Hey, they have a forum there! I didn't think to beg them for help! hahaha! Well, I'll have to learn how to beg in Russian! (I guess that's not as much of a joke as I meant it. I have a few Russian language textbooks, and what I can't do there, I'll have google give it a shot!).

So, I was flippant/incorrect there. The CY7C68033/34, which I have purchased, HAS a 8051 as its core processor, so, if I am to use it, I will have to learn 8051. This tells me that 6502 will be much diminished in my project, but maybe not. I still haven't looked at the USB on this forum page (I use public library internet and have limited time (time share? Why does that seem so very "1970's"?); last time I posted, I spent all day long "gibbering" and "jabbering".).

Also, looking at the CY7C68033/34 I notice that it has both a FIFO/GPIF mode and a Flash (NAND) mode, and they are register configurable. So, basically, as I understand it, I would have to pick one or the other? However, I think the USB port can run simultaneous to these other items. This will be useful, I think.

KAC9628 has a synchronous 12 bit bus, and I can feed it to the AL460 16MB FIFO memory. Unfortunately 16 bits wide, the FIFO, but I guess I will just pretend it is 12MB (tie the pins or float them? I forget; the specs will tell me, I hope?). This FIFO will read and write simultaneously, with circuitry to ensure no contention if read and write should "land" on the same address. It does need special heat dissipation "attention" for the PCB design, so I will read up on that.

I have not checked everything for oscillator frequency/power/logic level agreement, but hopefully at least SOME of the chips will agree with each other, because"level shifters" and things of that nature, will complicate things.

So, from the FIFO, it goes to a FIFO, of sorts. The Cypress CY7C68033/34 chip has an internal FIFO. The documentation is sparse, and I am here today, on the internet, looking for the technical reference manual, so I can get up to speed on how the CY7xxxx works.

So, I can feed the 12 bits from the 16MB (12MB?) FIFO into the GPIF/FIFO "mode" port of the CY7xxxx and, hopefully, if I understand things, fill up the 4kB internal-FIFO, then, switch the mode to NAND Flash, and shunt 4kB of info to the NAND Flash, at a time.

There are limits on the number and size of NAND Flash chips; page size and etc. The chips I was looking at might be too big (Micron, MT29G08CFACA... or some such number?). Now, if I understand things correctly, I can "USB" output some data (i.e. CY7C68033 acting as host) simultaneously to my "data-switching/shunting" between the FIFO and NAND? Maybve send the USB to the RAID array?

I will look into all of that. If the CY7C68033 is "too much chip" for me too handle, I will try to read the EnCoRE datasheet. It is cheaper anyway.

Now, if the CY7C680xx doesn't do what i said it will do (above), then I can also do the following; KAC-9628 to FIFO to FTDI60x (AKA, FIFO to USB bridge) to CY7C680xx (USB to NAND) to NAND (Flash mem, Micron MT29G08CFACA?). Its an extra chip in there, but its not a particularly expensive chip, so no sweat. Besides, I am not making a million of these things; one will do.

Now, I need to think about a "cycle steal" and/or "frame steal" circuit, once I get the stuff (above) settled. I read somewhere that a human being can recognize its mother (or any other item/image) after just 200 milliseconds (or was it microseconds? The latter would be tough, but do-able; the former, not quite easy, but certainly do-able.). That is from "eyes-set" to "demonstrable recognition".

Oh well, I am running out of time again. Cheers!

Whoops; still need a "USB host controller". I mean, could run it from a laptop or desktop, with drivers and such, but I am not a high level language guy, and it seems like cheating. I have to know this thing from the ground up.

Cypress offers a few USB host controllers, as does FTDI. A few others , in a quick internet search. Even found a few cheap, pre-made boards, which saves development time while not hindering learning , too much.

Haven't posted in awhile. In large part because I have strayed from a 65816 and/or 65xx project, into another realm. Still, I could probably use the 65xx. Still looking for a USB Host Controller, but have let summer intrude on my thinking. Next order I make should include a USB controller, or five or ten or so (the cheap and easy ones!).

My birthday is a nice month for my robot project. Sans Digital RAID array, TRM4+B, with 12TB capacity on 4 SATA (II?) HDD, if run as a JBOD. Capable of RAID 0,1,5,10 and etc. Got two cards to plug onto my desktop, but I only have a standard PCI socket and a PCI-Express is needed? Oh well. One of the cards is plain PCI, and outputs to a SATA cable. I think I can get a 5$ SATA to eSATA cable, because the RAID array takes an eSATA.

So, my desktop will probably be involved in the robot project for awhile. Ordered a laptop too (needed one... my desktop computer is eleven years old and my laptop was stolen in 2004!). Hopefully, one of the slots in the RAID array will be filled by my anthropomorphic retinomorphic focal surface image sensor array.

Robot bones came in the mail too. I feel like I got less than I ordered; I'll have to check receipt and order against received. Maybe my fault? Bad calculations? Whoops!

Was thinking about titanium bones, or central rod, but the carbon fiber pulltruded tubes forming the hexagonal "cylinder" of limbs is probably plenty strong. I had worried they might build a static charge, but my cheesy little ohmeter says they are very conductive. So, maybe just ground the CF tubes? Titanium also has high strength to weight ratio. Either one will be fine I think?

Also, SMAs might make for interesting robot muscles, but cycle time is a concern (and heat shedding and insulation of wires?).

So, have really let the posting on my project slip. Probably because I am such a neophyte! But also, when you "do the work" in the "real world" there is less time for speculation. Not ALOT of work done, but I have acquired more than a few parts, and the "outline" of a working device may come into view very soon.

I am a little bit sidetracked by Goldberg polyhedra and the math surrounding it. https://en.wikipedia.org/wiki/Goldberg_polyhedron http://www.pnas.org/content/111/8/2920.full.pdf?with-ds=yes http://www.pnas.org/content/111/8/2920.abstract

Neat to think that 2000 years--or thereabouts--has passed since people came up with a new polyhedra, and that some guy (actually, an eminent and known mathematician, if one knows how to look these things up. It took some tracking down, but I found many papers attributed/ascribed to Michael Goldberg, and it seems he was well known in the mathematical community, i.e. mid-atlantic states Wash. D.C. area, circa 1920s-1940s) could just dream it up in 1937 and be forgotten until Schein submitted his PNAS paper in 2013.

Now, what I want to know, is, if given that the sides of the regular polygons are all equal, i.e. s-sub-n = s-sub-1 = s-sub-etc. , and, if the Goldberg polyhedra is specified by only two numbers , namely, m and n, (thusly, G[m,n] ) , then, the question for mew, remains, "How close are these facets to the actual surface of the sphere, and, furthermore, how does on determine the ideal radius is on wanted the points to be coincident between the faceted surface and the spherical surface. This is an important question, and I wish I had the paper that describes such equations. Maybe such a paper does not yet exist?

I am preparing some rudimentary remarks on the subject, but, again, still keep my mind on "the prize" of a "curved focal surface array of planar process integrated circuits in anthropomorphic arrangement, with foveated resolution", or similar such description of my invention. Though I have some notes in my inventor notebook from circa 2011-2012 indicating that I could create a fiber optic image conduit curved focal surface field flattener (like, an extreme version of the one described in the patent literature by William Hicks, of American Optical company, as described-in/alluded-to-by the book city of lights, by Jeff Hecht), I am afraid one or two researchers may have "beaten me to the punch" in the actual realization of such a device (circa 2012-2013). Arianpour is the last name of the principal researcher (U of san diego maybe? or SoCal? I dunno?), and it is either a SPIE publication or IEEE publication? I forget?

Anyhoo, despite being almost-scooped, I soldier on.

My "tack" for tackling the Goldberg Polyhedra will be (1) read the rest of Schein's PNAS paper (duh!!! too lazy, so far!) and (2) try setting the Radius of the sphere, R, equal to the distance to the center point of ALL polygons and/or (3) try setting the Radius of the sphere, R, equal to the distance to the vertices of ALL polygons.

This may or may not work perfectly. I read somewhere that the "facets" on a GP are not necessarily "flat", rather, they might pivot in "three-dimensional angles", sticks--rotating like so many "ball and stick models"--around the balls, yielding a polygon with equal sides, but non-planar, i.e. "non-2-D".

So, I will tackle the math, and find the best solution, when given m, n, s, and R.

Got a CH341A EEPROM USB programmer for extremely cheap, and I hope it will function with the EEPROM I purchased. Haven't looked at my EEPROM chip in awhile, but when I do (tracking down the part numbers on line, here, today, so that I can see if the CH341A will "talk" to it) I will try to get the relevant software (if any) and check that the part number is agreeable to the CH341A manufacturer/device.

It seems to be a generic item, and I could not find the manufacturer name or pdf datasheet for the CH341A that i bought. Hopefully when it arrives sometime this week, such information will be included.

Also, I purchased, (cheap, also, 5-bucks-ish) a USB to TTL/RS232 bridge-chip/adapter, so that I can program my computer (Dell Desktop, WinXP, 2006-ish model) to tell my speakjet what to do. I had a crisis with the speakjet when i ordered it and could only get it to spit out the syllables in "demo mode", and I had (HAVE!Still have!) much work to do to get my mind thinking about communication links and wires and buses and protocols and such. It has been quite a journey so far, and it is a task that promises to bog mind down for at least 3-5 years to come! ONFi, NAND-flash, SATA, eSATA, USB and etc. But maybe, it will be easy to grab a DB-9 or DB-15 connector, cut the adapter off one end, splice the wires to the USB-to-TTL-RS232 device, and presto chang-o, we could have a speech chip! Things are rarely that easy (for a biologist like me!), but lets hope it happens!

Purchased TWO USB host chips, but have not lloked at them indepth. I now how plenty of chips on hand, probably more than necessary, and I should start to REALLY understand the data sheets and the peculiarities and promises of each product. Probably should have got that before I ordered, but it is nice to have options "on-hand", should I change my mind later.

Assuming my CH341A works, I could probably throw together a "mini-6502" project, rather, a first-venture into the 6502 hardware world. I have all the parts (RAMs, ROMs, 6502s/65816s, VIAs, etc.), and a few extras, even, so I could just put it on a project board and keep the clock speed slow, to avoid ringing and/or interference and or noise and etc. newbie-experimenter-board-problems.

I got a PC laptop, for my birthday, but gave it to my "mentee-kid" (who imagines himself a black-hat hacker), (Luckily, he's a good share-er (amazingly nice , for a black-hat! Kudos to his mom!) ). Who am I to snuff out those dreams? Besides, at his age, I thought myself a baseball player, and dreams are wonderful things to nurse. In a "Post-Snowden" age, it can be tough to know if there REALLY ARE "Black-hat", "White-hat", and/or "Grey-hat" hackers? The Gov't--and to no less an extent, the "free market"--blurred the lines, I think?

Anyhoo--I digressed; so I got a hexeditor on this PC and I also downloaded Stella, the Atari emulator with EVERY SINGLE Atari game known to man! So, what did I do with this newfound freedom? I altered pacman to make the ghosts into middle fingers (rude), and the pacman into a peace-sign (or victory, depending on what generation you grew up in) hand. I don't do too much programming, but I envision more than a few creative enterprises arising from these interesting tools. Stella shows the sprites-"bitmaps", the hex and binary, side by side, so it is easy to open the bin files in my BeHex (hexeditor) and find the relevant line to "monkey-with".

I guess that is all for now. Too much talking, and not enough working. I got my robot bones and am looking for elbows and knees and joints. Need some bearings/bushings, motors and gears. But this will be a slow process, determining what are the best parts to buy/use. Also, got some strain gauges, constantan, and if I get them working properly I can probably just move to "PCB, copper" strain gauges, of my own unique design. That would be cheaper (constantan strain gauges are about 1$ a piece, though possibly 50 cents to 10 cents, if bought in bulk; I haven't really found a good bulk supplier, yet.). Also, EM interference might be bad for strain gauges; perhaps optical strain gauges? I dunno?

The RAID array is still unused. I ordered the eSATA to SATA wire, but I keep getting the wrong kind, Ordered it five times and the sixth will be a charm. Just stupid mistakes on my part. I'll get it right this time!

It would be nice to publish in Science Robotics, since the new journal is being launched soon. I have spent significant time preparing such a document, but stalled in the past two weeks. Perhaps, begin again today?

American academia has severely hamstrung my efforts. I just realized (in preparing SR paper), the papers that are published every year that I don't have access to, are doing many GREAT things that I can only dream about, and I feel both AHEAD of the game, and BEHIND! Curse their knowledge monopoly! It serves to impoverish me, and ALL OF US!

If I can't get this robot built, I will never be able to "raise" it. I won't have time left, in my own life. Real "Strong AI" is unlikely to arise without a gentle "guiding hand".

Boy, public policy (especially illegitimate and disingenuous public policy) in the sciences can really screw things up for a person, can't it?

I have some notes in my inventor notebook that predate this instantiation (below) by about 2 years. Maybe I can post a scanned copy of my page(s), later. My ideas seem more extensive, and I have put my research into proper historical frame (i.e. citations of Schultz (1947?), Gliddeon (1929), and Kohler, (1890s/1900s)).

But it is nice to see somebody FINISH one of my ideas! http://psilab.ucsd.edu/publications/(journal_2013)_arianpour_(RS_An_optomechanical).pdf

(even if it isn't me!).

So, I think USB 2.0 is an adequate spine, for my vision project, but there is so much FUNCTIONAL product on the market for USB 3.0, it would be stupid for me to not start there. USB truly is universal and, crucially, it is backwards compatible. I am "in the thick" of the USB 3.1 Gen 1 and Gen 2 specs, right now, to be certain that I can use the USB 3.1 to establish the "speed limit", downstream, and then I can be certain that all upstream devices can be slower but still speak the same language.

All the USB books I have are in swahili. Well, not really. I mean, computer science writers are a special breed unto themselves, and rare is the comp-sci writer who might be nominated for a national book award, for clarity of concepts or cogent exposition. Also, these books are all old (and cheap!), and only cover up to the USB 2.0 specs. But I am getting the basics down, and the host and/or device controller pdfs are so unique for each product, I have perused a lot of them, while searching for the exact chips I want and/or need. So the USB-IF is the right website to go to for the detailed specs for developers, and I am looking at those for USB 3.1 Gen 2, right now.

Thunderbolt seems neat too, but Intel seems to hold its cards a little closer. No pdf for the controller! The pinout is interesting. Perhaps I should get some products and see how well they work? I dunno. Not made of money for such things!? I guess Thunderbolt 3 is supposed to match up with the USB-CType connector, so we'll see. Maybe I can use ThB without actually designing for it, from the start?

OK, that's enough for now. Have to go study CRCs and SOF and etc.

P.S. The TUSB7340 seems like a good host controller from Texas Instruments. I have found it in "ready-made-products" with four USB 3.0 ports. Cheap and ready to interface to PCIe. (but my desktop computer is so ancient, I only have PCI !!!! HAHAHA! Biologist Troglodyte, I AM!!! HAHAHA).

Also, I think I found a Cypress USB peripheral device controller chip that interfaces to an image sensor's parallel bus, directly. I will probably use this (skipping that large AL460 16 MB FIFO and the FIFO-to-USB bridge by FTDI Inc.), but I must read some more of that Cypress/IS-USB-device pdf, first. Also, there is a Cypress USB peripheral device controller chip that can interface to NAND Flash. Again, though, making my own NAND Flash devices seems superfluous, once the bus is set up. Off-the-shelf devices can easily be purchased and used.

Of course, using off the shelf devices places a "premium" on my high-level language skills which are, sadly, lacking at the moment. Perhaps its time I decided to study device driver design and higher-level languages?

Still plugging along, despite dragging my feet. KAC-9628 X 6 QTY (binocular X12 QTY) is still "in the works". I have noted on a wikipedia page "cordwood" construction, and I think, lacking my own PCB house or "decent multilayer PCB setup", I might utilize this method. I have found alot of my "mixed signal" PCB papers/books/pdfs, and know a bit more about analog and digital ground planes, mixed signal PCB design, and "ground loops". It shouldn't be too much of a problem, for my applicaton, because there are no "analog signals" per se. No radio signals. Just high speed digital parallel bus, for output. The "analog" portion of the circuit is "optical", i.e. the analog values of the light spots that are focussed on the sensor. The "analog" portion of the ICs (sensors) themselves are limited to ground and power pins, because the analog transistors for sensing the photogenerated charge should be free from noise, to the degree that is possible. So, a half decent design utilizing good ground planes, well regulated power supply, and decoupling cpacitors should suffice.

Also, I have some copper rivets. This should help on the "through holes".

Also, have been thinking about using Kapton (i.e. flexible PCB) material, and gluing to the top of a doble sided board. Three layers, this way? four layers this way? I dunno?

I f I do that, what sort of glue might I use? A youtube video on microscope slide PCBs uses a UV curing glue, but that requires see-through glass. Would a "evaporative curing" glue, tend to catch on fire? I think it might? Epoxy would be best?

Anybody know an EXACT material that allows gluing different PCB layers together. I do not have a oven-press for the traditional "Prepreg" PCB fabrication.

The price of getting custom board done has come down well within the reach of even very low-budget hobbyists. See our topic, "Cheap PCB stories" which starts at viewtopic.php?f=1&t=1913 . But since the topic was started over five years ago and things have improved a lot since then, you'll want to give more weight to the posts starting about 2/3 of the way down page 2. The 6502 primer has a page about custom PCBs also, at http://wilsonminesco.com/6502primer/CustomPCB.html .

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Unfortunately it looks like Radio Shack is going out of business now. I was undecided about whether this is worth starting a new topic for since it's hardly related to the 65xx; but there you have it. If anyone has more info, or wants to start a reminiscing Radio Shack topic, that should be fine. I built '02 computers on Radio Shack proto boards.

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

I've used plenty of those proto boards over the year. The sad thing was as time went on, that part of Radio Shack's business, sale of electronic parts, faded away and the sale of landfill fodder took over. They were destined to fail once that started.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I built my first SBC on a Radio Shack proto board back in 1997. Almost all of my later projects were built on Radio Shack boards too. I built a project this month on a Radio Shack board that can be seen in my photos.

I bought the boards at the Radio Shack in my town until it closed, then I bought them at the store in the next town over. That store closed too. Last month, I found that Radio Shack's website has all their boards on sale for about half price so I bought a small pile of them for future projects. I've tried various other boards but I keep coming back to the Radio Shack ones, probably just because I've been using them for so long.

_________________
- Mike Naberezny (mike@naberezny.com) http://6502.org

Hey, the mad scientist has returned. I have been a little embarrassed by the lack of "form" for my design, (and NOT a LITTLE BIT embarrassed about the lack of 6502--ALOT embarrassed--but please remember, I am a biologist, clinging to "the basics" and "what little I know about the digital/electronics " world)!

A 6502 chip WILL DEFINITELY be placed somewhere in the project.

So, today's post will have a little more "form" and some good specific questions. I believe I have advanced to the stage of requiring advice regarding power supplies and clocks.

I have a few books on USB (Jan Axelson's several works are good, but rather specific to niche purposes. One of hers' was a good introduction. The work I liked a bit better--for beginners and basics--was "USB explained", by McDowell and Seyer.

Anyhoo, I found some REALLY cheap FT245 (FTDI chips Inc) FIFO to USB modules on "the bay" (the one that comes after d-bay; I'm trying to avoid commmercial endorsements! Hahaha!). These have the benefits of having their own clocks and power supplies designed into their PCB, and two rows of simple "post terminals" for interface with a cameraboard--which is good, since electronic design is not my strong suit! However, on the FTDI inc. website, I think I prefer the FT600 (though the FT601 is "wider" I think it is extranous, since my 12 bit from the VGA sensor could have two sensors share one USB-FIFO bridge chip, but I think it is 32 wide, not 24? I forget... I just know FT600 is better for me, I think). I think FT 600 is 3.0 (and maybe even 3.1 USB gen 2 enabled)? Much faster than the FT245, I think? (For now, USB 2.0 is adequate for my design, since I have a few "choke points" where some commercially bought devices--bought on "the bay", cheaply--are only USB 2.0 enabled. Ultimately, the PCIe bus, where it plugs into the desktop computer will limit the bandwidth; but that shouldn't be a big problem, and I can always go "parallel" if I have to. I have a RAID, but it sucks. Doesn't recognize the second drive, so, it might be useless? There are many things I don't know about SATA and HDDs, and maybe I have to partition/initiallize the drives first, before the RAID can "do its thing"? But 2 drives at 3 TB a piece is a healthy amount for my robot, to start with. Maybe the RAOD controller card is bad? Maybe the software doesn't like my circa-2006 dell desktop, windows XP? I dunno... no time to care, much, but my oscilloscope is up and running, so maybe I can finagle it to read a differential signal from a SATA wire? It is such an ancient HP dinosaur; 25 MHz bandwidth! I forget freq of SATA/eSATA signals? probably higher!)

Also, I bought one AL460 FIFO, about a year ago, and though kind of pricey, I think it is probably the best fit for a fella like me,; with my mindset, and limited abilities. Performance plus ease of use, I think? Also, it is the biggest FIFO memory in the biz, or at least it was. Their website promised a AL462 would be out soon , and would be larger. I e-mailed and asked when's the release, but got no reply. So imagine my surprise when, in feb or march 2017, I googled FIFOs and found a press report from Oct 2016 announcing the bigger chip was available (same/similar price).

The heat sink thing was throwing me. I hadn't thought of heat budget. I was thinking I could drill a hole and wedge a penny in there! Hahaha! Then a good solid (1982 or earlier, penny) copper surface to attach a heat sink to! Peltier, maybe? Also, I think the bits are 16 wide on the AL460, so, 4 traces will be wasted. I thought of sharing FIFOs, cleaving off the four bits and "glomming" them onto some other sensor's bus, to save on chip count, but it didn't seem worth the trouble. Besides, By the time I get my KAC image sensors working, I will move onto another cheap sensor (the KACs are already obsolete, as well as the company that sold them, Kodak; I hear it rises from the ashes, or, at least, its name does. Brand. Logo. Poor old George Eastman; what's a guy got to do to get some good prints of his vacation? (Eastman was a bank clerk on vacation, and thought cameras needed to be easier and portable, hence, Kodak was born. I think the CMOS-imagers line was actually from National Semiconductor? I forget?)). Does anybody know if AverLogic is a good company? Fabless! Taiwan?

Anyhoo, my local radioshack had a three-to-five-year-long "going out of business" sale. Several times I went in there, they said "Seriously, this is the last month". Last I heard, though the one in Barre Vermont (North and East and Central) is gone (verified with mine own eyes!), the one in Burlington still exists, I think? I was told they did? Maybe national radioshack finally went kaput? I dunno?

I will do a modified-cordwood, I think. I have some very small rivets and some very small drill bits, though the collet on my very small push drill is funky. I have been gluing the bits into wood blocks/dowels, and swiveling them, handtool fashion. Not too hard to do, once the hole gets started.

I also purchased some very thin FR4 double sided laminates. I have some 1/16" diam, 5/8" long metal pins/dowels, for registration of designs, and such. That should work well, I think. So, I can build a board on the thin laminate, rivet the through holes, with a splash of solder, over top of the rivets, for good measure, then, build the bottom board, drill a final couple of holes--for the bigger copper rivets, to connect some power planes/GND planes--then with non-conductive epoxy (any brand suggestions? ideas are welcome! Thanks!) I can smoosh the two double sided board together, let dry. And a day later, I can drill/rivet the final holes, smooth/debur/solder them, and voila!

But, putting "ridiculous-me", and my "FINE, I'll do it myself" attitude, aside, the real question I have, today, is about powering , essentially, six different VGA cameras, and the FIFOs and FIFO-USB bridge chips. A repurposed computer power-supply? Maybe. EEVBlog (on "the-tube", the one that comes after the t-tube, and could be spelled u-tube; no endorsements) has an interesting design for building a lab power supply. I thought of just ripping it off, since he is offerring it, and all of his reasoning processes, for free!

Also, clocking such devices might be a concern, too. So, some questions might follow this post, and the next post. Next post will be my early power supply design (I have like ten handwritten (more like "scrawled", illegibly) pages on the subject, and more reading and planning than I can really handle or process! But the sheet I photographed today is a synopsis of my thoughts from today and yesterday; there is proabbly a much better answer out there. Perhaps you have it? If you do, please let me know.)

If you're trying to make a vision system, you don't want to be distracted by learning about power supply design - use the most accessible and affordable ready made system you can.

I've recently been saying that we all need three things to succeed in our projects: time, money, and enthusiasm. (Which is why negativity annoys me so much.) But perhaps I should add a fourth: focus, or discipline. It doesn't matter how much time you have if you spend it on the wrong things! And if you can't spend a reasonable chunk of time on a single given task, then there's a severe limit to how difficult a task you can successfully tackle.