahaha, I doubt you determine the time Achieve the clock, there is no risk.
[04:49:52] <RR101> It is not a secret, if little known. But you said you like challenges
[04:51:57] <RR101> The sixth rule is that the designs should not be similar.
[04:53:53] <RR101> z64555: What will be the format in which the designs are generated?
[04:54:34] <z64555> .SVG comes to mind. although I am sure there are other vector graphics formats
[04:57:41] <RR101> It is the format currently in use. The procedural generator can create multiple designs simultaneously or just one to one?
[04:57:41] <z64555> unfortunately I must go for now, best of luck
[04:58:18] <z64555> eh, probably just the one, since you have mentioned it will be running on an Android device
[04:58:45] <z64555> unless you meant the designs will be used on Android devices, and implied that the designs could be made on a strong computer
[04:59:08] <RR101> I like I contact you for further technical queries in future?
[04:59:12] <z64555> but anyway I must go. It's been a pleasure talking with you
[04:59:52] <RR101> mm
[05:00:52] <RR101> In android the finished design is executed, the generation is done from a powerful desktop computer.
[05:02:05] <RR101> It is notable for its technical knowledge, I can get on this channel in the future?
[05:04:48] <deshipu> it's a public channel, you don't need permission to come here :)
[05:07:04] <veverak> syntax checkers are awesome
[05:07:12] <veverak> one don't have to run the code to see it sucks!
[05:07:14] <veverak> :)
[05:07:28] <RR101> deshipu: You know about procedural generation?
[05:18:13] <RR101> deshipu: Tengo que conseguir a la persona correcta para desarrollar el generador de SF.
[05:27:53] <RR101> deshipu: All I know about you is that you like bad jokes, and you know how to program in low-level language.
[05:42:18] <deshipu> RR101: no, sorry, I only have very general knowledge on the subject
[05:45:01] <RR101> deshipu: Its power is in the hardware?
[09:54:58] <SpeedEvil> http://xkcd.com/
[12:22:17] <Loshki> I want one of those Tesla automated suicide machines...
[12:46:22] <rue_shop4> you can never understimate the stupidity of people
[12:46:39] <rue_shop4> why do I ahve 2 idle 3d pritners?
[12:53:45] <deshipu> smash them, you will have 0
[14:46:54] <i-make-robots> rue_bed: rue_house rue_shop2 rue_shop4 hey, at the VMMF I took pictures of your pull pull arms. I can't find the pics any more. Do you have any online, please?
[14:52:13] <i-make-robots> and everyone else here should check these pics out, the arms are gorgeous.
[17:13:04] <ferdna> it will be cool to use one of these:
[17:13:09] <ferdna> https://www.massdrop.com/buy/8bitdo-fc30-pro-or-nes30-pro-bluetooth-gamepad
[17:13:17] <ferdna> to control a
[17:13:26] <ferdna> rotor plane
[17:13:35] <ferdna> will i need anything else?
[17:13:43] <ferdna> or i have to purchase a full controller
[17:13:48] <ferdna> radio*
[17:19:36] <anonnumberanon> Incomplete information. Not able to make informed answer.
[17:23:58] <RR101> cameronfr: hi?
[17:33:56] <RR101> A Programmer Alghoritms Procedurals?
[17:38:07] <RR101> A Programmer Alghoritms Procedurals?
[17:46:49] <anonnumberanon> RR101, Do you have a problem?
[20:22:54] <ShadowZ> Wonder if I'll get around to installing the power rings on my quadrotor today
[21:47:03] <rue_shop4> pull pull arms...
[21:47:06] <rue_shop4> huh
[21:50:40] <ShadowZ> pull pull arms? what?
[21:52:19] <rue_shop4> I dont know
[21:52:40] <rue_shop4> maybe he was reffering to the robot arm with the bowden cables?
[22:58:00] <Casper> hi there, mig welding question... fluxcore vs C100... is there really a big difference? I'm out of fluxcore, weld infrequently... I have the hose and regulator, just need the adapter and the tank (200$+gas)...
[23:46:19] <rue_house> screw flux core
[23:46:21] <rue_house> get gas
[23:48:57] <Casper> do you have some photo that compare the result before and after cleaning of both?
[23:49:07] <Casper> or video or whatever
[23:58:20] <rue_house> you cannot clean up flux core welds, they always look like the result of a volcano sneeze
