-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
Dear luke, Although I can't find the post, I do remember you writing about the fex file and how hard it is to modify it. I've encountered some similar problems when programming and I've always thought of doing things like Donald Ervin Knuth recommended. That would be, I would use literate programming. I've not read the official paper through because I did not know it was available (and I did not buy the book which I did know about). As it is, when I went looking for you for some information on the subject I found more then I had expected for you. A paper: http://www.literateprogramming.com/knuthweb.pdf Website: http://www.literateprogramming.com/ Criticism: http://akkartik.name/post/literate-programming
The criticism shows how this method can be taken to an extreme and be a complete waste of everyones time.
Hope it helps, David
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Thu, Jul 18, 2019 at 4:28 AM David Niklas doark@mail.com wrote:
A paper: http://www.literateprogramming.com/knuthweb.pdf Website: http://www.literateprogramming.com/ Criticism: http://akkartik.name/post/literate-programming
The criticism shows how this method can be taken to an extreme and be a complete waste of everyones time.
"Literally the first line of code shown is a macro to access a field for presumably a struct whose definition — whose very type name — we haven't even seen yet. "
i stopped reading the akkartik.name post at this point. use of sarcasm to criticise one of the most highly regarded and influential people in computer science is just going to piss people off.
#defines are *KNOWN* to be a pre-processor stage. if you don't know that, get f*****g educated for f***'s sake.
i'm getting quite alarmed at the amount of blatant ignorance and misinformation about c that seems to be propagating through the internet at the moment.
"Let us change our traditional attitude to the construction of programs: Instead of imagining that our main task is to instruct a computer what to do, let us concentrate rather on explaining to human beings what we want a computer to do."
this is why there is an entrepreneur that flatly refuses to employ computer science majors: he employs ENGLISH LANGUAGE majors and PAYS to have them trained to program.
the reason is extremely simple: they make better communicators, they make better team co-workers, they write better documentation and they write maintainable and understandable code.
l.
Hope it helps, David -----BEGIN PGP SIGNATURE-----
iQIzBAEBCAAdFiEEL2N7+xWmVOJDQxWGm3XCrhg2YP8FAl0v5yIACgkQm3XCrhg2 YP8cLBAA6ujfFmCm6sFq73l5WLf3L1GTABduglRrkUj5LwhnyTFhII/fiTy+mNLS d6CkKSuJp3iu/wizYy4Pyb/EIt6n8D4dG6Vq7nGTQWIABS/4KW6n5c2iIFmQajXK /LkRkCdFvMJHF9Ok1XutvCNTFQm5EBh1RdZQ2buERLkOJpjcfGWRo4SztouQzhpW AK0ekRRxqolQOuS1hKeRG+rUb6hnGj+aMQcpYVcMFIKmzNGsOAVsA2Hkxyc5m6ZB Yn1sU7IQ/zia9nirfoeOZWA70/iSpWUn13UeU7m84Mw3NPoJ8utsL5zqJ1PtiITG 8PXOH1FIQZhdqpqvOb+h79KdGs1U4EAHEMIYHoSGXdsK74foyBcN+maKK0E0gMGS Cb5f7RtC2t8tNZ+0pthcwOlOUYqtoiWyIOeWCX1lReknzVcdO4rgOUu45WK2amCC RCb+O5w6Tvptq6RFH0BJypI1SbgVRi5n8S3dKSPlXFfVecsdPU91vUERpqlFpGRH qWFNfNsLH9MgKg3LNceN+Rm2SRqnwKpxQFoFv/eeeBiH6EpOXrJgPKWzsV8CUtBo 8iFqMMvXccNWyHnyXiUB5lT+ysBMV6UdD+7njRLc+aPWG4ngkbqkZ0Wevdh3qC8g QaWpdjMDCfZ8uou+yVk6b1vh92lT/9H0S58QE49JHkok83VpxyU= =9fnZ -----END PGP SIGNATURE----- _______________________________________________ arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
On Thu, 18 Jul 2019 06:35:53 +0100 Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Thu, Jul 18, 2019 at 4:28 AM David Niklas doark@mail.com wrote:
A paper: http://www.literateprogramming.com/knuthweb.pdf Website: http://www.literateprogramming.com/ Criticism: http://akkartik.name/post/literate-programming
The criticism shows how this method can be taken to an extreme and be a complete waste of everyones time.
"Literally the first line of code shown is a macro to access a field for presumably a struct whose definition — whose very type name — we haven't even seen yet. "
i stopped reading the akkartik.name post at this point. use of sarcasm to criticise one of the most highly regarded and influential people in computer science is just going to piss people off.
#defines are *KNOWN* to be a pre-processor stage. if you don't know that, get f*****g educated for f***'s sake.
i'm getting quite alarmed at the amount of blatant ignorance and misinformation about c that seems to be propagating through the internet at the moment.
"Let us change our traditional attitude to the construction of programs: Instead of imagining that our main task is to instruct a computer what to do, let us concentrate rather on explaining to human beings what we want a computer to do."
this is why there is an entrepreneur that flatly refuses to employ computer science majors: he employs ENGLISH LANGUAGE majors and PAYS to have them trained to program.
the reason is extremely simple: they make better communicators, they make better team co-workers, they write better documentation and they write maintainable and understandable code.
l.
Actually, I just skimmed the article and I was referring to the part of his post where he criticized the documenting of basic includes (we all know what stdio.h is). In the case of defines I assumed that he was talking about things like the common MAX. I tend to assume people are arguing intelligently. Next time I'll be more careful when referring you.
But going back to the point of my original email, would literate programming be any good to you in this case?
Sincerely, David
On Thu, Jul 18, 2019 at 3:44 PM David Niklas doark@mail.com wrote:
But going back to the point of my original email, would literate programming be any good to you in this case?
honestly, just other people helping out - *actually* helping out with u-boot, linux kernel etc. - would be a huge relief.
l.
On Thursday, July 18, 2019, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
honestly, just other people helping out - *actually* helping out with u-boot, linux kernel etc. - would be a huge relief.
I've been itching to get involved since I picked up my prototype in England but waiting (too patiently? Should I have bothered/reminded you about it earlier?) for the video how-to you promised to make on how to non-destructively plug the unenclosed CPU card (without the guides) into the micro-desktop case.
Would the cable-only setup be a valid way to get started? Should I expect the micro HDMI plug to be alive?
I guess the earliest debug output (u-boot?) will likely be on the serial console available through the EOMA68 connector.
On Thu, Jul 18, 2019 at 8:13 PM Richard Wilbur richard.wilbur@gmail.com wrote:
On Thursday, July 18, 2019, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
honestly, just other people helping out - *actually* helping out with u-boot, linux kernel etc. - would be a huge relief.
I've been itching to get involved since I picked up my prototype in England but waiting (too patiently? Should I have bothered/reminded you about it earlier?)
probably :)
for the video how-to you promised to make on how to non-destructively plug the unenclosed CPU card (without the guides) into the micro-desktop case.
ah, rats.
ok. it's fine if you get it wrong as long as it's not powered up. try it and send me some photos.
* put MD PCB on a flat surface * processor goes face-up * push PCB in until it meets pins * make sure that the gap amount on each side (left, right) is equal - about 3mm * make sure it's about 1mm off the MD PCB * make sure it's DEAD level * push in GENTLY, making sure that the force applied does not make the Card PCB "rise up" (be anything other than dead-level)
if you then inspect into the tunnel, you should see no pins. actually it's perfectly fine to *deliberately* get it wrong, as long as you don't apply power.
send a photo looking into the tunnel, ok?
Would the cable-only setup be a valid way to get started?
actually that's a good test - you should see the standard allwinner USB device with "lsub"
Should I expect the micro HDMI plug to be alive?
no, not really.
I guess the earliest debug output (u-boot?) will likely be on the serial console available through the EOMA68 connector.
yes. the schematics PDF shows the pinouts. in the past i've stuffed serial port wires directly into the PCMCIA header, *and* soldered direct to a socket (!) - soldering to the MD PCB is a *lot* easier.
l.
Thanks, Luke. Looking forward to following your instructions next week after I'm back home. On holiday in a different timezone right now.
On Sat, Jul 20, 2019 at 6:43 AM Richard Wilbur richard.wilbur@gmail.com wrote:
Thanks, Luke. Looking forward to following your instructions next week after I'm back home. On holiday in a different timezone right now.
cool!
Ok these are the microdesktop with a USBUART attached, you should be able to zoom in and see tx rx on the UART PCB.
You will find quite often that you get a ground loop between the main DC PSU and a laptop's PSU which is often not earthed, over the USB, through the USBUART, to the MD PCB that will, on powerup of the MD, cause the USBUART to either generate random data, or cause it to spike and lock up entirely, or in some cases of extremely cheap FT232 USBUARTs, cause them permanent irrevocable damage.
This is just how it is.
I buy lots of them, and in cases where difficulties occur unplug them before powerup. Of course, uboot needs to then be configured not to autostart too fast, and also you miss the very early uboot-spl debug messages, ho hum.
You need a 5.5mm jack with a 2.1mm centre hole, middle is +ve (aka "pin positive").
Anywhere between 7 and 21v is fine, minimum 1A, 1 5A is better. One of those multi power thingies is fine (5 7.5 9 12v you know the ones I mean, get them at RadioShack... or at least you used to), just make absolutely sure outer is -ve and inner (pin) is -ve.
You almost certainly have a random 12VDC PSU somewhere in the house already, for a toy, a charger, or many 12 VGA monitors use a 5.5mm jack, it's an extremely common standard, can take up to around 4 to 5A no problem.
If you have a multimeter for goodness sake use it.
The Card can be powered up whilst attached to USB OTG however although power gets through the PCMCIA connector there is a SY6280 current control IC set at 1A which acts as a diode in this case. You MUST therefore power the PCB from the DC Jack in order for the USBUART to work.
DO NOT wire up the 3.3v supply to the USB UART. ONLY solder on 3 wires: GND TX RX. Those work by sinking current to GND and are otherwise floating, I believe.
However they *may* be tied to 3.3v pullups actually on the USBUART and I have had this be sufficient in some cases to actually power parts of the A20 processor (!!) even when otherwise fully powered down.
L.
On Mon, Jul 22, 2019 at 10:50:24AM +0100, Luke Kenneth Casson Leighton wrote:
Ok these are the microdesktop with a USBUART attached, you should be able to zoom in and see tx rx on the UART PCB.
You will find quite often that you get a ground loop between the main DC PSU and a laptop's PSU which is often not earthed, over the USB, through the USBUART, to the MD PCB that will, on powerup of the MD, cause the USBUART to either generate random data, or cause it to spike and lock up entirely, or in some cases of extremely cheap FT232 USBUARTs, cause them permanent irrevocable damage.
This is just how it is.
I buy lots of them, and in cases where difficulties occur unplug them before powerup. Of course, uboot needs to then be configured not to autostart too fast, and also you miss the very early uboot-spl debug messages, ho hum.
You need a 5.5mm jack with a 2.1mm centre hole, middle is +ve (aka "pin positive").
Anywhere between 7 and 21v is fine, minimum 1A, 1 5A is better. One of those multi power thingies is fine (5 7.5 9 12v you know the ones I mean, get them at RadioShack... or at least you used to), just make absolutely sure outer is -ve and inner (pin) is -ve.
Presumably one of these "-ve"s should be "+ve".
-- hendrik
You almost certainly have a random 12VDC PSU somewhere in the house already, for a toy, a charger, or many 12 VGA monitors use a 5.5mm jack, it's an extremely common standard, can take up to around 4 to 5A no problem.
If you have a multimeter for goodness sake use it.
The Card can be powered up whilst attached to USB OTG however although power gets through the PCMCIA connector there is a SY6280 current control IC set at 1A which acts as a diode in this case. You MUST therefore power the PCB from the DC Jack in order for the USBUART to work.
DO NOT wire up the 3.3v supply to the USB UART. ONLY solder on 3 wires: GND TX RX. Those work by sinking current to GND and are otherwise floating, I believe.
However they *may* be tied to 3.3v pullups actually on the USBUART and I have had this be sufficient in some cases to actually power parts of the A20 processor (!!) even when otherwise fully powered down.
L.
--
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Mon, Jul 22, 2019 at 3:24 PM Hendrik Boom hendrik@topoi.pooq.com wrote:
^^^ } } You need a 5.5mm jack with a 2.1mm centre hole, middle is +ve (aka "pin } } positive"). ^^^^
Presumably one of these "-ve"s should be "+ve".
thx.
arm-netbook@lists.phcomp.co.uk
-
David Niklas -
Hendrik Boom -
Luke Kenneth Casson Leighton -
Richard Wilbur