#avr Logs

Nov 04 2017

01:06 AM BongoShaftsbury: when setting SPI clock, how do i determine an appropriate clock freq
01:09 AM day__ is now known as daey
01:56 AM rue_mohr: datasheet of what your talking to
02:10 AM BongoShaftsbury: the xbee says 6mhz and the teensy says 1/2 clock w/ clock @ 72mhz, so take the min of that?
02:12 AM rue_mohr: read the datasheet for an spi device
02:12 AM ZeroWalker: i think you usually want the slave to be a minimum of 1/2 of the master clock, but i think it will only work during the best cases and you usually need a bit more headroom
02:12 AM ZeroWalker: i meant 2x* not half xd
02:13 AM ZeroWalker: so if master is 8mhz, the slave needs to be 16mhz+ to ensure it doesn't miss anything (might be totally wrong though)
02:13 AM BongoShaftsbury: the xbee says 6mhz and i calculated 36 mhz for the teensy with the teensy being master but figured since the xbee wouldn't be able to make sense of anything > 6mhz
02:14 AM BongoShaftsbury: i think you're trying to apply nyquist theorem to it
02:15 AM BongoShaftsbury: where your sample rate has to be 2x your freq or you miss data
03:46 AM rue_mohr: (x<B9>,y<B9>)
03:46 AM rue_mohr: drat
05:12 AM nux_ is now known as nuxil
05:20 AM polprog: Isn't SPI clock there so you can freely manipulate the frequency?
05:21 AM polprog: like, keep it low wwjile your cpu executes an ISR or someting like that
05:21 AM polprog: spi is a clocked interface
06:32 AM nuxil: mornings.
06:33 AM nuxil: question. when dynamicly changing the prescalar on the fly. should one stop the timer 1st then set the new prescalar or dose it not matter ?
06:34 AM Lambda_Aurigae: umm.
06:34 AM Lambda_Aurigae: I'm sure the datasheet mentions this, but, to keep things from going wonky, yeah, stop it, change prescaler, then restart.
06:35 AM nuxil: alright. thanks :)
06:36 AM nuxil: yea. things are going wonky indeed. will try and stop it 1st then set the new one.
06:36 AM Lambda_Aurigae: when changing prescaler there might(might) be a little jitter.
06:36 AM nuxil: i expect that
09:53 AM polprog: https://puu.sh/yezR7/425d4e67d9.png
09:53 AM polprog: \o/
12:12 PM cehteh: uhm .. what are common forward voltages on zener diodes?
12:12 PM Lambda_Aurigae: depends on the zener.
12:13 PM cehteh: wanted to know just some ballpark, seems like normal silicon diodes ~0.8V or more, too much .. forget the idea
12:14 PM cehteh: do you know if there are zeners with very low forward voltages? 0.2V or such?
12:14 PM polprog: that would be shottkys
12:15 PM Lambda_Aurigae: do you mean forward voltage or voltage drop?
12:15 PM Lambda_Aurigae: looking around the minimum forward voltage is like 0.8 to 0.9 volts on most datasheets I just found.
12:15 PM rue_mohr: "do you mean forward voltage or voltage drop?" -> do you mean forward voltage or reverse breakdown
12:16 PM rue_mohr: fets and an op-amp :)
12:16 PM cehteh: voltage drop, doesnt the almost equal voltage drop?
12:16 PM Lambda_Aurigae: I think in this case forward voltage is the same as voltage drop...me and lack of correct terminology in my head.
12:17 PM cehteh: so i need a shootky-zener :)
12:17 PM polprog: pretty much :D
12:17 PM Lambda_Aurigae: is there such a thing?
12:18 PM cehteh: dunno
12:18 PM cehteh: i just forget about the idea
12:18 PM Lambda_Aurigae: it does not, apparently.
12:19 PM polprog: you can try to use a normal schottky as a zener, it will break down eventually
12:19 PM cehteh: i brought a small 5V powerbakn which i want put between PS and raspberry pi as USV
12:19 PM cehteh: i know this powerbank can charge and supply power at the same time
12:19 PM Lambda_Aurigae: polprog, at least once...
12:19 PM cehteh: but when it looses power there is a few ms powerloss when it switches over
12:19 PM Lambda_Aurigae: BIIIIG honking cap on the power input of the rPI
12:20 PM cehteh: thought i just add a goldcap + zener on the rpi yes
12:20 PM polprog: why zener
12:20 PM Lambda_Aurigae: 16V 470000uF should do it...about the size of a coke can.
12:20 PM cehteh: 4.5V zener to keep it charged and limit inrush current
12:21 PM cehteh: yeah thats way too big
12:21 PM Lambda_Aurigae: I have a box of 200 of them.
12:21 PM Lambda_Aurigae: screw terminals on top.
12:22 PM cehteh: maybe i just try a godcap i have laying around
12:22 PM Lambda_Aurigae: ok...off to go shoppin wif da wifey.
12:22 PM polprog: have fun
12:25 PM cehteh: http://tinyurl.com/yc8bz93a
12:26 PM cehteh: alike
12:26 PM cehteh: does that idea even work :D
12:27 PM polprog: that zener causes it to jump between 2V and 5V
12:27 PM polprog: let me draw something
12:28 PM cehteh: and didnt goldcaps have some rather low limit at what current you can charge them?
12:29 PM polprog: some have
12:29 PM polprog: some can melt screws as thick as a tjumb
12:29 PM polprog: supercaps*
12:29 PM polprog: not sure what a goldcap is
12:30 PM cehteh: charge, not discharge
12:31 PM cehteh: anyway i have some old bke taillight somewhere which has a supercap, can try with that
12:31 PM polprog: http://tinyurl.com/y93wt6tr
12:32 PM polprog: try this
12:32 PM cehteh: but the diode drop makes this thing impossible
12:32 PM polprog: not if you use a schottky
12:32 PM polprog: there's no schottky in that sim though
12:32 PM polprog: but the general rule applies
12:32 PM polprog: the diode is there to prevent the current going from the cap to the battery
12:33 PM cehteh: i dont think the 'battery' side needs a diode here, actually its a powerbank which may already have proper circruit for that
12:33 PM cehteh: yes
12:33 PM polprog: may have
12:34 PM cehteh: otherwise its infeasibel, even shottky drop would be bad on that side
12:35 PM cehteh: well i try with that goldcap, w/o diode or anything, either it works or not
12:35 PM cehteh: have to find it first anyway and check if its a 3.3V or 5.5V type
12:39 PM twnqx: couldn't you use the mosfet diode replacment trick for lower drop than even shottky?
12:46 PM cehteh: too much trouble., i just scrap the whole idea
12:53 PM Ameisen: you kids and your fancy EE
01:38 PM polprog: Ameisen: excuse me, i started as a programmer :P
03:07 PM Ameisen: Well, the main AVR maintainer for LLVM seems pleased with my changes so far
03:07 PM Ameisen: though there's still a lot to do
03:08 PM Ameisen: polprog - since you expressed interest, let me put the list somewhere
03:08 PM Ameisen: https://pastebin.com/qnpG67dX
03:09 PM Ameisen: I also have one outstanding, severe avr codegen bug in g++ that needs to be fixed that goes back to like GCC 3.
03:09 PM Ameisen: That particular bug does not exist in Clang
03:26 PM ZeroWalker: hmm, so if i have a DMA from SPI1 on my STM32 and when it's done i want to basically passthrough the data to USB Serial, but if i try to write to the USB in the DMA interrupt it basically breaks and i have to reset the device, why is that
03:26 PM polprog: thanks, ill take a look
03:29 PM Ameisen: polprog - a few of the optimization things I'm not sure are entirely doable within the context of the existing middle-end
03:30 PM Ameisen: the IL in both gcc and llvm are pretty high-level
03:30 PM Ameisen: some of the optimizations require lower-level IL to actually derive
03:30 PM Ameisen: like, how the heck do you get a compiler to emit LAT?
03:30 PM Ameisen: load-and-toggle
03:30 PM Ameisen: that requires a very specific IL instruction, and no amount of lowering is going to come to the point that it thinks that that's a good idea
03:34 PM Ameisen: at some level, I wonder if a custom middle-end is the best end-goal
03:34 PM Ameisen: the complexity of middle ends like llvm's/gcc's are in the strength of their scheduling which really isn't relevant to AVR
03:35 PM Ameisen: they have a ton of work set up to improve cache locality/the ability for SIMD/multithreading to work
03:35 PM Ameisen: none of which means anything to AVR
03:35 PM Ameisen: on the other hand, their middle-end limitations tend to increase register pressure in AVR
03:35 PM Ameisen: due to forcing things like 16-bit registers on any 16-bit type regardless of usage
03:36 PM Ameisen: so if all your non-paired registers are free, but you're using all of the paired registers, and you try to copy a 16-bit memory location to another, it has to vacate one register pair to do it
03:36 PM Ameisen: even though there's zero reason it couldn't do it with existing registers
03:36 PM Ameisen: it has no concept of it
03:36 PM Ameisen: and it requires basically another optimization pass before register allocation
03:46 PM polprog: keep in mind they have to make a compiler that works well on everything
03:47 PM polprog: i'd be curious how other compilers apart from gcc and clang do on avr
03:47 PM polprog: sdcc doesnt support avr sadly
03:51 PM Ameisen: Sure. This is why I'm saying that forking a compiler could work
03:51 PM Ameisen: take the frontend, write a new middle-end/back-end
03:51 PM Lambda_Aurigae: ZeroWalker, because your code is wrong.
03:51 PM Ameisen: dedicated for AVR
03:51 PM Ameisen: or just 8-bit chips
03:51 PM Ameisen: the frontend is often the hardest part
03:51 PM Ameisen: especially for C++
03:51 PM Ameisen: so using an existing one can be nice
03:52 PM ZeroWalker: well that much i know, but i don't understand what hte problem is, i guess it's that i don't send the data fast enough before the next DMA occurs
03:52 PM Lambda_Aurigae: well, without code it's hard to say...other than, something's not right.
03:52 PM Lambda_Aurigae: no clue what stack you are using, or anything.
03:53 PM ZeroWalker: stack?
03:53 PM Lambda_Aurigae: I might recommend asking in ##stm32 however.
03:53 PM Lambda_Aurigae: usb stack.
03:53 PM ZeroWalker: oh ther'e actually a channel for that
03:53 PM Lambda_Aurigae: the software that you use for communicating over usb.
03:53 PM ZeroWalker: well it's just Serial.write, and them talk to COM Port
03:53 PM ZeroWalker: not sure what's going on in the background, new to this
03:53 PM Lambda_Aurigae: so you are using some HAL that you know nothing about.
03:54 PM Lambda_Aurigae: again, ##stm32
03:54 PM Lambda_Aurigae: they will likely have more of a clue.
03:55 PM ZeroWalker: well this is the first time i ever do anything with these sorts of things, so it's very new, will ask at that channel, thanks
03:56 PM Ameisen: Any of you guys ever done self-modifying code in AVR?
03:56 PM Ameisen: I imagine it's more difficult than on most archs, since you have to do program-memory loads/stores
03:56 PM Ameisen: and stores for program-memory are a PITA
03:56 PM Lambda_Aurigae: only with an interpreted language.
03:57 PM Lambda_Aurigae: well, interpreted-scripty-language.
03:57 PM Ameisen: Working on things like 3d printers/CNC
03:57 PM Ameisen: instead of gcode, I was considering pre-compiling gcode to AVR machine code
03:57 PM Ameisen: and then loading that as the 'script'
03:57 PM Lambda_Aurigae: I find other architectures better for self-modifying code.
03:57 PM Ameisen: it'd be smaller/faster than gcode
03:57 PM Ameisen: though potentially too large to fit in flash
03:57 PM Ameisen: whereas gcode can be loaded in parts into RAM for an interpreter
03:58 PM Lambda_Aurigae: g-code interpreters exist for avr and are pretty fast.
03:58 PM Ameisen: sure
03:58 PM Ameisen: I'm just thinking of going even faster.
03:58 PM Ameisen: usually, AOT compilation is as fast as you can go, but AVR has limitations
03:58 PM Lambda_Aurigae: your limiting factor is likely to be your hardware rather than processor.
03:58 PM Ameisen: Sure, I just want to reduce time taken for gcode so I can throw it at other things :)
03:58 PM Ameisen: That, and having it AOT compiled makes a few things easier, like movement prediction based upon future instructions
03:58 PM Lambda_Aurigae: then I would go with a second processor or a faster processor.
03:59 PM Ameisen: since such things can be pre-encoded
03:59 PM Lambda_Aurigae: if it was me, a second processor usually.
03:59 PM Ameisen: that's an option as well
03:59 PM Ameisen: I tend to like AOT in principle - written a few AOT compilers
03:59 PM Ameisen: always fun
03:59 PM Ameisen: writing an AOT recompiler for MIPS->x86-64 was fun
03:59 PM Ameisen: if weird
04:00 PM Lambda_Aurigae: I'm currently starting work on a BASIC interpreter for MIPS...pic32 specifically.
04:01 PM Lambda_Aurigae: modeled loosely after the commodore vic-20 interpreter.
04:02 PM Lambda_Aurigae: hoping to fit the BASIC interpreter and USB interface and bootloader into 32K of flash and maybe a light DOS for SD cards in another 32K.
04:04 PM Ameisen: Lambda_Aurigae - my mips emulator is for MIPS32r6
04:04 PM Ameisen: so, if you're using MIPS32r6 at some point, you can use it
04:04 PM polprog: hey, wonder if you could compile freedos for that
04:04 PM Ameisen: it's the fastest MIPS emulator I know of
04:04 PM Ameisen: though it doesn't target any particular MIPS chip
04:04 PM Lambda_Aurigae: polprog, there's an open source dos implementation that will work on it.
04:04 PM Ameisen: it's a generic emulator for a generic MIPS32r6 chip
04:04 PM Ameisen: that happens to meet the spec
04:04 PM Lambda_Aurigae: I don't generally use emulators.
04:04 PM Lambda_Aurigae: I just run on the real hardware.
04:04 PM polprog: Lambda_Aurigae: that's what im talking about
04:04 PM Ameisen: Lambda_Aurigae - mine is spec-compliant, and faster than any real MIPS system
04:05 PM Ameisen: also allows line-by-line debugging of MIPS code in Visual C++
04:05 PM Ameisen: which is still the neatest thing ever
04:05 PM Lambda_Aurigae: so, useless here.
04:05 PM Ameisen: prolly
04:05 PM Ameisen: it allows line-by-line debugging in gdb as well
04:05 PM Lambda_Aurigae: as visual C++ won't run on this computer.
04:05 PM Lambda_Aurigae: or any of my computers.
04:05 PM Lambda_Aurigae: as it requires windows.
04:05 PM Ameisen: it relies on the visual C++ gdb bridge
04:05 PM Ameisen: so... gdb works too
04:05 PM Ameisen: lldb as well
04:05 PM Lambda_Aurigae: but, can't remember the last time I used a debugger.
04:06 PM Ameisen: I just use VC++ usually
04:06 PM Ameisen: heh
04:06 PM Lambda_Aurigae: I just debug on the hardware.
04:06 PM Ameisen: first part of vemips development, I printf'd a lot, and did outputs
04:06 PM Ameisen: once you have a proper debugger though...
04:06 PM Ameisen: so much nicer
04:06 PM Ameisen: especially a proper visual debugger
04:06 PM Lambda_Aurigae: i/o pins and serial port for debugging here usually.
04:06 PM Lambda_Aurigae: I've used VC++...way back when.
04:06 PM Ameisen: nice to just tell it to stop on a line, and hover over variables to see their values
04:06 PM Lambda_Aurigae: never found anything about it nice.
04:06 PM polprog: i did that with avarice and DDD
04:06 PM Ameisen: yeah, people either love or hate vc++
04:06 PM Lambda_Aurigae: along with visual basic and visual java and all that too.
04:07 PM Ameisen: if you love VC++, you will hate anything else, and vice versa
04:07 PM Lambda_Aurigae: used to have to work with them years ago.
04:07 PM Ameisen: but if you were using Visual Java...
04:07 PM Ameisen: the last VC you used was VC6
04:07 PM Lambda_Aurigae: sounds about right.
04:07 PM Ameisen: which... isn't particularly representative of current versions.
04:07 PM Ameisen: Though if you're on Linux, there's Visual Code
04:07 PM Lambda_Aurigae: I played with visual studio 13 at work.
04:07 PM Ameisen: which people seem to really like
04:07 PM Lambda_Aurigae: found it unusable for the way I think and work.
04:07 PM Ameisen: I use Visual Studio 2017
04:07 PM Ameisen: but yeah
04:08 PM Ameisen: if you aren't big into IDEs, you'll hate VS
04:08 PM Lambda_Aurigae: yup.
04:08 PM Ameisen: I just generally prefer it if you _can_ use an IDE if it exists
04:08 PM Ameisen: I hate projecs that enforce one viewpoint or another
04:08 PM Ameisen: that's one of my annoyances with GCC, and one thing I like about LLVM
04:08 PM Lambda_Aurigae: it comes from my early programming on basic based machines and unix systems waaay back when.
04:08 PM Ameisen: GCC? You're using a text editor.
04:08 PM Lambda_Aurigae: yup..usually vi
04:08 PM Ameisen: LLVM? Can be built/debugged in basically any IDE. Or a text editor.
04:09 PM Ameisen: thus, LLVM has better accessibility, and on a project that large, full IDEs are very helpful sometimes
04:09 PM Ameisen: when you have to jump across 10 files to figure out what a symbol means
04:09 PM Ameisen: and the IDE links it together for you
04:09 PM Ameisen: annoying with vi/gedit/whatnot
04:09 PM Ameisen: but, it's all mindset
04:09 PM Lambda_Aurigae: if I have to do that then I didn't comment my code properly.
04:09 PM Ameisen: I can say that usually, IDEs increase productivity noticeably, though if the programmer hates IDEs, they resist the IDE
04:10 PM Ameisen: and their productivity goes down :|
04:10 PM Ameisen: my productivity went down with xcode
04:10 PM Ameisen: worse IDE ever
04:10 PM Ameisen: it's incredibly unintuitive
04:10 PM Ameisen: at least what VC++ does makes sense
04:10 PM Ameisen: xcode randomly opens files sometimes, sometimes tabs, sometimes doesn't do something sane at all
04:10 PM Ameisen: I want consistency, dang it
04:10 PM Lambda_Aurigae: ok...back to making sawdust!
04:10 PM Ameisen: mmmm, sawdust
04:11 PM Ameisen: I need woodflour if you have any
04:11 PM Ameisen: from a safe wood
04:11 PM Ameisen: like oak
04:11 PM Lambda_Aurigae: turning an attic stairwell into a closet.
04:11 PM Ameisen: I make formicariums
04:11 PM Ameisen: so woodflour is useful sometimes
04:11 PM Lambda_Aurigae: mostly pine here.
04:11 PM Ameisen: pine is safe if kiln dried
04:11 PM Ameisen: and not PT
04:11 PM Ameisen: there's a few species like black walnut that are outright not safe
04:11 PM Lambda_Aurigae: but I do most of my cutting outside and don't collect the sawdust for the most part.
04:11 PM Ameisen: even for humans
04:11 PM Ameisen: and fatally toxic to dogs and horses
04:12 PM Lambda_Aurigae: I work with oak and cherry and pine primarily...for construction it's all pine.
04:12 PM Ameisen: I tend to use fir for construction
04:12 PM Ameisen: I used to use oak a lot, been moving towards cherry or similar
04:12 PM Ameisen: oak is just so damned brittle
04:12 PM Ameisen: you look at it wrong and it splits
04:12 PM Ameisen: I have it split when making pilot holes to prevent splitting
04:13 PM Lambda_Aurigae: I have a pile of cherry planks about 1x4 to 1x6 and 4 to 6 feet long...maybe 100 or so pieces.
04:13 PM Ameisen: it's the only wood I have where GRK screws cause it to split
04:13 PM Ameisen: the damned GRK screws are designed to make wood _not split_
04:13 PM Lambda_Aurigae: I wet oak to drill it.
04:13 PM Lambda_Aurigae: steam wet it.
04:13 PM Ameisen: I may have to try tha
04:13 PM Ameisen: that
04:13 PM Ameisen: though doesn't that warp it?
04:13 PM Lambda_Aurigae: small area only.
04:13 PM Ameisen: At the moment, though, I am moving towards cherry - similar strength
04:13 PM Ameisen: but not brittle
04:13 PM Ameisen: and cheaper, IIRC
04:13 PM Ameisen: oak is surprisingly expensive
04:13 PM Lambda_Aurigae: this cherry I have I got free.
04:14 PM Ameisen: and not particularly resistant to environmental things
04:14 PM Ameisen: insects, in particular, love oak
04:14 PM Ameisen: though I keep my carpenter ants in oak nests
04:14 PM Ameisen: I don't remember if my biggest nest is in oak or cherry
04:14 PM cehteh: http://tinyurl.com/yauyk76t .. so how do i get that FET into saturation?
04:14 PM Ameisen: once it's a formicarium, it looks the same
04:14 PM Lambda_Aurigae: I also have about 30 or so poplar planks about 1x6 and 3 to 4 feet long.
04:14 PM Ameisen: also, I keep ats
04:14 PM Ameisen: ants
04:14 PM Ameisen: if you didn't gather that
04:14 PM Ameisen: I sell queens, too, if you live in Illinois
04:15 PM Lambda_Aurigae: yup.
04:15 PM Ameisen: you live in IL?
04:15 PM Lambda_Aurigae: the formic gave it away.
04:15 PM Lambda_Aurigae: Iowa.
04:15 PM Ameisen: oh
04:15 PM Ameisen: so close
04:15 PM Lambda_Aurigae: we have too many damned ants here already.
04:15 PM Ameisen: yet just far enough to be illegal
04:15 PM Ameisen: oh, general ants are easy to find
04:15 PM Lambda_Aurigae: no kids to have an ant farm either.
04:15 PM Ameisen: queens with workers? Harder to find
04:15 PM Ameisen: especially for hobbyists
04:15 PM Ameisen: I have no kids and am thirty
04:15 PM Ameisen: still have formicariums
04:15 PM Lambda_Aurigae: I just poison ants whenever possible.
04:15 PM Ameisen: don't poison all ants
04:16 PM Ameisen: that's how fire ants became a problem
04:16 PM Lambda_Aurigae: thinking of getting into beekeeping though.
04:16 PM Lambda_Aurigae: that could be fun.
04:16 PM Ameisen: you kill 'good' ants, which makes hardier, more resistant 'bad' ants spread more
04:16 PM Ameisen: just like grass vs weeds
04:16 PM Lambda_Aurigae: I don't poison them outside...just in my house.
04:16 PM Ameisen: some ants develop slowly but aren't a problem
04:16 PM Ameisen: some are ridiculously fast spreading but weak overall
04:16 PM Ameisen: even carpenters are beneficial if they aren't in your house
04:16 PM Ameisen: they develop slowly, too
04:16 PM Ameisen: very slowly
04:16 PM Ameisen: unbearably slowly if you're keeping them
04:17 PM Ameisen: https://i.imgur.com/Mw0AlXm.jpg
04:17 PM Ameisen: one of my founding formicariums
04:17 PM Lambda_Aurigae: I do encourage spiders and snakes in my basement though.
04:17 PM Ameisen: very small, but so is the nest
04:17 PM Ameisen: about ~20 workers
04:17 PM Ameisen: Camponotus subbarbatus - doesn't have a common name
04:17 PM Ameisen: I call 'em dwarf striped carpenter ants
04:17 PM Ameisen: they make tiny nests, usually have only a few hundred workers
04:17 PM Ameisen: they're technically pests, but even if they nest in your house, the colonies are so small that they are unlikely to cause damage
04:18 PM Lambda_Aurigae: ok...gotta go start on the next step in the project....building a door to the attic.
04:18 PM Ameisen: unlike their big brothers C. pennsylvanicus, the black carpenter ant
04:18 PM Ameisen: twice the size, and usually have tens-of-thousands of workers
04:18 PM Ameisen: also, C. subbarbatus are surprisingly pretty
04:18 PM Ameisen: evenpeople who hate ants note that they're attractive
04:18 PM Ameisen: because of the golden stripes
04:18 PM Ameisen: also surprisingly rare and hard to find
04:18 PM Ameisen: still happy I found a queen
04:30 PM polprog: cool hobby, ants
04:40 PM enh: really cool, Ameisen
04:40 PM enh: Where can I lern more about them?
04:45 PM Ameisen: ants?
04:45 PM Ameisen: lots of places
04:45 PM Ameisen: looking up species, sites like antscanada or formiculture
04:46 PM Ameisen: keeping them and seeing what they do
04:46 PM Ameisen: to keep proper formicariums, you at least need a queen (preferably newly mated), though survival/success rates of fresh queens are quite low
04:46 PM Ameisen: usually catch more than one
04:46 PM Ameisen: see which survive
04:46 PM Ameisen: my C. subbarbatus thankfully has done OK
04:48 PM enh: Where can i keep them?
04:48 PM enh: A formicarium is made of what?
04:49 PM Ameisen: anything really
04:49 PM Ameisen: seen them made out of tons of things
04:49 PM Ameisen: mine are wood and acrylic
04:50 PM Ameisen: people use plastic, cement, hydrocal, plaster...
04:50 PM Ameisen: basically, if the ants can't destroy it, and it's non-toxic, it works
04:50 PM Ameisen: though there's obviously some engineering that goes into it
04:50 PM Ameisen: usually, at the start, queens go into test tubes that have water halfway, stopped with cotton
04:50 PM Ameisen: simulates their claustral chambers well
04:50 PM enh: And do they live the formicarium? Can you keep them away from your food?
04:51 PM Ameisen: formicariums are generally designed so they don't escape
04:51 PM enh: ok
04:51 PM Ameisen: that one is vinyl-tubed to an 'outworld' that simulates outside for them
04:51 PM Ameisen: which I put food in
04:51 PM enh: very interesting
04:51 PM Ameisen: though it's november, they're preparing to hibernate
04:51 PM Ameisen: so they aren't using the outworld much
04:51 PM Ameisen: I like C. subbarbatus. They're small but not tiny, make small colonies
04:51 PM Ameisen: they're like perfect for keeping
04:52 PM Ameisen: and they develop relatively quickly
04:52 PM Ameisen: hard to find though
04:52 PM Ameisen: they're not rare, just they nest in weird places
04:52 PM Ameisen: like twigs
04:52 PM enh: here sometimes i have to poison them. But I really do not like doing that. I have those small black ants that like sugar.
04:52 PM Ameisen: depending on where you are, that can be a few species
04:52 PM Ameisen: common tiny black ant pests here are Tetramorium spp. and Tapinoma spp.
04:52 PM enh: brazil
04:53 PM enh: NE
04:53 PM Ameisen: no idea for Brazil
04:53 PM Ameisen: not sure what ants are there
04:53 PM Ameisen: I know you have Tapinoma and Camponotus, but they're different species than are here
04:53 PM Ameisen: like, C. subbarbatus is only found in in the northeast of North America
04:53 PM Ameisen: from Kentucky to like Quebec
04:54 PM Ameisen: killing them is usually the same. Feed them something tainted with boric acid.
04:54 PM Ameisen: shuts down their digestive system, they share it with the rest of the colony, etc
04:54 PM Ameisen: I'm not fond of killing wild colonies, but in the house is generally a different matter
04:54 PM Ameisen: especially if they're a pestiferous species
04:55 PM enh: Here in brazil we have enough species. From tiny yellow ones that bites like taking off your finger to 5cm ones that scare the hell out of me
04:56 PM enh: are you aware of grebbenikov's work?
04:57 PM Ameisen: nope
04:57 PM cehteh: https://wiki.openwrt.org/_media/media/mikrotik/rb951g2hnd-03.jpg lol that layout, priceless
04:57 PM enh: you may find it interesting
05:02 PM polprog: wow, cool traces
05:13 PM enh: Ameisen, what do you fill the formicarium with?
05:13 PM Ameisen: ants
05:14 PM Ameisen: Mine are generally just carved wood covered with acrylic
05:14 PM Ameisen: but that's personal preference
05:14 PM Ameisen: don't know what you mean by 'fill'
05:14 PM Ameisen: I don't normally make formicariums that they dig in, because they tend to dig in a way that obscures visibility
05:16 PM enh: I thought they were filled with earth, or something similar
05:17 PM Ameisen: if you want them to be able to dig
05:17 PM Ameisen: I don't like ones with digging media
05:17 PM Ameisen: they tend to dig with dirt in front of the glass so you can't see the
05:17 PM Ameisen: them*
05:17 PM Ameisen: as it blocks light
05:18 PM enh: I remember to have seen flat ant houses, with earth between two acrylic plates near each other
05:35 PM enh: I've been experiencing a very weird update problem in my system. Working form more than a week to find the bug. Just found it. Such an idiot I am.
06:10 PM polprog: night
06:11 PM Lambda_Aurigae: nighters polprog
06:12 PM enh: sleep weel
06:12 PM enh: well