OESF Portables Forum
Everything Else => Distros, Development, and Model Specific Forums => Archived Forums => PocketPenguin => Topic started by: Da_Blitz on July 27, 2006, 11:57:14 pm
-
Well we finally have our own fourm for discussion and i intend to break up the monolithic highly technical posts in to as many smally highly technical post as i can
Anyway here is where we will keep an updated list of whats in the pocket penguine as well as a place for FAQs about the project
Q: What is it:
A: This is our attempt to build our own PDA to our own standards. allowing us complete control over the hardware in an attempt to build a totally open PDA platform [/artistic babble]
Q: What is in it:
A: as best summed up by ferret simpson:
CPU: Freescale iMX30 - 533mhz ARM11
RPU: Freescale MXC300-30 3G ARM11 with StarCore DSP. 533mhz.
RChipset: Freescale RFX300-20 3G Radio Platform.
RAM: 2Gb (512MB) Laptop DDR266 Memory in 2x1Gb chips.
HDD: Internal ATA CF/MD interface, User provided Card
SD: 2x 4bit SD slot, SDIO capable with 3rd Party Commercial stack.
NAND: 16MB RO for Bootloader only. Will kExec any Kernel from a variety of Media.
RFlash/Mem: Indeterminate NAND and SDRAM. (16Mb Flash 128MB ram minimum)
Wifi: Zydas ZD1211 chip on CF16: 802.11G-54, WPA.
BT: Unknown Bluetooth V2.0 on USB-1.1
Rad: RPU connected over SPI, Audio, and USB480
IRDA: Fir?
Keyb: Internal Backlit KBD,
Mous: Internal Trackball with "Select" function and illumination,
DPAD: Internal with OK/Canc.
Roll: On Landscape Rear.
L/R On Landscape Rear
ACC: 3D Accelerometer.
BIO: swipe-bar Fingerprint scanner on SPI.
AUD: 5.1 with Microphone, and BlueTooth routing.
MSC: 3.5-4 inch, VGA, with integrated Resistive Touch Layer.
SSC: 4bit Serial on rear of device, for Phone display and viewfinder for Camera shots.
CAM: On main screen panel, above TFT in portrait mode. Supports Video and Stills.
CPLD to act as Programmable ASIC for SPI interconnection to devices.
Possible MicroProcessor to control Buttons and roller for use with phone when PDA disabled.
External full-size 480MBPS USB2 port.
Video out, Possibly 768i HDTV compatible, will be used for connection to external monitor.
Socket for network operator SIM card.
Security:
Key-signing of OS to prevent Unauthorised OS replacement.
Possible GSM trace functions including local MAC addresses and signal tower strengths.
Biometric identification of allowed users.
1800mAh Battery: Easily available Digital Videocam model.
Customised Aluminium case.
Q: Whos working on it (offically)
A: Da_Blitz (post to be added to the list)
Q: I have very few technical skills but want to help, how can i do so
A: Poke fun at our ideas as well as come up with some of your own, we intend to inovate with this platform as we dont really care about profits so we can chuck nearly anything we want in it.
If you want to help us futher than help contribute to the wiki (see Da_Blitz's tag) or start reading the chip documentation, you dont have to know electronics to understand the chips (it helps) a basic knowlge of how computers work is fine
Q: I have a highly technical background in electronics/kernel hacking how can i help
A: Introduce yourself and have a look at what we need done, i will try and maintain a list of what needs to be done but dont count on its reliability (check first). we will need driver writers and people with a background in security (i have a side project or two we will need for this thing)
if you have experince with mobile phones or GSM stacks please get in contact with us as we could really use someone in that area for writing a bit of firmware
Q: will this cost less than my car? what are you wildly estimating to be the shipping cost (with the understanding that if you answer this this early you won't be held to the figure-- looking at the specs, I'm thinking I might have to save a few pennies..)
A: We are aiming for the price of a c3x00 when fully loaded (ie full ram and storage) shipping should be included in that cost whoever we dont expect shipping costs to play a big part in the final price
Q: when do you expect to finalize the design and begin production?
A: as soon as posible, current goal is no latter than the end of 2006 but dont hold us to that
Q: what standard os is it shipping with? I'm guessing something built on OE?
A: Inital efforts will be focused on linux. at this point we may not ship with an OS on board as all you willl have to do to load an OS is use gpg to sign a binary and copy that binary to a CF/SD card for it to boot. therefore we may ship with several options on CD that a user may chose from and load to a CF/SD card to try out in a LiveCD fassion
Q: the screen description is vague. Still looking for parts? I would hope to see the 4" transfective from the zaurus 6000, or or better... mostly, outdoor light capability would be nice.
A: Yes we are still looking for parts or more to be more presice suppliers, at the moment we are aiming for 640x480 4" transreflective
Q: Whats the holdup on starting the design rather than talking about it
A: Before we can really design this thing we have to know which chips we are using in the final design, some of the critical things that we need to know what we are using are the screen, Wifi, RPU (radio processing unit) and keyboard.
-
quettion 1: will this cost less than my car? what are you wildly estimating to be the shipping cost (with the understanding that if you answer this this early you won't be held to the figure-- looking at the specs, I'm thinking I might have to save a few pennies..)
question 2: when do you expect to finalize the design and begin production?
question 3: what standard os is it shipping with? I'm guessing something built on OE?
question 4: the screen description is vague. Still looking for parts? I would hope to see the 4" transfective from the zaurus 6000, or or better... mostly, outdoor light capability would be nice.
-
Updated FAQ
-
I meant cost, shipped as described, sorry. approx 6-700 usd? damn, that would be absolutely amazing... would blow away everything on the market.
I certainly am intereted !
-
RPU? I thought we'd settled on the Freescale MXC300-30 - Well supported because Freescale and Motorola love Linux, Cheap, and with a matching 3G Chipset. . .
-
i used RPU because i was under the impresion that was what you used to refer to the MXC300, ah who cares, i just copied and pasted it anyway
btw is this in any other lin phones we know of because that would only sweeten the deal
-
i used RPU because i was under the impresion that was what you used to refer to the MXC300, ah who cares, i just copied and pasted it anyway
btw is this in any other lin phones we know of because that would only sweeten the deal
[div align=\"right\"][a href=\"index.php?act=findpost&pid=136723\"][{POST_SNAPBACK}][/a][/div]
The generally accepted term for that is 'Baseband Processor'
-
I know that RPU is the MXC, since I DID write it, but you said it's a critical thing we need to decide on. >.< WE ALREADY DECIDED!!!!!!!!!!
Baseband processor is too ruddy long for me. I like Acronyms, XP
Can anyone confirm the CPU's used in the Moto Smartphones? I'd expect them to be using their own baseband and applications processors. . .
-
I know that RPU is the MXC, since I DID write it, but you said it's a critical thing we need to decide on. >.< WE ALREADY DECIDED!!!!!!!!!!
Baseband processor is too ruddy long for me. I like Acronyms, XP
Can anyone confirm the CPU's used in the Moto Smartphones? I'd expect them to be using their own baseband and applications processors. . .
[div align=\"right\"][{POST_SNAPBACK}][/a][/div] (http://index.php?act=findpost&pid=136737\")
That's why people call it a 'BP', and see [a href=\"http://wiki.openezx.org]http://wiki.openezx.org[/url]
-
Hmmm. . .
None of the Motorola phones uses Motorola Processors.
On the flip side, Nokia use the MXC300-30 in the S-60 running Symbian. . . So the chip's been proven, if nothing else.
-
very intresting, do you have any more info on the BP?
-
Not yet, that's all I've seen so far. I'm surprised though, that Motorola still don't use their own processors for their phones. . .
-
Not yet, that's all I've seen so far. I'm surprised though, that Motorola still don't use their own processors for their phones. . .
[div align=\"right\"][a href=\"index.php?act=findpost&pid=136877\"][{POST_SNAPBACK}][/a][/div]
Not really, sharp doesn't use their own ARM cpus in their PDAs.
-
I didn't know they manufactured them. Still, If they do, as you say. . Then I'm surprised that they don't either!
AMD sell a small range of Development PDA's, using their own processors. It's generally considered worrying if you don't use your own products. After all, if you don't trust them, or think they're good enough. . . Why should anyone else?
After all, if you're a plumber or a computer techician, you don't generally get another company to provide you services, when you can provide them to yourself for a lower price?
It's one reason why I worry about Microsoft servers. Last time I saw an error on Microsoft.com, it was an Apache host. . . Why don't they use IIS?
-
they do use ISS but have apache cache servers that if i remeber correctly are not run by them, it was on slashdot awhlil back
funny thing to note is that thier cluster of ISS machines are rebooted every 5-10 minutes (there was a video documentry with the microsoft guy in charge of the website) but it is still a great setup (i have yet to see them down but then again its not my fav site
when it comes to using your own products, no one really cares. few people look inside thier own equipment but it does matter to some chip compaines more than others. it might also be intresting to not that from what i can see on the outside most of those companies seperate thier ic buisness off from the rest of the company so it might be a contributing factor
-
KK, well back on topic: We should set a definate date to finalise the hardware specification, and I think it should be a little earlier than you first estimated. If we finalise in December-ish, then the product won't be availbe for a good few months after that, by which time the list of available hardware and how it's interfaced will have changed and we'll have to begin again.
I'd say we should freeze the hardware list in early october-time for ALL developers. That would put us on track for something along the lines of a Q1-2/2K7 release. . . Right? But definately, we should have a Date, and GMT time so there's no confusion. Any posts after that can then be reasonably ignored, without "But you never said. . .!"
-
mmm, if you had read the link for the pcb people i gave out then you would relise that they are "rapid prototype specilists", i have had boards back from them in 1 week from the order date.
hardware gets frozen in the next couple of weeks, i dont have time to fiddle around for that long, the PCB design will take me alot of time
you seem to have in your head that we will be doing a limited production run, in fact there will be no production run. what happens is i give you the files, you send it off to the rapid prototyping company and they send you back a bouard, that way i dont have to collect any money and therefore i cant run away with it all and leave you guys with nothing while i buy a cluster of Z's
i expect it will take them 2 weeks to build the board and ship it from the day i send my order out.
-
mmm, if you had read the link for the pcb people i gave out then you would relise that they are "rapid prototype specilists", i have had boards back from them in 1 week from the order date.
hardware gets frozen in the next couple of weeks, i dont have time to fiddle around for that long, the PCB design will take me alot of time
you seem to have in your head that we will be doing a limited production run, in fact there will be no production run. what happens is i give you the files, you send it off to the rapid prototyping company and they send you back a bouard, that way i dont have to collect any money and therefore i cant run away with it all and leave you guys with nothing while i buy a cluster of Z's
i expect it will take them 2 weeks to build the board and ship it from the day i send my order out.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=137116\"][{POST_SNAPBACK}][/a][/div]
If you sent me just a board, I'd be pretty disapointed
-
heh, The point of this is that we'll be buying the case etc separately ADF. We'll design it, and then there'll be a universal blueprint that we can have made up to order.
This way, we don't have to get FCC verification, or the equivalent in every other country we want to ship it too.
DB, I was making the point that it usually takes some time to put together the schematic, get one built, bug test it, get a second prototype built, bugtest that.
-
hopefully there wont be a second prototype, however the funds avalible just went up a bit
i am aware of that, but keep in mind that i am running on my own schedule which i havent made public and most likly wont (so you cant hold me to it). i am intrested to see how quickly i can achive everything
-
Security:
Key-signing of OS to prevent Unauthorised OS replacement.
I hope I'm not the only one who sees a problem with this.
-
i would take a guess and say "yes you are"
i dont really care, i have said all along that the user will hold the keys, not me. i will keep your keys in storage if you email them to me but that will be a security risk (if posible it will be done with a crypted email using pgp)
i will not ship with the keys in place, its optional but i will tell you how to make the device unhackable as posible.. you will be in full control. but if you F**** it up you will not be able to do anything with the device without compleatly lossing all data on the device.
also i have recently seen an in kernel key mechinism that will only allow signed kernel modules to run, its currently in fedora core 5 and has intresting posibilities for helping to enhance security
-
I can see about 20 problems with the keyboard, But the security? No. . I don't see a problem. . .
Generally, if you can assemble a PocketPenguin from the component parts, and you can use what is likely to be a Slackware or Gentoo based Operating system. . .
Then you should really be able to handle basic security. . .
-
that brings up an intresting point, out of the box you will need to know somthing to get it running, even if thats just which file to put on an SD card on first boot
of course the security i am trying to provide (ie against an offline attack) is not somthing i expect most people how know about securtiy to know how to implement properly
actually i would say that the key signing stuff comes more under the integrity column rather than security, while it is a way to help improve security it is an entity in its own right.
-
Ok, the project seems to have ground completely to a halt, with us all getting sidetracked by laptops and watches and all sorts of other fun!
Where does it stand right now, on the PocketPenguin!?
-
well its basically about chipsets, i know most of the chipsets i want but if i wanted to i could start wiring everything together and add things as required
pcb software is also somthing i have been wondering about as i do have some top end sotware but i cant exacttly let you guys have a copy if you want to modifiy it (basically not open source ) i could do what the open graphics card guys do however and print the design to PDF but i think that wouldnt be nearly as usefull to someone who wonts to submit a patch
as it is however i think i would be the only one working on it anyway
plus i might have found a radio engineer who could give me some advice (for UTMS and such)
-
maybe it would be a good idea to make two versions of the pocketpenguin, one would be just a pda (without the baseband processor etc.) and the other would be a phone/pda. i just dont believe in pda/phone combinations because the batteries are still not up for it. i have a motorola a925 and i often find myself wasting two batteries in less than 8 hours which basically leaves me phoneless. and i need my phone to be on 24/7. i dont understand why should we (well, mostly da_blitz :-> ) try to build a proper phone/pda. there are lots of good phones out there, lets just concentrate on the pda/pocketpc aspect of the thing. i for one would love the price to go down to about 500$ and a form factor thats between the zaurii and the sigmarion 3.
oh yeah....
AND A BIG, FAT KAZILLION MaH BATTERY!!!! *_*
-
yes i would agree with you but i am still of the mind that if you dont want it then dont specify the chipset when you order it and it wont be soldered in
some things are unavodable, the gps chipset will in use the same chipset so some radios chips will be required as well as the wifi and such sharing antennas, but the option to drop most of the chips there and i only have to make one design then
i have a htc tytn now which i got just in time (fried the Z) and i get good battery life on it but i think the real killer is charging off of USB and possibly USB host (still researching that one as well as the posibility it has a disabled gps chip in it)
-
sharing antennas? does that mean that it isnt possible to use wifi and gps at the same time? do you have a final candidate for a gps chip?
-
all an antenna is is a peice of wire, for maximum power transfer it should be 1/4 the wavlength long so an optimol wifi antenna is diffrent in length from a fm radio
dont worry everything will work at the same time i hope
for the chipsets i am getting very sneaky (perhaps too sneaky) and using the same chipsets as those that have linux ports already if posible. otherwise i do have 2 chipsets in mind.
-
Q: when do you expect to finalize the design and begin production?
A: as soon as posible, current goal is no latter than the end of 2006 but dont hold us to that
Unfortunately, I don't have enough time to follow the very interesting discussions about the Pocketpenguin. I'd love to help and give my ideas etc. but have too many own projects running...
When do you plan to be ready, now that end of 2006 has passed
All the features sound very promising. However, in order to produce such a device in even small amounts, you need to have some very sophisticated and probably expensive production tools. So how do you plan to keep the price for the device as low as a few hundered dollars?
Just curious. Of course I understand if you can't give details...
Now that Sharp accounced the end of the Zaurus line, even if there will be used units for a long time to come, it may be good to reorientate slowly.
daniel
-
well in an ideal world an OSS company would sponser me but that wont be happening
so i am going to have to do it the tough way, it means the prototypes wont be fully populated, only with what will be tested and i will be hand soldering and butchering the prottypes (ie recycling them) to keep costs down
i also hawe a fpga board on the way (http://cgi.ebay.com/XILINX-USB-FPGA-BOARD-SPARTAN-3-200K-ETHERNET-WEBSERVER_W0QQitemZ220069714762QQcategoryZ50913QQcmdZViewItemQQisPrinterFriendly
1QQpvZ1) that will allow me to build virtual hardware for some things (ie looking at buses and doing logic analysis) and emulate some parts in softhardware
i entend for it to cost as much as the top range PDAs, but for each country the price is diffrent and for each individual, what is considered top end is diffrent
from my perspective i am looking at $1000 AUD for a final price as a maximum but even making 100 units will bring down the price, in fact 20 or more units will shave about $20 to $40 off of the price due to a lower retooling cost for PCB creation
at the moment i have a rapid prtotype manufacturer in china that i use (the US stuff is 10x more expensive, gives a lowwer quality product and forces me to wait 3 months instead of 3 weeks. not to mention the chinese people have better english skills than i do)
so they make it and i test it, repeat. you guys wont pay for dev costs in the final price. you will pay ONLY for raw hardware costs. so that saves a bit
but basically it comes down to my ability to debug without fancy tools and lukily the fact that the hard stuff is only RF which there are established practicies for and high speed busses which includes the ddr ram which is accounted for as well (more best practicies)
but yeah it comes down to skill and i hope i have it. i just have to remeber that along time ago when SDram came out they didnt have all these fancy tools so they had to inovate, the fpga helps but the thing that helps the most is old electronic magazines and infomation on old design practicies and debuggining teqnieqes
sure i wont have a 100Mhz ossciliscope but i will have a high speed logic analyser that i could wire a high speed ADC to to make a Digital storage ossciliscope
-
good luck!
-
- Is the first post in this thread accruate wrt to hardware at this time?
-
fairly, things change and i drop some things i consider "minor" and add them back all the time, it depends on the moment.
at least its a rough draft, basically things that change are mainly cameras and "gadget" features. the core ram and cpu stuff as well as conectivity options will stay the same
i would love to throw a small fpga in there but it will most likly be a cpld instead and non hackable, upgradable yes but not somthing you can change at a whim (it will be doing alot of interfacing and "language" conversion
biometrics look to be dropped as there are better ways of authenticating and i belive it will be abused. there is an option however that if a camera is included that it could do iris verefication.
anyone still want IR? also mini sd or full sized, i cant remeber every ones postion on that (or mine for that matter) or one full sized and one micro?. the micros are ones that shouldnt be pulled out to ofteen as they are a PITA to remove (ie this is the rootfs)
-
fairly, things change and i drop some things i consider "minor" and add them back all the time, it depends on the moment.
at least its a rough draft, basically things that change are mainly cameras and "gadget" features. the core ram and cpu stuff as well as conectivity options will stay the same
i would love to throw a small fpga in there but it will most likly be a cpld instead and non hackable, upgradable yes but not somthing you can change at a whim (it will be doing alot of interfacing and "language" conversion
biometrics look to be dropped as there are better ways of authenticating and i belive it will be abused. there is an option however that if a camera is included that it could do iris verefication.
anyone still want IR? also mini sd or full sized, i cant remeber every ones postion on that (or mine for that matter) or one full sized and one micro?. the micros are ones that shouldnt be pulled out to ofteen as they are a PITA to remove (ie this is the rootfs)
[div align=\"right\"][a href=\"index.php?act=findpost&pid=152641\"][{POST_SNAPBACK}][/a][/div]
1 gig micro-sd ( or even 512MB, any size over 128MB should be enough) holding the rootfs and the basic operating system files sounds good. for security reasons you can always do a crc32 checksum on all the data on the card to verify it has not been tampered with or even sign it with a private key and have the public key in the processor's memory. that way you can have upgrades released and still be able to make sure it is secure. upgrading the operating system would be much easier since all you have to do is pop the card into any card reader and do a image onto it without having to worry about flashing the device itself (this can make it really hard to brick your device), you can even have two partitions on the card, one for the operating system that can be checked for integrity and another that is user rw and holds the /home directory.
for lower volume production the micro-sd would actually be cheaper than buying flash chips and having that much more board design work, since you can get 512MB micro-sd for less than 10 USD.
the whole design goal would be to make the hardware as flexible as possible while still making it simple to design and be cost effective. (when you get down to it these are not easy goals to hit)
-
i like it because it futre proofs it
-
- Hard to argue with stampsm, but I'd like one full-size slot as well, unless we can get three or four micros and do that Raid thing. (wolfish grin) I think I prolly want a full 'cause they go to 4Gb, and with SDIO there are non-memory peripherals available for it. Not that I own any, mind you, because the Z uses CF for this, but the PP won't have CF. Seems like a nice thing to have, though.
-
the huge flash on the iphone is tempting but its somthing that can bite you on the arse unless you go all out (like apple did). at lesat with the slots you can chose or reuse what you have
im guessing the full sized will be the external one might as well make the micro one externaly acsesible for devs or does that partially defeat the security, must think about that. probelly not however
CF is nice and i cant cemeber where i am at with it (if its in or not) i think its really a case of lets see how much room we have left. if it did go in it would most likly not be removable for some reason (there was somthing technical holding it back, somthing to do with the ata mode)
-
the huge flash on the iphone is tempting but its somthing that can bite you on the arse unless you go all out (like apple did). at lesat with the slots you can chose or reuse what you have
im guessing the full sized will be the external one might as well make the micro one externaly acsesible for devs or does that partially defeat the security, must think about that. probelly not however
CF is nice and i cant cemeber where i am at with it (if its in or not) i think its really a case of lets see how much room we have left. if it did go in it would most likly not be removable for some reason (there was somthing technical holding it back, somthing to do with the ata mode)
[div align=\"right\"][a href=\"index.php?act=findpost&pid=152878\"][{POST_SNAPBACK}][/a][/div]
make the micro one an internal one like on the ROKR cell phone. it makes for a very small size holder with no mechanics needed to eject it. if it has operating system files you do not want the user to "accidentially" take it out while it is running anyways. it is easy to make it secure while still making it easy to remove if you have an asymetric encryption system(like pgp) that has a private and public key running in the small firmware (rom) inside the device. you just check the keys on the main files or file clusters and if they match up with the private key you know the files have not been tampered with (well there are ways around this, but NO security is foolproof). if the files are detected to be tampered with you can make the device do whatever you programed it to do from continuing to run to "scorched earth policy" (wipe the device clean of certain files set as secure.
-
- Hard to argue with stampsm, but I'd like one full-size slot as well, unless we can get three or four micros and do that Raid thing. (wolfish grin) I think I prolly want a full 'cause they go to 4Gb, and with SDIO there are non-memory peripherals available for it. Not that I own any, mind you, because the Z uses CF for this, but the PP won't have CF. Seems like a nice thing to have, though.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=152869\"][{POST_SNAPBACK}][/a][/div]
yep full size slot and micro one would work. full size for user access and micro for operating system files (or something like this). and CF would be nice IF you can find the space to squeeze it in, or use the CF (actually IDE) port on the processor for peripherals.
-
i think you misunderstand pgp encryption. it only uses rsa or dsa for encryting a key that was used for a symectrical cypher on the data (such as AES) security is covered elsewhere, but its encrypted and tamper proof (no bit flipping allowed) but for the kernel and initrd only as all other data is changable so i expect that you rely on the fact that the initrd and kerenel are garenteed to be correct and use them to verify the rest of the device.
i want to provide a secure path of high assurence booting to the kernel, what the kernel does afterwards is its problem but i also want to leave the kernel to do the security stuff rather than the bootloader/bios the only way to do this in a secure manner is to ensure the integrity fo the kernel and then let the kernel ensure the integrity of the filesystem you boot from
the rokor idea is a good one. i dont know how devs would like it i guess put it at the edge and allow 2 diffrent sockets to be used and provide a cut out on the case
oh well glad the size thing is sorted, best of both world really. if there is a Cf card in there it would be IDE and not CF so non removable (unless the chip has hotplug support for ata drives) this is so that you can give it an ata drive if you wish from an ipod for eg
-
well since SD has only 9 pins it would not be to hard to have a one of those flexible brown cable thingies (not sure what they are called) that break out the internal micro-SD slot into a regular SD card holder. it would be less work for the devoloper to keep swapping out cards since you do not need to take out the battery just turn the device off. and you can have 4-5 different SD cards you just turn the device off and plug a different one in and boot into a whole new operating system configuration. plus you can have up to 4 gigs of space (plus possibly 4 more gigs in the other SD slot) to work with and possibly even have enough space to do native compilation of the programs (some programs like openoffice.org need to be natively compiled, though i have heard it can take a few days to compile on an ARM system) . basically you take the same case the consumer gets (or close to the same) and just make a breakout cable to plug in a full size SD card holder externally.
-
i think you misunderstand pgp encryption. it only uses rsa or dsa for encryting a key that was used for a symectrical cypher on the data (such as AES) security is covered elsewhere, but its encrypted and tamper proof (no bit flipping allowed) but for the kernel and initrd only as all other data is changable so i expect that you rely on the fact that the initrd and kerenel are garenteed to be correct and use them to verify the rest of the device.
i want to provide a secure path of high assurence booting to the kernel, what the kernel does afterwards is its problem but i also want to leave the kernel to do the security stuff rather than the bootloader/bios the only way to do this in a secure manner is to ensure the integrity fo the kernel and then let the kernel ensure the integrity of the filesystem you boot from
[div align=\"right\"][a href=\"index.php?act=findpost&pid=152971\"][{POST_SNAPBACK}][/a][/div]
i guess in my madness i did not explain myself very good as that was exactly what i was thinking. you do some sort of check to verify the kernel and key system files have not been tampered with then load them and hand over control over to them. then they can be configured to do as little or as much security checks of their own on the rest of the system as you boot it.
-
yep its the whole verify the kernel and initrd thing in a nut shell. only problem is it then means you have to have some change detection in the initrd. however if your / filesys is RO then you can just cheksum/hash the raw block device. might want to keep it small or do the zaurus thing and not reboot many times (people were amazed when i told then i hadent rebooted the thing in 6 months)
nice idea with the breakout part.
well compiling is not much of an issue. i have been thiniking about "blades" o these things or stripped down versions with only a serial port ram, and basic flash. does a net boot and then does distc over the network. i was thiking plain ethernet but i thoght it would be easier to build it all into a high performance backplane and connect that to the blades via a fpga thats mmaped as sram
means if you are doing unique things (like running a modified sql server that offloads the search to a distribtued search engine that take a single cycle regardless of entries less than its maximum) then you could quite easily chuck some hardware at it
gives me an ecscuse to egineer a fpga interface in the linux kernel that supports dynamic loading of "programs" or hardware designs while running
-
One vote for CF,
it's great to have the peripherals it allows, ie: ethernet/wifi/bluetooth/audio adapter etc.
-
actually i am pushing for ethier ethernet on board or move everything to usb2.0
wifi and bluetooth should be built in and i havent seen a cf audio. usb does it fine and then i dont have to worry about driver issues that we get on these boards all the time about new cf cards
the other thing is how much space it takes up
-
been skipping posts--just read the last two.
imho--USB 2 & cf would both be useful. don't sweat on board ethernet, a usb port is more adaptable. keep a cf slot. the tech is driven by cameras and is likely to continue improving without weird drm-ish crap happening to it.
my quite possibly irrelevant .02usd
-
- I think if CF is in it should be external, particularly if ethernet won't be on-board. I don't like using USB as my "network" connection. It's okay for peripherals, but for actual networking I vastly prefer ethernet over USB and some kludge to make it look like a NIC, that only works on the one computer it's running on. Ikky. I'd much rather plug into my network like a real computer, when I'm on the wire.
- I like the micro-SD idea for the OS, with miminum internal flash, and I really like stampsm's idea of using a "break-out" ribbon cable for development. I'd still like to keep at least one, and without CF more than one, full-sized SD slot.
- Biometrics are a nice-to-have, but not necessary, imho. I'd rather type in a password and have Raid-on-SD than have a bio scanner, or even a camera.
-
there is a CF audio card, discussed in another forum, but it's designed for high-end. A more generic twin-channel analogue I/O system which could be pressed into service as an oscilloscope would be nice - use at 44 or 48kHz, 16 bit * 2 and it's an audio capture/playback system, use at 1MHz, 8 bit, it's a simple 'scope.
-
well there are always usb to ethernet adaptors, and with a usb hub you could have 4
haha RAID on SD, sounds good still in 2 minds about the finger print reader, it dosent really bring any security to the device like the smartcard does. but at the moment its out as i cant find a supplier
only reason i like it is because even on my Z i used very long passwords (23 chars or more) and its a pain to type.
there will be a slot for a CF card but i am tempted to make it cf on ata rather than apure CF card so that if you wanted a larger model and you had room then you could attach an ipod drive to the thing. i need to work out the space allocation first
-
How thin can this hardware be made? The thickness of my Z has always been the thing that bugs me the most. A clamshell device that is long, narrow and thin, would have a big enough screen and keyboard for normal use, and still fit in a jacket or back pocket.
-
well it wont be "razar" thin, i hate that type of desgin, in fact i felt the Z was a good thickness
at the moment with the case design i have in mind it would be as thick as the keyboard part of the c3000 with the keyboard indented (grips on sides of keyboards)
anyway you can chose which case you want, but you might have to use a smaller battery
-
I want one of those (even if it's not in a case to start with as i like to fiddle with new gadgets (and maybe do some kernel hacking - since i don't have any of the tools (oszilloscope, logic analyzer,..) to do anything on the hardware myself really.
will there be a possibility of getting one (with or without case) early on to do testing and stuff?
-
yeah there is actually, it will be beta but basically i will have the plans on the internat as well as where to order the board from, you send them the design and they send you a populated board
for the dev model not everything has to be soldered on, there is also the posability of a cheap dev board that is nothing but RAM and the CPU (image is downloaded over a parrelel port jtag cabel or usb aand exectues a linux kernel with ramdisk from there for testing, great idea for a cluster of these (my compile farm in the futre))
-
- Your OpenPDA link says "FORBIDDEN" on April 5, Da_Blitz. I presume any data on getting a dev system together will appear there?
- I'd prefer an Ethernet port. My experience is that "networking" with USB does little more than complicate an incredibly simple and reliable system. "Progress", I guess...
-
i dont think a full sized ethernet port is a good size for this thing
i know it sasys forbidden,im still trying to get the site set up, dont want to use apache or any form of dynamic page generation where posible however i will have to make an exception for the wiki and webmail but i might move that to another host
i have the basis for a webpage generation system that uses make to build the entire site when somthing is updated, i have my resons for staying static.
as for the power supply, reson its not up is its really a monolithic chip rather than several seperate chips so there is no real point in putting it up
-
- Unfortunate, the poor Ethernet. It's no larger than USB A, but that's a lot of area compared to mini-A, I guess. (sniffle) I was so hoping to have real networking. USB networking is "real" about like Winbloze is a "real" OS.
-
i dont think a full sized ethernet port is a good size for this thing [div align=\"right\"][a href=\"index.php?act=findpost&pid=158311\"][{POST_SNAPBACK}][/a][/div]
I think that a regular cat5 ether 10/100(/1000 would be nice) is essential. I would rather than one less USB socket if needed to make room.
-
actually you might want to know there is an actual product called "winbloze", it was an excellent parody of windows and its users about 7 years age if my mem dosent need to be fsck'd
the one advantage of usb ethernet is it can be faster than ethernet, however YMMV
it has USB host so why not stick a gigabit USB2.0 ethernet adaptor in the port
-
giga ether full duplex is going to be somewhat faster than even usb2.0 full speed (480 mega bits/sec); it's a pity it can't be provided on the CPU.
-
true but my point is why stick with 100mbps when yu can potentially get up to 400mbps half duplex, how you split it is up to you (i mean how many operations these days are 50/50 apart from p2p)
i have to credit the NSLU guys with this one, there are a couple of people doing it to get bettre than 10MB/s transfer on the slugs, using a gigabit adaptor in a usb2.0 port gives you up to around 35MB/s if you are lucky (ie low protocol overhead)
-
Hi everyone,
As u seen, I am new in here. When I dug the old posts, I am quite confuse for the spec of pocketpengium and hence I summarize it and ask for validation:
LCD: 5" 16(15?):9 transflextive (touchscreen?) slide-like, not tweak model like Cxx00
keyboard: prison5-like
USB: 2 full size full power host-client usb port
SD: 2 SDIO
CF: no
ethernet: use usb-ethernet
audio: usb-audio
CPU: freescale
Ram: 128Mb
Storage: ??
power: ~3.xV
Battery: Nokia battery put into serial for swapping
shape: clame shell like, 2 parts link by connector.. Upper part only contain a CPU, Ram, LCD and a battery. THe lower part contains I/O, keyboard and batteries.
built-in wifi, bluetooth, camera
Just let you know My personal pocketpengium usage objective:
1) Notetaking (for science classes which need to draw a lot charts and diagrams)
2) ebook reader for a long while
3) programming and research platform
4) (VNC terminal to) my desktop ^^
5) simple PIM stuff
6) dictionary
7) Handy Calc
8) never mp3 player and phone, but may be GSM
So power is very important for me....
I also have some suggestions and questions too,. Since the topic are board, I don't seperate them into different thread which is hard to trace. BTW, I don't have electronic background and low level background, plz don't bind if I raise some stupid idea. :)>
1) If we have so much built-in thing, can we cut the power for them to safe the power? I heard about the Cxx00's wifi of CF is a power monster.
2) Can the device set to "bypass" the battery when charger is used so the batteries won't hurt?
3) Is there anyway to use PSP's 5V external batteries? Those batteries are really cheap now....
4) I suggest to have 1 or 2 SOny Clie's "jog dial" like "gear"? It sounds crazy, but it is very useful for small screen device, especially for tasks like note-taking, drawing and web browsing.
5) If someone still think keyboard is too small, maybe we can try the IBM thinkpad's butterfly like keyboard.
6) But if we don't use cxx00 size keyboard, are there anyway to use keyboard without a table? It would be cool to play w/ my pocketpengium in a long line or inside public transport and watch the others carry a heavy M$ laptop which ran out of power.
Sorry for my long post :)>
Finally, I am not pushing, but when is the pocketpengium device come out?? I want it deadly........ What about this summer??
-
thans for your usage case, you have no idea how handy that is to have, i mostly get one or two "i do this or that" or i dont like that because i wont use it, having a usage case allows me too chose features and have justification behind this
now to answer your questions
(1) is a kernel thing, wifi has been and always be a power hog unless you tune the power settings (at the cost of responisvness and latencey, which is fine for teh web but not as good for ssh)
(2) yes, however keep in mind this can be dangerous if you pull teh plug and you get a 20ms power out, hence there is a mini backup reserve. not that you need to know or care about it but i think its worth putting here
(3) Perhaps, depends on avalability and sizing, this thing dosent even have a set case
(4) I like jog dials but i prefer a proper wheel, think htc hermes or trinity (i have both) its like a jog dial but keeps turning, great for mp3s and afaik i planned to have 2 already
(5) links, i have propsed several keyboard designs but anything other than the cxx00 tends to be rowned upon, my fav was a fold in half one that was a keying/corded keyboard when folded up. however most people dont want to learn a new keyboard layout.
i personally felt that for a Z a cordded keyboard would be nearly as fast as it allows you t better maintain a grip on the thing and means less odd keyboard combos (ctrl+c, which i use all the time)
i do have to keep telling myself that others are not willing to give some wierd things a go, i try everthing once (typing this on dvorak where i went cold turkey) and not everyone has a heavily modified keyboard layout (all my Fxx buttons are bounde to common funcitons)
(6) well my summer is in december, the current holdup is no one badgering me about it and no one technical to make a point or two and lend a bit of a hand (eg, i know you havent looked at ityet but have you looked at chip yxv for function x)
i have an idea of nearly every chip i need just not which exact chip yet, the only ones really remainng are dedicated chips for the rf radio and misc fetaures.
so the best way to join the project and help out is to annoy me as much as possible on msn or by email (perhaps i sholud set up a mailing list), look at chips that sound like they should do the job (note no electrical experince required, i just need sugestions of chips that sound like they might work) or find me some egineers woh want a PDA they built themselves
ooh and i am looking for software guys or anyone who have some crazy UI ideas or dreams, i have a few bizare ideas for things running in the background to improve the uesr experince (syncing and such)
and if you have read some of my posts and seen the time stamps then you would know your post is mearly "introductionary" in length and my spelling sucks
-
actually you might want to know there is an actual product called "winbloze", it was an excellent parody of windows and its users about 7 years age if my mem dosent need to be fsck'd ;)
- Funny. I wonder how long it took the "real" Winbloze to sue them? (wry grin)
the one advantage of usb ethernet is it can be faster than ethernet, however YMMV
it has USB host so why not stick a gigabit USB2.0 ethernet adaptor in the port ;)
[div align=\"right\"][a href=\"index.php?act=findpost&pid=158611\"][{POST_SNAPBACK}][/a][/div]
- As speculatrix mentions, USB isn't faster than gigabit, it's just faster than 100Mb. Whoopie. So is Firewire, whose latest incarnation is also faster than the latest USB, just like the first incarnation was. So why not put Firewire in because it's faster? But faster isn't the point.
- The point is that Ethernet is the world-wide networking standard. USB is unfortunately now the standard for peripherals, but it is not routable. Just because someone's made a do-dad to allow USB to connect to Ethernet and pretend to be routable doesn't mean that solution is desirable. It just means it works. If USB were really faster or better for networking in any manner what-so-ever, the world would have migrated to it from Ethernet. But the world hasn't. Neither should the Pocket Penguin.
- On a personal note, I don't want to have to carry any more dongles around with me than I have to, or have them hanging off the back of the Z like some goiter. I would prefer that my pocket Linux workstation be able to natively connect to a real network like a real workstation. Wifi is not secure enough to be trusted as the only Ethernet-like network connection. A wire is essential, and USB faking Ethernet is not.
- You mileage may vary, but that's my half-nybble, such as it is.
-
all good points however the one that kill it for me is the crazy power usage of gigabit chips and the reqired isolation transformers (bulky)
-
Much as I love reading about this potential device and I would seriously think about buying one as soon as there is a stable operating system.
I want to know, are we any closer to building one than when this thread started in 7/6 (if you're american 6/7)?
Chas
-
i wish i wolud staop hearing about it and had a finished product, the only reason there isnt one is that i have been very easy to distract for about a year now (lots of stuff going on) and it becomes hard to keep motivated about somthnig for this long esp. when there is no input
i will seee what i can do about wiring up the power planes tommorrow, that should be a major win
-
- I'm not advocating using gigabit. I'm advocating having an Ethernet port on the Pocket Penguin at whatever speed the battery can handle. I'd rather have Ethernet at 1Mbit than have to rely solely on USB "networking" at any speed. I'd be happy to have a PP either way, but I'd prefer if it had Ethernet.
- Everyone has other obligations. Far as I'm concerned the fervor has died down and now we're into the long slog. No biggie for me. I've been in the new house eight months and haven't set up the cross-dev workstation yet. I bet if someone told you, Da_Blitz, they had an OS ready and needed a development system to start working on it, you'd move along faster. (shrug) I don't see that OS manifesting itself, either. The Pocket Penguin will get done, or it won't, and either way I expect we'll enjoy ourselves, or we wouldn't even consider building it.
-
I agree with Ragnarok. If the hardware exists and is well documented, the programmers will come. The forums are full of clever people who would buy up the heir to the zaurus throne, it just needs to be a worthy contender!
I think even a part working prototype would really get people's appetite going, but so far, sorry to say, we've had a lot of buzz, many ideas for features, but not even a parts manifest or even a prototype PCB on its way.
p.s. I think if you got a part working board with the promise of a free final sample, you'd get one or more of the Angstrom people working on it - they couldn't resist the challenge.
-
p.s. I think if you got a part working board with the promise of a free final sample, you'd get one or more of the Angstrom people working on it - they couldn't resist the challenge
you saw right through me, actually i was hoping to get a cut down board out the door first that could connect to a backplane. with just ram on board with the CPU and use the usb bootloader to load the image
would make a nice small cluster for some services i want to run (virtualisation only goes so far)
oh well just need to knuckle down and do it, just keep on bugging me here and on IM
-
p.s. I think if you got a part working board with the promise of a free final sample, you'd get one or more of the Angstrom people working on it - they couldn't resist the challenge
you saw right through me, actually i was hoping to get a cut down board out the door first that could connect to a backplane. with just ram on board with the CPU and use the usb bootloader to load the image
would make a nice small cluster for some services i want to run (virtualisation only goes so far)
oh well just need to knuckle down and do it, just keep on bugging me here and on IM
[div align=\"right\"][a href=\"index.php?act=findpost&pid=161831\"][{POST_SNAPBACK}][/a][/div]
absolutely minimal necessary circuit board to get started:
* CPU (it's quite useful)
* RAM
* a boot-strap NOR flash to boot from USB,MMC/SC
* USB slave
almost a necessity:
* on MMC/SD socket
* USB host
I'd guess the keyboard would be a usb device? I've been looking around for a tiny usb keyboard and there's nothing to match the zauruses - they're all either mini laptop sized or thumboards, nothing in the middle.
-
keyboard was gioing to use the dedicated keyboard handeling logic with keyscanning and such.
as i said you dont need flash to boot the thing, the device can download a kernerl and initrd into ram over usb client (built in bootloader over USB, no external device required)
nice thing about the freescale chip is that it can boot off of NAND flash, so we only need one chip for teh entire thing and we can use the cheap stuff
-
connect to a backplane
- What backplane would that be? PCI slot? S100? (grin) Or something proprietary? Just curious how this will all shake out ... if it's booting over USB it doesn't seem like it would need much a "backplane", per se.
keyboard was gioing to use the dedicated keyboard handeling logic with keyscanning and such.
- Is this a normal keyboard, or a "custom" scanned keyboard like the Z has? Or something I'm not thinking of? Again, just curious how this "bare-bones" system will work. For instance, if I had one, how would I interface to my wintel (even if that wintel happens to be running linux) boxen to start developing for it?
-
keyboard was gioing to use the dedicated keyboard handeling logic with keyscanning and such.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=161923\"][{POST_SNAPBACK}][/a][/div]
wouldn't it be better to keep as many general purpose I/Os free as possible, can be used for bit-bashing serial and parallel ports, I2C, A2D converters...
-
using the keyboard scaning logic makes it eaiser to go to low power mode, sure the usb spec can handel it but turning off the USB logic in suspend means greater power saving
as i see it there will be a power button for turnig on the device initially (or perhapes power on when power is applied) and let any keypless wake from suspend
however a dedicated usb device does allow for some wierd posabilities, for example using the usb "bridge" feature to connect the usb keyboard directtly to a host computer
if i remeber correctly there should be 24 gpio's free if i use the keyboard handeling stuff
as for the "backplane" idea, usb is fine for loading the firmware and you could use it for comms but i had a better idea in mind
make a pcb with some flash, ram, a fpga, an ethernet port and multiple controlable powerdomains on it with several PCI connectors, then make stripped down cards with the cpu/ram/fpga combo on it. basically you can swtich power to any slot on and off, allows hotpluing of cpu cards (usb handels detection) and use a fpga to implement a layer 1 and perhaps 2 OSI comms model. eg 32bit DR @ 133Mhz (233Mhz effective) the fpgas have this logic built in to talk to ddr ram but it would make for a nice way for these things to talk
its a cluster in a box of sorts. but its a better way of giving these things gigabyte comms in some ways (or @ 64bit, 400Mhz you get 3.2GB) you mem map it as sram and you get low latency comms between cards with enough speed to make rebote hardware (in this case hardware not attached to the chip) appear as local, which is my goal as well as some cluster and load balence stuff. its cheap
as for the actual bus, could be wishbone, wont be pci but i would rather make it seems like an ethernet card with RDMA capabilities and some ultra heavy offloading capabilites (eg scatter grab and assembley of packtets done by the card instead of the kernel)
but its a fpga so chuck whatever you want in there, and wihle your at it you could add a cryto offload egine in the spare space
lets just say i have refound my intrest in the project as well as finding a way that this design may be useful to more people
-
- "Cluster in a box" sounds nice, but that seems like it'll be a bit of overkill for a proto system to slap an OS on for testing. (crinkly grin) If they are that low latency we should make a "hypercube" out of them and market the whole thing as an affordable Grid computer. "Yes sir, I know the CPUs are slow, but there are hundreds in this one suitcase-sized box. Your brain switches at 1Khz and look what it does. Throughput on this will knock your socks off."
- But I digress.
- Let's say hypothetically I decide to try to port an OS to this beast. What would I expect to receive as a prototype and how would I interface to it? Pretend I have a pretty decent familiarity with programming in C/C++ and know hardware well enough, but haven't designed squat with electronics for 20 years so my understanding is almost hopelessly out of date. Would this be one board and an USB connector to load the firmware? Something else? If it's just a CPU board how does one interface to it for "keyboard" and "display" to see it do something?
- Or am I just going off on a tangent? (shrug)
-
you would get a pcb with a cpu and 64MB of ram, ram size may change but for the dev version i doth think there is needs for alot
it would have a pinout for a usb connector but wuold be fitted with a edged card like the PCi socket, so you can solder to that or plug it into a back board
there would most likly be a fpga on board for a bit of fun and IO stuff. and thier would be a serial line as well
the hope is that it would all plug into a backboard and they would all have an ethernet equivelent connection between each other and nfs thier root fs off node 0, all force fed over the devices usb client port, if you wanted to run it without a back board then you lose the network connection and do everything over usb, includhing loading firmware
so load firmware then move to usb-net in the downloaded (into ram) rom
-
- Cool. What about keyboard emulation and such? If the system should boot and load drivers how will the dev board do that? Or will it simply not do that initially to get the basic OS whipped up and bootable? If it doesn't worry about trivialities like keyboard and display initially, what about later? All these devices are supposed to be part of the final unit, but none will be on the prototype board. Will those be added later, somehow, then the drivers bolted on then?
- I presume this CPU board would be a precusor to the bona-fide Pocket Penquin. Perhaps that's off the mark and you're concentrating more on "blades" than on the PPC itself?
EDIT:
- Just read up on FPGAs. Wowzers!! Are we using a small one glue, or is this thing going to be built from heuristic hardware?? (wolfish grin)
-
I am concentrating on the blades more, they are the base for the other designs i have but i was betting on a device comming out that would match my needs or wants closly, and it was anounced today (http://www.engadget.com/2007/06/14/compal-showcases-prototype-umpc-running-windows-ce/ http://translate.google.com/translate?u=ht...language_tools) (http://translate.google.com/translate?u=http%3A%2F%2Ftechon.nikkeibp.co.jp%2Farticle%2FNEWS%2F20070607%2F133773%2F&langpair=ja%7Cen&hl=en&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools))
time will tell if it makes it to market but i have a feeling it will, but i also feel they will shave the ram down a bit and make the case nicer (in black please: http://64.233.179.104/translate_c?hl=en&ie...language_tools) (http://64.233.179.104/translate_c?hl=en&ie=UTF-8&oe=UTF-8&langpair=ja%7Cen&u=http://techon.nikkeibp.co.jp/article/NEWS/20070607/133773/%3FSS%3Dimgview%26FD%3D291939425%26ad_q&prev=/language_tools))
and it supports USB host ( http://64.233.179.104/translate_c?hl=en&ie.../language_tools (http://64.233.179.104/translate_c?hl=en&ie=UTF-8&oe=UTF-8&langpair=ja%7Cen&u=http://techon.nikkeibp.co.jp/article/NEWS/20070607/133773/%3FSS%3Dimgview%26FD%3D174474337%26ad_q&prev=/language_tools) , look at the specs)
i was thinknig about a spartan 3, you can get some heafty devices in that range in many diffrent price rages that are pin for pin compatible
for anyone else reading this, look up what a fpga does, i can give you a brief but its mind blowing stuff.
essentially its the equivelent of what software is to a cpu only for hardware. its dynamically reconfigurable hradware that can one seccond be a tv tunner, the next a distributed key cracker and the next a PC. and they are MASSIVLY parrelel, one of these doing an optimised fast furior transform is over 50x faster than the best x86 computer you can put together
x86 can only exectue one stage at a time of the fft, this thing can do all stages in parrelel (so around 2048 operations in parrelel per clock)
-
device comming out that would match my needs or wants closly
- Interesting do-dad, but I missing what the CPU speed is, though 256MB RAM looks nice. It also only has a single micro-SD slot and no hard drive; the former (currently) doesn't support as much storage as full SD nor does it have peripheral capability, but that may not be an issue with WiFi & BT built-in. The latter is more of a restriction imho because any "full" Linux will run better with swap than without it. Even pdaXrom runs faster and is happier with the hard drive in a 3200 compared to no drive/swap on a 1000, based on how T3_slider says it runs on his Z and how mind-numbingly slow it is on mine.
i was thinknig about a spartan 3
- ???
-
spartan 3 is a xilinx fpga, has small caches of ram and some embedded logic in it. on the pda design i would use a cpld as i wouldnt need the higher logic fuctions, only things like adress decoding (CPLD's are lower power but much simpler devices, you couldnt cram a uP in a cpld)
still undecided, i have a couple of things happening all at once that are related to this but i should have had a design by now. i think the one thing that has held me back is trying to satisfy everyone. i should just skim the features to what i want, add in anything else i can and run with it.
one of my "wants" however is a profesinal looking case
-
- I hear ya. I finally lost Hiroshi and ordered an UMID BZ to replace it, so I'm Z-less until further notice. There it is...