Quotes Thread For The Blog


(PND continued from post above)
skeezix said:
It's an ISO that can also be squashfs'ed.The filesystem in the pnd can be many things; iso is one, cramfs is others; its not all pnds are one type,.. its just a given pnd may be one type, and another pnd may be another type, up to developer preferences.

In theory, we can support dozens of filesystem types, but in practice we're only supporting a few right now. (its purely a 'magic' file thing, in pnd_run.sh, really.)

Even windows can make cramfs (we do it :)

jeff

As to post-release.. I'm not a pnd-nazi; libpnd we've made and LGPL'd, peopel can go nuts :) I'd liek sensible reviewed alteratins and additions .. no need to go hog wild and break it all day-2, but whatever. Extensions, changes, even long term changes.. all good.
-----------------------------------------
also..
You can just stick pnd files in any old directory you want; you don't have to use the menu or desktop auto-icon business at all .. you can stick them into /games on your SD, and then just use the file browser and double-click on the pnd you want in there to run it.

No one is forcing you into anything :)

What we are doing is making automagic instant easy to use behaviour an option ... seems desirable to me, but you don't have to use it :)
-----------------------------------------
Also, we're ignoring one of the big benefits.. pnd is drag and drop and play. PEople are really asking about.. wanting it more like windows? _Windows_? With dlls in one place, apps in another, your user data in a third place, your registry entries, and home-doir hidden stuff? And oyu have to run an installer and an uninstaller?

We have that beat :) Stick your pnd files in one place (or any place you want really), and your data all goes to one folder, with a subfolder per app. ITs easy to back up, easy to delete, and all automatic.

The only case I can see here, and its an interewsting idea, is to make the appdata-dir be the same dir as the pnd-file, as an option; that'd not be too hard a change, but lets leave that as a post-release feature request. Right now, all pnd user data will go to /pandora/appdata/app-name; in the future, if you want it so /games/foo.pnd writes its appdata to /games/foo/, we or someone can code it in; its a change to pnd_run.sh really, and maybe a conf file to toggle it.

As it is, its a very good system; whats going on here is 'bike shed' -- arguign over little details :)
ED(tweet) said:
I uploaded a YouTube video -- Pandora: Surfing the web, doing some messenging and eMailing http://youtu.be/1QUDH-Ii8Cc?a
(Pandora SDK)
skeezix said:
I sent the 'WORK IN MOTION' set of libs to atiti there, so now you have .. CSL + usual headers + some arm libs to link against + libpnd -- so you can complete your VM :) (or just use cygwin, or build your own gcc, or whatever :)

jeff
atiti said:
Hi guys,

I've torrented up the stuff skeezix has put up and its now available at:

http://www.patup.com/torrent/1841/Pandora%20Feb23%20ARM%20libs%20-%20Skeezix.html

Or a direct link to the torrent:

http://www.patup.com/dl/1841/pandora-feb24-usrlib.tgz.torrent

Happy leeching! :D

EDIT:
!!OH yes, please read this!!:
"This is purely a set of random ARM-built libs for pandora image;

they _will_ change (more libs coming, possibly some dropping, and likely some version changes.)
I've also done some symlinking to make them easier to link against with gcc and CSL-gcc.
I compile and link against these quite often, and drag tgh resulting binary over to real pandora devboards and works great :)
Combine these + Codesourcery + linux headers == win! :) "

P.S.: Thanks skeezix
skeezix said:
That's a snapshotand will change as I said

it is git/cvs pulled, with manual versions bumps..

Future os versions will only add libsand versions, not remove, and as space permitswe'll follow versions

jeffphone
(Cpasjuste)Pandora Sdk Installer An unofficial pandora toolchain/sdk installer for debian/ubuntu
Cpasjuste said:
Hi, i did re-write my little pandora toolchain/sdk scripts for an easier and better installation.
This stuff is writed/tested for deb based distributions ( debian, ubuntu ... ), 32 and 64 bits.

It comes in 3 files :

toolchain.sh : will download and install codesourcery arm toolchain ( arm-2007q3-51-arm-none-linux-gnueabi-i686-pc-linux-gnu.tar.bz2 ) and set the PNDSDK path variable ( /usr/local/pandora/arm-2007q3 ). It can be used alone to install the basic toolchain.

pnd_libs.sh : will download and install some libraries from the angstrom repository, based on the third file list. It need the "libraries.list" file.

libraries.list : a list of libraries URL from angstrom repo. So if pnd_libs.sh script isn't too crappy, we (you) should only need to update this list to update.


I encourage you to report me libraries that may miss you ( i just added a very few one for gles, x11 and sdl ), if they are part of the actual pandora base image, i'll add them to the "libraries.list" file.
The "toolchain.sh" script and "pnd_libs.sh" scripts will leave a "tmp" directory where it was launched with downloaded stuff, you can safely delete it.


toolchain.sh
pnd_libs.sh
libraries.list
(Midori CSS bug)
EvilDragon said:
Na-Noo said:
Did ED's vid say that there's a css bug that's been fixed in the windows version but hasn't made it to the linux version yet?
Nope, it's fixed on normal Linux desktop, just didn't make it into the Angström repo yet :)
(automatic clock setting)
skeezix said:
Elanzer said:
Ideally you would want to run a shell script upon execution of each program which configures all the clocks to their optimal speeds for each emulator.

Like if SNES emulation can run fine at 300mhz OPP2 or something like that with the SGX clocked at 55mhz, you'd save a good amount of battery life, then upon closing the emulator, trigger the usual speeds needed for basic desktop/browsing type use. Likely you could keep the SGX at 55mhz and only ramp it up for things that even need 3D.

There are so much possibilities to further extend battery life beyond the 10 hours, it sounds like much more will be possible after software gets optimized.

That is in fact what happens more or less .. if a pnd-file requests 200mhz (say), and that is allowed by the user, then the system downclocks, runs the pnd, and restores clock on exit. Sort of goofy if you're running 5 apps at once, but the user knows the risk of doing that :) (ie: slowing down other apps, and the whole risk of running 5 apps at once, etc.) Anyway, yeah, we do that :)

When the LCD is closed, we turn off the screen (but leave clock alone in case you're burnjing thrugh mp3s, say.) But if your mp3 player mneeds only 100mhz, set up a pnd for it that requests that, and bam, every time you run your mp3 player, you're running at 100mhz; close the lid, and you're talking pretty low consumption :)

Theres lots of options..

jeff
 
Last edited by a moderator:
(Regarding PND and overides available)
skeezix said:
We've stated many times.. the system has a built in option for overrides; you can override the developer set category, right at the pnd-level; presumably the menus will have that option coded for them, or someone will code up a override editor tool. But the pnd-system has overrides built in .. so if the dev says 'default to category Games', but you want it in "Applications", you can do that. The goal is.. we might just save you the hastle, it has a good chance of just being right. You want to move it, shoudl be a piece of cake. (maybe not day one, but the system is there, so day 3 maybe :)

Likewise for icons; presumably menus will let you override the icon in some way or another, or do it at the pnd level. OR worst case, use a pnd-editor and just slap a new icon on it (ED is working on just such a tool.)

We have gone out of our way to make 'automatic, but not forcing you'; seems to me you're making an issue out of somethign that its not -- ie: you had to do all these things _before_ -- developer included an icon foo.png; how many peopel _Really_ edittedf that icon and replaced it? So you can do the same thing now, its just that instead of having a separate file, its built in.. that is _better_. The menu could still allow an file with the same name tobe shown as a replacement icon.. who said the menu authors couldn't? So whatg we have is better.

Nothing is lost over what you had before.. we've just added sensible defaults in a single package :)

So your 'easy solution'.. yes, thats there, menu authors can do it; and at the pnd-level (PXML level), thats built in too, but rather than have peopel futzing with PXML.xml, we want someont to write a tool for that (be it in the menu, or a standalone tool), which will come.

See, I thought of that ;)

This has come up about 5 times; people will keep bringig it up, but it pains me every time.. we beat it to death, and same output :)

jeff
 
Last edited by a moderator:
(PND FAQ)
dflemstr said:
First shot at a FAQ: http://pandorawiki.org/PND:FAQ
ED answering PND questions: http://www.gp32x.de/board/index.php?/topic/52318-pandora-directory-structure-question/page__view__findpost__p__824448
(question about CeBIT)
EvilDragon said:
I'm there, yes, having a presentation about OpenSource gaming handhelds at the Linux Hotel stage from 12pm to 1pm :)

After that it's press interview time (also video interviews).
Will have a busy time there, yes :)
(SD slots)
skeezix said:
It's springliaded like mad and rocks

it's tight enough to hold an ejected sd wihout losing it (awesome) and has a really secure hold and solid spring

it's actually one of my favorite design choices they made :)

Jeff
(Battery monitor analysis) http://www.gp32x.de/board/index.php?/topic/52375-pic-battery-monitor-analysis/
MWeston said:
There is a register in the fuel gauge where you can put a number in to represent what you think the system will use at its peak. The idea is that software can check the time remaining under maximum load to see if it should cut the system off early because if you were to suddenly need that extra current, the battery would be too low to do it and data corruption could occur. Since Pandora can be running at less than 300mA or need 1.5A if you were driving everything all out, this register isn't all that useful. It's more for devices with no chance of jumping around much. The battery warning would trip way too early if it was expecting 5x the current at any moment.

Since there is no real way to know what the ideal value would be other than to put some crazy high number in there (ie: the USB host was being maxed out), I just put some value in that seemed like a good number under hard gaming only. It was probably 600mA based on the readings but I don't remember right now. It doesn't do anything if you don't write software to need it (which we don't) and it can be reprogrammed at any time since we don't seal the gauge for read only.

The most accurate reading based on all my testing is the "TimeToEmptyAtConstantPower" and 12 hours is pretty cool although it dips down to 10ish when playing Hexen 2 (my new favorite) or something using 3D.
notaz said:
Currently it's reading TimeToEmpty which seems to be more optimistic (as seen in the photos), can change it to TimeToEmptyAtConstantPower if you think it's better (personally I don't have a clue).
(ED commented on assembly at CeBIT)
JayFoxRox said:
On CeBIT ED confirmed the 22nd march another time, he also named 15th and 18th of march. I think one was about the boards and one about the plastic parts. ED also said that there is a "padding" between the 22nd and when they start working. They are worried that an additional delay (which is unlikely but could happen) would delay assembly, so they would sit around for 5 days or so with no work. Thats why they will meet a bit later than the 22nd to assemble the Pandoras.
He also said that not everyone can or should help with assembly once more. ED said those who repaired notebook for years or those working in the technical sector and who really take their time for it are invited tho.
craigix said:
Well we don't like to say too much anymore because a) our information always relies on other people/companies and B) when ever we are let down our predictions are quoted and thrown back in our faces and we get the blame.

However, I'd expect a significant number (about 500-1000 units) should have been shipped by this time next month.

It is now just a situation of getting all the bits (cases, boards) here in the UK, then the team + assistants here, and building them.

What I am personally doing at the moment is preparing the hall for the work, as always this job is more work than you may imagine.

Then I have to prepare to have maybe 8 people staying in my house too.

Then I'll have to make sure the cellar is well stocked, and the larder, oh and any paypal donations to that drink+food fund will be gladly received! Maybe we could sign your case on the battery compartment with a knife to say thanks (no idea if that would look good).

Oh and I have to set up that webcam (with sound?) too. Maybe we can have a big brother style close up with audio now and again to report what is happening.

That's about all the news right now.
craigix said:
jumpman said:
Hey Craig, since your in a question answering mood I've got one for ya! Are all 4000 unit parts expected on the 22nd? Or just a portion? Just curious, I'm assuming all 4000, but I wasn't sure!

Thanks,
Chris
For everything but the boards, yes. The boards is still unknown. We will have some but maybe not all, but that isn't a big deal as we don't need them all at first, they can arrive as we make the consoles.

If we have 1000 in March that is easily enough to keep us going for the first week.
(about the boards)
craigix said:
The testing process is slow, so it looks like they will arrive in batches. A tricky part is making sure the hand soldered nubs are centered.

They are getting there though, the Pandora has a lot more hand assembled parts than most systems, it's very Rolls Royce.
(CeBIT pics)
http://forum.gp2x.de/viewtopic.php?p=115618#p115618
http://forum.gp2x.de/viewtopic.php?p=115647#p115647
http://yvbe.kicks-ass.net/~daniel/PandaCeBIT/
mirror1: http://s69.photobucket.com/albums/i75/rabidpoobear/EvilDragon/
mirror2: http://www.feelthesource.com/titcherdon/ED/
http://jayfoxrox.bplaced.net/content/scratchpad/media/pandora/
http://forum.gp2x.de/viewtopic.php?p=115732#p115732
http://forum.gp2x.de/viewtopic.php?p=115780#p115780
http://forum.gp2x.de/viewtopic.php?p=115836#p115836
http://blog.radiotux.de/2010/03/07/cebit-2010-michael-mrozek-opensource-auf-spielehandhelds/
(Videos)
http://www.gp32x.de/board/index.php?/topic/52403-cebit-video/page__view__findpost__p__825524
http://www.gp32x.de/board/index.php?/topic/52403-cebit-video/page__view__findpost__p__825664
http://www.gp32x.de/board/index.php?/topic/50806-media-reports/page__view__findpost__p__825460
 
Last edited by a moderator:
[Quote limit is ANNOYING]

Screeny said:
Some more pics and a little report about ED & the CeBIT
(Please use the "translate" button)

http://www.pandoratoday.de/exklusiv/der-cebit-bericht/
(PND, unique ID)
skeezix said:
Againm all stuff we've been through many times :)

I'm a man of 'reasonably unique', without overkilling it. We don't want pure random numbers or to make peopel use a CGI to generate a ID from a central repository.. too much baggae for dev, and too much annoyance for users. I've been a fan of 'appname.devname.something' for awhile, but it might be too ugly. still.. battlejewels.skeezix.0001 is a pretty unique name, imho, but if someone collides, they have to be pretty going out of their way to do it :)

(and in my case "0001" means "state format" not "version of game" sinceI don't much believe in breaking state between versions.)

Anyway, currently the unique-id is used for unique-id and for appdata name, but I will be splitting it when I get a moment, so the dev can set the unique-id and also set the appdata-name (if other devs agree), so that the dev coudl be as free to use a crypitc name, or a common name, as he wants, rather than being forced into it. (I liek forcing devs into a pattern if its a good one, but clearly therte is some disagreement, so .. maybe I'll play it nice ;)

AS to:
"You mean if someone creates "AwesomeGame" and then someone ELSE creates "AwesomeGame", can it handle them both?
Or someone creates "AwesomeGame" as "AwesomeGame.pnd" and then updates it with "AwesomeGameV2.pnd" will they both use the same appdata directory?
"

Yes, of course we thought about it; we're professionals here. Spare time sure, but we're not idiots :)

Unique-ID collisions will vbe annoying, but shoudl not be common (the common scenario woudl be one dev copies is OldAwesomeGame to make NEwAwesomeGame and fortgets to update the unique-id; so worst case is one dev breaks his own shit, really, until he fixes it. No big deal, right? Same risk has always existed for every app for every OS ever, as well.) For Awesome1 and Awesome2, the unique-id and appdata-name can be updated at developer leisure -- if he wants to use a new appdata, or not, is purely up to him (as it always has been for every app and OS.) -- if if the state data is needing to be new, he can bump the id and get a new dir. Consider how many times for some apps you had to wipe out the old data before copying in the binary, whereas others you don't have to. That shoudl all be magicly taken care of now.

So yes, we've thought of most scenarios,m and tried to make it 'fail gracefully' :)

jeff
 
Last edited by a moderator:
(More PND)
skeezix said:
It is an icon appended, not screenshots. Fat media such as screens, audio preview and video should be in the pnd. Only pxml and icon are appended since those are needed immediately to help he user know what's inside

(a menu will mount the pnd to pull out the fat media as mounting is pretty quick on flash media)

jeffphone

skeezix said:
Converting to pnd is trivial to .. very easy on *nix right now; we've got some sh-scripts up on the site to just take a directory and turn it into a pnd in one step; add that to your makefile and *poof*. OR do it by hand with a couple easy steps, all documented.

For Windows, you can do the same sort of thign with batch files if you like (its document with downloads to tools in the libpnd hub page.) A few people (our own Evildragon even!) are working on GUI tools for editting PXML and making pnd files .. so shoudl be a piece of cake.

jeff
skeezix said:
Laurencevde said:
If I understand correctly, the PXML.xml is appended to the filesystem-image that constitutes the pnd? hmmmm, ok....
Might it be an option to also be able to store that metadata into extended attributes? I guess that'll be quicker.

So, that'll prefarably become:
Check to see if the FS can handle extended-attributes
if so, check to see if the PND has its metadata stored in extd-atts
if so, use those
if no ext-att support, or no metadata in it, check for a PXML.xml appended to the image
if so, use that, and put the info into extd-atts if possible
if not, mount the pnd, optionally append the PXML.xml stored inside it to the pnd, and add the info into extd-atts if possible


extended atts are often lost when copying/sending files around, so being able to auto-regenerate it would be very welcome. And auto-appending PXML.xml means less work for the developer, but that breaks the idea of non-changing pnds...

We thought about it a year back when brainstormning, and it really came to..

1) Consistency and simplicity is ebtter than having things vary by container, and most filesystems can't do that very well
2) You'd have to mount to get that info then; consider 1000 apps (or 500, or 10,000 :) and mounting every one just to get the info is a lot of work on the device (in real time while user waits, potentially.) Just pulling off the last couple K of a file (one read on flash) and grepping it for the values is consistent, easy, and fast.

(It may not be perfect, but one has to consider we at the time only had two months (!!) to esign and build everything, and then we've been extending and patching and fleshing out ever since as we never really knew we'd have an extra year to work on it :)

jeff

skeezix said:
Pnd is a dice-roll; we built a lot of cool stuff; we didn't "over do" it, since when you super over tightne the shoes, everyone complains about that too; to become 'common usage' somethign has to be easy, useful, easy for developers to pick up, and well communicated and so on .. pnd can work really well, but devs might just not use it at all; a plain executable works well too, and would be just lke every other device out there. Nothing stops people from just saying 'heres a zip, unpack it, works fine' .. it just turns into that mess wherre ebery app works differently, files spread out all over the place, etc.

So we hope pnd takes off, it coudl become a really cool thing over time.... or total fizzle.

Either way, thats cool; we throw some cool stuff at the wall, see what sticks; we've tried :)

Its not all or nothing either; theres lots of neat stuff everyone has built. From the software side, we've got several menus, we've got the every-device libpnd stuff (common ways to set brightness, simple conf files, adjust clock speed, all that sort of tstuff), makes developers lives easier. (And I hope devs will keep fleshing out and sharing rather than making 500 custom ways of doing the same things and fighting with each other..) -- we've included a few basic OS icons in the device for everyone to use, rather than make their own; we've put an effort into making everything open source and out there.. its Open Pandora. All these little bits.. some of it will stick, some of it won't. Thats cool, that is the way of things :)

jeff
 
Last edited by a moderator:
Prices-

MWeston said:
Actually the cost of populating the bare boards and testing them is now more expensive than talks I had with them dating back to spring 2008 when they were verbally agreeing to pricing we were budgeting for the entire build. Basically, doing it ourselves is only preventing us from sinking into a deeper hole on these first couple of batches.

The first batch is going to be the most expensive and we should be able to get the price down a bit for future batches. There were a lot of setup costs and learning processes, plus some manual efforts that we can't automate until the next batch. Still, it will never be as cheap as discussions from two years ago and we don't have the manpower to move assembly to China without wading through a possible disaster as the supervision required would probably out weigh the cost of still doing our small batches in the US (boards) and England (cases).


CeBIT Videos:


.
.
.
.
.
.

Jeff's Minimal Menu:
skeezix said:
I've been goofing around the idea of making a uber-minimal launcher; Craig really wanted me to do it, so we coudl have something basic and fast and sort of like gmenu2x.

Seeing as we're shipping soon, this is quick and last minute .. wish I'd known to write something months ago ;)

Since I'm in a hurry I made it depend on SDL, which buys us the advantage of working on framebuffer or X11 no problem, so can use it inside or outside of X.

A quick pic from my desktop:
http://www.codejedi.com/pandora/minimenu/screenshots/minimenu_0005.png

A quick pic on the pandora:
http://www.codejedi.com/pandora/minimenu/screenshots/IMG_0174.JPG

A quick video, which doesn't even show me switching tabs or scrolling up and down, though it can do those things. You know by now, I suck at video.
http://www.youtube.com/watch?v=BlIx3ZukTDo

I've got some loving to give it yet, as best I can squeeze the time in.

I've made it open source, GPL, though the code will bring shame on the House of Skeezix for being rushed out the door, but you go and write a menu in 4 nights when you're on babysitting duty ;)

jeff

Anyone like the basic looks of it? Theres no features .. no movie playing, no fancy menus, no battery indicator (so far, not sure I see the need, we have a blinky LED right?), and so on. It just comes up, lets you shut off the pandora, or run apps. Maybe I'll add a directory browser for running apps from the filesystem.
skeezix said:
Its currently all png-files, named in a conf file (and with their position and various options set in conf.) ie: You want a background image? Specify a filename with the right key in the conf, and bam. You want a grid thats 1 column, or 8 columns, just tell it the spacing and padding and scaling sizes for icons, and it'll take care of that. Borders, arrows, etc, all just png-files.

So folks ca customize it 'reasonably', but not as much as cpasjuste's pmenu or fancier menus.

But my goal was to have it start up in .5s (which it does), and then scan for apps, and then do stuff. I can speed up the app-scan by removing the 'moving pandora icon', and some stuff), but I'm not worried about it until I've got about 1000 pnd files to do stress testing :)

jeff


PND-
skeezix said:
A pnd file may contain many apps within it, wih their own categories and params. Think Vice.. C64 and vic20 etc in one pnd file. When in pndfile it can only easily get one icon for all subapps, but if unpacked it can be an icon for each subapp .. But for say windowed versus fullscreen versions of same app, same icon makes sense

ED Twitter-
Evildragon said:
Nice, just saw on http://www.pygame.org/news.html that they did post the Pandora video where I do live pygame editing :) Cool thing.
 
Last edited by a moderator:
craigix said:
Today I watched this: http://www.bbc.co.uk/iplayer/episode/b00rfgmt/b00rfgl6/Fake_Britain_Fake_Goods/

Which if you can you should watch, it's the sort of thing I need people who complain about our seemingly high prices to see.

Fixed up some heating in the hall (this isn't finished and is a big job, several huge heat exchangers required to be installed).

Did various wiring jobs, fixed a water leak, and I'm going to pick up the first of the supplies for the assembly sometime this afternoon.

Chased up the moulding company, got some details of the LCD + battery shipment prices, just made sure everything will be arriving this month and that we will have what we need to get going when it arrives.

Tomorrow I have a meeting with EBV and more work setting things up.
craigix said:
Oh yeah, you guys are so much cooler than me, I love your appstores. Oh.. wait..

In other news, screenshot and video size is now equal to half the Pandora res (widescreen), works well. Almost ready for round two.
 
Last edited by a moderator:
New picture on MWeston's photo album, looks like the battery time remaining calculator explained:
http://s198.photobucket.com/albums/aa99/MWeston_2007/?action=view&current=batt_levels.jpg
 
(Power saving, battery life)
skeezix said:
IT boots up pretty fast imho, but yeah, generally you can just..

1) Close the lid; thus turns off the screen, but leaves cpu at current (so you can play music etc); this cuts power usage down by probably half (vimacs and so on have real numbers if anyone cares)
2) Hit thw power button to toggle low and high power mode; in low power mode the CPU is down to 15mhz or something like that, the screen is off, wifi and BT are unloaded; when you toggle back to high power mode, it returns to where you were (cpu goes up, lcd back on, wifi and BT loaded if they were before.) So this doesn't shut anythign off, just turns everything way down.

On low power mode, you'll get (currently) what, 2.5 days life on a battery, but probbaly can tweek another half or full day some more out of that .. a few things were still going when we did that test.

The battery is beefy -- I just let mine sit on all day today (like 8 or 9 hours), with some usage, and killed about 3/4 of it .. it really is about 12 hours or so of not-heavy usage, or like 7+ hours of heavy usage :)

Boot up duration really depends on which environment you pick, too; if you go wiht a lighter env like pmenu or minimenu, you're up in half the time as full desktop, but full desktop lets you do a lot more (and eats more RAM away from you :)

jeff
ED(tweet) said:
notaz just managed to get solid 50kB/s WiFi connection on the Pandora! And I managed to connect my Pandora via Bluetooth PAN to my mobile :)
ED(tweet) said:
That was pure WiFi speed though. There are probably still packet losses and encryption slows it down a lot. And encryption slows down a lot.
(PSU)
Craig(tweet) said:
http://twitpic.com/17lt25 - It begins...
Craig(tweet) said:
http://twitpic.com/17lv56 - Photo of the Pandora PSU (UK) as requested.
(Bluetooth tethering)
EvilDragon said:
Well, found some more time to play around with Bluetooth.
I now managed to successfully connect my Pandora via Bluetooth PAN to my mobile phone.
I get an IP address via dhcp, I can resolve the DNS, I can ping my phone, route is setup correctly.
I just can't connect to any site on the internet, but I guess that's due to my phone - I use a custom baked ROM on my phone, so that might be the problem. Never used BT before :)

Need to find another phone I can try.

That means you should be able to use your phone if it supports bluetooth PAN for mobile internet on the Pandora on-the-go :)
Defo said:
New blog post by ED:
Keep it going.

Long time no blog post - and a lot has happened, really!

First, let's start with some OS development.
The OS is getting better and better. We will probably add another minimal menu, similar to gmenu2x with no fancy effects and graphics but PND-Support.
This is for all the purists of you.

Also, notaz and DJWillis are slowly stepping forward in getting WiFi working.
While it's not perfect, the packet loss is decreasing and things are getting better.
It might still be a long way (or it might be finished soon, you never know with such things), but it's improving.

I also successfully managed to setup a Bluetooth PAN connection to my mobile phone. I got an IP address via DHCP, I got DNS resolving and I could ping the phone.
Pinging websites outside of the phone didn't work yet, but that might be my mobile phone (custom cooked ROM, never tried Bluetooth PAN before).
As I did use the terminal to set that up (it only takes a few lines to do so), the next step is including this into the Network Manager.

Next let's go on to the Mass Production:
As most of you already know, we moved the assembly from US to UK - we'll do it ourselves!
This is because the company in US finally gave us a price for doing this - a ridiculous high price.
The boards will still be populated there, but we will assemble the units ourselves. Craig is currently setting up the village hall to turn into a proper place to do this, including buying everything needed (ESD stuff, etc.)

Our current roadmap is like this:
(note: It still depends on the case production and on the board production. This is based on the dates they did tell us - but as we already experienced a few times, we are a small client for them and therefore have low priority... let's hope they keep their dates this time!)

March 18th: Production of 1000 cases is finished and will be shipped to UK.
March 22nd: 1000 cases will arrive in UK and will be inspected. If all is well, the next 3000 will go into production.
March 22th - March 26th: 500 populated and tested boards and all parts will be shipped to UK.
Approx. April 1st: Assembly will start.

Yep, another promise broken - whereas the company in Texas told us they can test 1000 boards a week before, it's now down to 500 boards a week.
At least they kept their shipping dates for all the prototypes and testing boards so far, so March 22th - March 26th should hopefully be fine.

Some parts already started to arrive in UK, e.g. the PSUs are there.

We decided to start on April 1st to have a small bufferzone for the stuff to arrive.
It doesn't make much sense if we all sit around in the village hall without having any parts to build together :)

The complete assembly process will be broadcasted live via Webcam. That will be fun!

BTW:
The Pandora stirred up a lot of interest in Germany currently.
I've been at the CeBIT in Germany doing a presentation about the OpenSource gaming handhelds and of course the Pandora.
It was crowded.
Thanks to Radio Tux, you can listen to the presentation.
You can also download the PDF I used for it here.

They also did another interview here.

And if anybody wants to watch it, we've got the full thing available as video on Youtube here.

Beware, it's all in German!

After the presentation, a lot of guys (some from the community as well) could get their hands on the Pandora and play with it.
They all loved how well the controls and the keyboard work. Looks like our design is working well :)

That's it for today - hopefully we have good news within the next two weeks!
http://www.open-pandora.org/index.php?option=com_content&view=category&layout=blog&id=2&Itemid=2〈=en

edit: fixed links
EvilDragon said:
Well, what should I tell you?
We're probably more pissed about the delays than anyone here can be.
Because it's US who will go down with a big bang if anything happens.

We're trying to do the best we can, we eve build them with our own hands and you can be sure we try to push the companies as good as we can.
Being a small player in this big business surely is not easy, and being lied by the companies defintely doesn't make this easy - but I surely don't want to lie to you so I made this blog post which has the most recent information we have got.
(Board testing)
MWeston said:
I'm assuming the testing process is going to be much like a train ride. It chugs slowly out of the station and builds up speed before slowly getting to that final cruising speed. They had to build more test jigs in Texas in order to test more boards simultaneously and that was only ordered recently after a very long and slow process of getting the procedure right (took longer than expected for reasons I won't get into). Even though those thousand boards were built a while ago (whenever I mentioned it), testing them has been a slow process and as more test jigs get built and tested themselves (lots of soldering work and mechanical fitting of pieces, it's quite a complex rig), more human work stations can be set up to run the automated tests on them.

Some of those first rev5 boards were being manually tested like they have done on prototype boards which involves typing in all commands manually and observing the results. There's no way you would go through thousands and thousands that way but it had to be done so they could lose less time making sure their procedures were accurate and the boards were soldering properly. As more test jigs are built, the process takes more of the human out of the equation.

The idea is to automate as much as possible but a human still has to man the station and load the boards into the jig, run the software, plug in all the connectors, click Y to begin the test sequences, etc. They are keeping their estimated yields low as they chug out of the station (ie. 500...then 500 more...then hopefully 1000, then another 1000, then the last 1000).

ED tells you what we know today but that can change tomorrow. The total time frame is not going to be known until it is basically all done. ED has given the worst case but if we get things rolling smoothly, that time could start coming down more and more. The part that is killing everyone is that if we just wanted to do a power on test of each board and ship them out, we could probably do all 4000 in a week (assuming cases arrive and man power is enough) but both the Pandora team and the assembler are not satisfied with such a thing and so every circuit on the board is being tested in an effort to shoot for a 0% RMA list after launch. Imagine being that guy who finally gets his Pandora, tears open the box only to have to email support an hour later because a bunch of stuff won't work. Nobody wants to be that person. The trade off is time and a fair bit of it but if it means your Pandora runs solid for years instead of minutes, we've got to do it. There's just no other way.
(Pickle/dflemstr) DOSBox frontend
http://www.gp32x.de/board/index.php?/topic/52296-dboxfe-qt-help/page__view__findpost__p__825799
(Skeezix) PND program help idea
http://www.gp32x.de/board/index.php?/topic/52443-idea-automatic-pnd-infoinstallhelp-good-or-bad/
 
Last edited by a moderator:
Thrake said:
warum lässt ihr erst am 22.-26. die bestückten und getesteten Board nach england verschickten und nicht jetzt schon sof ort?
Weil sie es vorher nicht verschicken. Der Termin kommt von denen, wir haben ihnen gesagt, so schnell wie möglich.
http://forum.gp2x.de/viewtopic.php?p=116179#p116179

EvilDragon roughly translated said:
Thrake said:
Why are you having the populated and tested boards sent to England on the 22nd to 26th instead of right now?
Because they don't ship earlier. The date is from them, we told them as quickly as possible.





Die 1000 Gehäuse war unsere Entscheidung - wären sie irgendwie fehlerhaft sitzen wir schön blöd da und haben immense Kosten.
Bei 1000 ist das nicht so wild. Und da die Gehäuseproduktion auch deutlich schneller abläuft als wir die Dinger zusammenbauen können, sollte das nicht stören.

1000 Boards sind fertig produziert, das ist richtig. Aber es sind noch nicht alle getestet (es wird ja jede Funktion der Pandora getestet).
Da haben sie eben gemeint, dass sie 1000 pro Woche testen können - und haben sich nun (nach Testbeginn) auf 500 korrigiert
http://forum.gp2x.de/viewtopic.php?p=116188#p116188

EvilDragon roughly translated said:
The 1000 cases was our decision - if they were faulty in any way we would be in quite a pickle and have immense costs.
It's not as bad with 1000. And since case production goes much faster than we can assemble the things it should not disrupt anything.

1000 boards are produced, that is correct, but they are not all tested yet (every function of Pandora is tested). Then they meant they can test 1000 a week - and have now (after starting the tests) revised that figure to 500.





Warum hier über die 1000 Gehäuse gerätselt wird, obwohl ich das schon zwei Posts übere Nupfis schon erklärt habe ist mir zwar ein Rätsel, aber naja.
Warum die 1000 ungetesteten Boards rumliegen statt bereits versendet worden zu sein, dürfte eigentlich logisch sein - ursprünglich sollten sie ja nicht woanders zusammengebaut werden.
Jetzt haben sie eben, damit sie alle versenden können, die grossen Tests angefangen und festgestellt, dass sie nicht so schnell nachkommen.

Warum nicht früher getestet wure? Tja, fragt das Werk. Vermutlich wieder einfach die Priorität zwischen einem Kleinkunden, der nur ein paar Tausend Boards bestellt und einem Grosskunden mit mehreren zehntausend....

Glaubt mir, wir sind mindestens so angepisst wie ihr (wenn nicht noch mehr). Bei uns steckt ja die Existenz mit drin.
Wir versuchen alles so schnell wie mglich zu erledigen - aber was sollen wir denn hier noch machen
http://forum.gp2x.de/viewtopic.php?p=116209#p116209

EvilDragon roughly translated said:
It's an enigma to me why you are puzzled about the 1000 cases when I explained it two posts above Nupfi's, but whatever.
Why the 1000 untested boards are lying around instead of being shipped should be logical - originally they were not to be assembled somewhere else. Now they started the big tests so they can ship them and found that they can't do it as quickly.

Why didn't they test earlier? Well, ask the factory. Probably again the priority between a small customer ordering a few thousand boards and a big customer with several tens of thousands....

Believe me, we are at least as pissed off as you are (if not more). With us it's our whole existence that's on the line.
We try to do everything as quickly as possible - but what else can we do here





Wie Michael im Englischen Forum geschrieben hat, sind die 500 pro Woche zunächst ein Worst-Case-Scenario. Wenn in Texas mehr Teststellen aufgebaut werden (und die Tester mehr Übung haben :) ) geht das auch schneller.

Es ist NICHT geplant, nur 500 zusammenzubauen. Es kommen nur zunächst 500 an.
Der Rest wird weiterproduziert und getestet während wir bauen.

Ich muss ohnehin mehrfach nach UK und zurück, da ich mind. 1 - 2 Mal die Woche in Deutschland sein muss (als Dozent oder beim Eishockey).

Fatih setzt die Chinesen auch ein wenig unter Druck, damit sie auch wirklich die Gehäuse produzieren - er hat die Abholung schon organisiert und ihnen den Termin auch mitgeteilt... und sie haben bestätigt.
http://forum.gp2x.de/viewtopic.php?p=116254#p116254

EvilDragon roughly translated said:
As Michael has posted on the English board, the 500 per week are a worst case scenario right now. When there are more testing positions (and the testers have more practice :) ) this will be faster, too.

It is NOT planned to assemble only 500. It's just that only 500 will arrive in the beginning. The rest will be produced and tested while we assemble.

I have to go to UK and back several times, since I have to be 1 - 2 times per week in germany (as lecturer or for ice hockey).

Fatih is pressuring the Chinese a little, so they actually make the cases - he has organized the pickup and told them the date... and they have confirmed it.
 
Last edited by a moderator:
^ You are awesome, Lafazar :)

Blog said:
Didn't we all dream about that the case mass production should become reality?

Well, it has. We got some brand new pictures for you.
The mould has been successfully calibrated and mounted on the machine.
The pieces you see are the first that has been spit out by the machine, so those are testing ones which are being checked.
Not all have the finishing applied - one nice picture that does show how it will look like with finishing applied is the one with the battery lids.

Those samples are being checked - and if they are okay (and they really do look fine), they will continue to spit out the cases and apply finishing.

Enjoy!

http://www.open-pandora.org/index.php?option=com_phocagallery&view=detail&catid=21%3Acase-production&id=142-dsc06958&tmpl=component&detail=0&buttons=1&lang=de
 
Last edited by a moderator:
From Fatih;

mfk said:
The pictures are from China. We don't need to confirm anything about the cases. Everything is on track right now.
 
Last edited by a moderator:
skeezix said:
<in a hurry stream of consiousness, so don't take this for much>

Tell you the truth, with mine .. I've been so utterly busy in every second of the day with either Pandora coding or real work (damn real work is in the way!), that I've not had time to have 'fun' with it ;) (I mean, I'm a coder through, so I really enjoy coding. Real life is overtiming me tthis last year, and I'm spending as much time iwth my kid as I can.... But we're polishing as much as we can for launch in a couple weeks, so when we're not doing our day jobs, we're all working our tails off. No time for the misses, its Pandora or Work :p ) I'm pretty vocal (thats how I vent my stress right? :) , so you saw two weeks ago I started minimenu, and its 'nearly done' now; I think two weeks ago I built in the ability for pnd-files to support multiple apps per pnd, so Vice can have one pnd but offer up like 5 apps (vic20, c64, etc). I built the 'info' system to pop up install guides or app docs the other day; we've debugged a bug in the pnd system that we found; and how many other things, just in the last week or two? :) every day a couple hours work on this or that subsystem. And thats just me .. we're killing ourselves here :) notaz and djwillis have had some good progress on wifi, and djwillis got the new opengl drivers and new X drivers couple weeks ago, and they're forever doing patches for this and sorting out that. ED keeps saying he's not a developer but hes scripting and configuring like mad, so I think he's cutting himself short. Craigs working on the store and archive, etc and etc, everyone is busting a nut.etc and so on, I'm not going to list everyone.

That said ..

Pickle and cpasjuste's port of Quake3 is really coming together nicely, as I think you saw a few times. I tried it a few weeks ago, and it was pretty clumsy; tried it a week ago, and its frakking great. Default config isn't there much, so you go to Config -> Move and set the d-pad and left-nub to move and strafe, and right nub to mouse look, and shoulders to fire and jump and suddenly its pretty good. You can't top keyb/mouse for Quake, but pandora is pretty good :) Quake3 multiplayer people!

Last night I played Descent 1 (D1x in fact, which is better looking than original Descent), also from Pickle; he didn't have nubs coded up yet, but I rebound the keys so it was the d-pad and some keyb keys, and it was playable.. I beet the first level in no time :)

I've fiddle with DosBox .. few weeks ago, the old X drivers, I had problems with mouse calibration; new drivers and some guidance form Pickle and Dosbox is pretty usable, and will be great soon. I played a bit of Wizardry 7 and Civ 1 and fiddled with Warcraft 2 again. Dosbox is going to be great.

(Pandora having a keyboard is a huge thing, for me. I'm also a text adventure wanker as you may know, and working on Kronos, etc. Thats coming along nicely. I also have BattleJewels working great, Hatari, and some other stuff.)

Overall I'm quite pleased, and its going to be a great system.

Its a little bulky and heavy (the battery mostly); Theres been a couple gotchas on the case, but mine's prototype and a little different than the final units, so I'm not going to go into it.

I'm a dev, so I don't notice rough edges, and I'm sure there are some on the software; but overall its working really well imho. Video is on software and not hardsware decoding, and I'm okay with that; some peopel will be pissed they can't fire up their latest h264 etc, yet. But as I've accepted current limits (since its partly my faulkt, right? :) , I'm okay with those thing.s I know we'll fix it ove rtime as more eyes and hands get on board.

So dont' put it on a high horse.. some rough edges. And its great.,

(And for comparison.. go spend $500 on an Archos 5, its a fantastic device; oh wait, out of the box the damned thing barely works, you have ot patch the living hell out of it, to get it to work; and it still crashes every 10 minutes ;) Apple makes a polished first impression on people, but you get that polish because they restrict the heck out of it :)

jeff

In response to: http://www.open-pandora.org/images/phocagallery/Recent/Photos/CaseProduction/thumbs/phoca_thumb_l_DSC07000.JPG

craigix said:
They are cases for the battery(s) when they are out of the Pandora.

You can see the little bump for the PCB at the ends.
 
Last edited by a moderator:
(Wifi driver status)
EvilDragon said:
Well, AFAIK there are no packet losses anymore, but WPA2 is not working yet and speed is still a bit slow.
That seems to be connected to SDIO power management, as far as I did understand DJWillis and notaz.
Disabling the power management of the WiFi driver did stop the packet losses, now the speed issue comes from power management in SDIO.

There are lots of patches and fixes that need to be done to the driver, but they are getting there :)
(Old closed source Wifi driver)
notaz said:
The closed source TI driver became open after G1/HTC Dream was released thanks to Google, but it has a problem - it doesn't support Linux wireless stack properly and needs custom tools to operate. Some time later Nokia developed another driver which does fully support Linux and is part of mainline kernel, which we are using now. The old TI driver doesn't perform any better than Nokia one (after my adjustments) so there is no point using that.
Fishbong said:
Hi,
apologies if this isn't News Zone material, but i thought you'd be interested in a translation of ED's CeBit speech on open handhelds, so here you go.

http://www.youtube.com/watch?v=T89mHvizwVg
http://www.youtube.com/watch?v=k-toU59IWF4

I'll translate the other parts in the next couple of days if there is enough interest.
-------------------------------------------------------------------------------------
Joy and jubiliation! It suddenly and unexplainably works again. Perhaps my ISP is to blame?
http://www.youtube.com/watch?v=Qt27l77hoXc
-------------------------------------------------------------------------------------
http://www.youtube.com/watch?v=ZH1kYkvXWPY
-------------------------------------------------------------------------------------
What's so bad about using Firefox that you can't at least have it installed?
Anyway, fifth part:
http://www.youtube.com/watch?v=6ub_d0H3aFQ
-------------------------------------------------------------------------------------
And it's done! This is the best part of them all.

http://www.youtube.com/watch?v=dE-oxcAA5NA

I worked a bit slower, but more thorough this time.
(notaz' Pandora)
notaz said:
Mine is typically lying on the table running some minimal kernel to debug some driver or app in isolated environment:
http://notaz.gp2x.de/img/mess/pnd/cp05.jpg

Still have to use the devboard occasionally because the cased unit has broken wifi and different LCD. Haven't really played much of anything on my unit due to lack of time, but tested Q3, N64 and some other stuff.
Pickle said:
Loonie said:
Hopefully, we'll have a port of Zdoom that will cover the Heretic and Hexen side of things (along with the Doom and Strife side of things as well, of course).
be nice if they cut that fmod crud out
edit: fyi ive port hheretic and hhexen and prboom, so the only thing not supported is strife
WizardStan said:
I killed a few hours today by building Dink Smallwood for my BeagleBoard.
Sound works, but not music. This could just be because of problems with the sound mixer.
The mouse doesn't work, but the only time it's used is on the main menu screen (to select Start, Continue, or Quit) but you can move the cursor around with the arrow keys as well, so no major loss there.
It can't load the default font for some reason. Probably something to do with it expecting SDL_ttf version 2.0.9 and only having 2.0.3 built for ARM. Not sure which version will be available on the Pandora, but I got around it by just grabbing a basically random font (FreeMono.ttf) and using that instead. It's not a very pretty font, it's a little hard to read. If I can't get the default font to work, I'll experiment with some others until I find one that looks nice.
Other than that, it works pretty good. My only concern is that I'm running at 800x600, and I don't know how it will look on an 800x480 screen.
WizardStan said:
WHEEEE! I asked Skeezix to give it a try on his Pandora, and it worked beautifully. He even snapped a picture for me. It's magic!
So thanks to him I'm now even more giddy to get my Pandora than I was before, assuming that's even possible.
(file extension registration in Linux/PND)
http://www.gp32x.de/board/index.php?/topic/52485-file-extension-registry/

(Cpasjuste's Pandora SDK updated)
http://www.gp32x.de/board/index.php?/topic/52343-pandora-sdk-installer/
Cpasjuste said:
I updated the toolchain.sh script to install arm-2009q3-67 as suggested by skeezix :)
Cpasjuste said:
I just updated the "libraries.list" too with latest gles packages. (previous one was not available anymore, so it was giving an error)
Remember to reports any missing libs, i'll add them to the "libraries.list" file if available on the "final" pandora image.
(skeezix) Libpnd Api - Read Nubs/dpads, Etc
http://www.gp32x.de/board/index.php?/topic/52510-libpnd-api-read-nubsdpads-etc/
 
Last edited by a moderator:
EvilDragon on German Boards said:
Welche Art von Powersave Modus unterstützt das BS der Pandora?

Smartreflex --> Bei der CPU wird dynamisch die Core Spannung runtergesetzt.

CPU Scaling --> Die Taktfrequenz wird dynamisch geändert.
Meines Wissens dürften wir im Moment CPU Scaling im Einsatz haben, kann das aber nicht sicher beantworten.
http://forum.gp2x.de/viewtopic.php?p=116459#p116459
EvilDragon said:
Which kind of powersave mode does the Pandora OS support?

Smartreflex --> core voltage at CPU is dynamically lowered

CPU Scaling --> clock rate is dynamically changed
AFAIK we currently employ CPU scaling, but I can't answer it with certainty.



EvilDragon on German Boards said:
Ich habe den Thread fast komplett gelesen, aber immer noch ein paar Fragen:

1. Die Pandoren werden jetzt "privat" bei ?Craig? zusammengebaut. Wieviele Leute sind daran am Batch 1.1 (500 Stück) beteiligt?

Es werden immer mind. 8 Leute am Schrauben sein.

2. Wer ist am Batch 1.2,1.3 und folgenden beteiligt?
Das gleiche Kernteam + Helfer oder angeheuerte Studenten.

3. Kann sich Craig überhaupt 4 Wochen Urlaub nehmen?
Craig ist - wie ich, MWeston und Fatih auch - selbständig. Wir brauchen keinen Urlaub nehmen.

4. Hat er nach der 2ten Woche Dauer-Pandora-Zusammenschrauben überhaupt noch Lust weiterzumachen.
? Seit wann geht es beim Arbeiten um Lust?

5. Haben die Helfer noch Lust weiter zu arbeiten, nachdem sie die ersten zusammengebaut haben und fleißig am daddeln sind?
Die werden verschickt und nicht in der Halle zum Daddeln freigegeben...

6. Werden sie "privat" zusammengebaut, weil das Geld knapp wird?
Ja und nein. Ja, da wir bei Batch 1 eh schon draufgezahlt haben und das Werk nun meint, dass sie pro Pandora 100USD zum Zusammenbau verlangen müssten (was 400.000 USD mehr wären).
Nein, da wir das auch gemacht hätten, wenn mehr als genug Geld vorhanden wäre - der Preis ist einfach unverschämt.
Nein, da wir bisher auch noch ohne Investor oder Kredit gearbeitet haben und bei einem fertigen Produkt, wenn nur noch der Zusammenbau fehlt, auch hier was hätten machen können.

"Privat" ist auch ein falscher Ausdruck. Wir bauen eine Halle um, damit sie den normalen Standards für solch einen Zusammenbau bietet.
Damit hat die OpenPandora Ltd. eine Halle mit Arbeitskräften, die das durchführen zur Verfügung.
Genau wie bei Sony oder Nintendo oder anderen Massenproduktionen - nur kleiner und dass wir eben viele freiwillige Arbeitskräfte haben, die da kostenlos aushelfen wollen.
Ansonsten ist kein Unterschied: Menschen sitzen in einer Halle und Schrauben.
Wer meint, man braucht Fachkräfte für sowas: Jeder Schüler kann im Urlaub am Fließband Geld verdienen (sehr beliebt hier bei Audi). Jede Zeitarbeitsfirma verhökert Leute an Massenproduktionsstellen. Man braucht keine großartige Ausbildung für sowas.

"Privat" wäre es wohl, wenn alle auf Craigs Sofa vor dem Fernseher sitzen würden und nebenbei Pandoras mit Essensbrösel zusammenschrauben.
http://forum.gp2x.de/viewtopic.php?p=116615#p116615
EvilDragon said:
I read the thread almost completely, but still have a few questions:

1. The Pandoras will be assembled privately at Craig's. How many people are involved on Batch 1.1 (500 pieces)?

There will always be 8 people assembling.

2. Who is involved with batch 1.2, 1.3 and following?
The same core team + helpers or hired students.

3. Is craig able to take holidays just like that?
Craig is - like me, MWeston and also Fatih - self-employed. We don't need to take holidays

4. Will he after the second week of continuous Pandora assembly still be in the mood to continue?
? Since when is working about the mood?

5. Will the helpers be in the mood to continue working, after the first few are assembled and they [the helpers] are busy gaming?
They [the Pandoras] are shipped and not approved for gaming in the assembly hall...

6. Are they privately assembled because money is getting tight?
Yes and No. Yes, since we paid extra on batch 1 anyway, and now the factory means they have to demand 100USD per Pandora assembly (which would be 400.000 USD more).
No, since we would have done this even if there was more than enough money - the price is simply insulting.
No, since we have always worked without investor or credit so far and we could always have done something there since we have a finished product that only needs assembly.

"privately" is the wrong expression. We are converting a hall so it offers the normal standards for such an assembly.
With that the OpenPandora Ltd. has an assembly hall with workers who can produce this at its disposal.
Just like at Sony or Nintendo or other mass productions - only smaller and we have many volunteer workers who help us for no charge.
There is no difference otherwise: People sit in a hall and assemble.
Whoever thinks you need qualified workers for that: Every student can earn money during holidays on an assembly line (very popular here at Aldi). Every time share company sends people to mass production jobs. You don't need overly much training for this.

"privately" would be if everyone was sitting on Craig's couch in front of the TV and assemble Pandoras together with crumbs of food.



EvilDragon on German Boards said:
Nicht mehr als 300. Vor zwei Jahren waren die Teile vielleicht noch so teuer, aber heute ist sie in der jetzigen nicht das wert wofür sie verkauft wird. Also 199 Euro ist noch so gerade okay.

Hmm, wär ein schöner Preis, aber würde gerade mal die Produktionskosten decken (ohne die ganzen Kosten für Webserver, Shop-Mitarbeiterkosten, etc. zu berücksichtigen)
http://forum.gp2x.de/viewtopic.php?p=116663#p116663
EvilDragon said:
...199€ would be an ok price for Pandora...

Hmm, would be a nice price, but that barely covers the production costs (without considering all the costs for webserver, shop workers, etc.)



EvilDragon on German Boards said:
Nein, keine Pandora bisher hat ein Finish.
Beim Finish dürfte das auch alles verschwinden, denn schließlich werden da ja mit einem Sandstrahler Unebenheiten beseitigt und das ist ja auch eine (durch verschiedene Materialdicken entstandene) Unebenheit.
http://forum.gp2x.de/viewtopic.php?p=116620#p116620
EvilDragon said:
No, no Pandora has a finish applied so far.
This should all disappear with the finish, since they remove unevenness with a sandblaster, and this too is a unevenness (caused by varying material thickness

Dammit, I'm having problems with the forum software, there seems to be a limit on the number of quotes allowed inside quotes or something. I'm just using italics now for quotes inside quotes.
 
Last edited by a moderator:
(Win 3.11 in DOSBox)
Pickle said:
I loaded up windows 3.11 last night on the pandora, boots pretty fast and gui interactions is pretty good. Touchscreen mouse control needs some work, but i can control via the nubs so it wasnt too bad. Fired up minesweeper.
ED(tweet) said:
Pheww, we all did a lot of WiFi stuff during the last days. Slowly narrowing the problem. And I just packaged up Exaile as PND. Nice player.
EvilDragon said:
Yeah, we're all pretty busy with this :)
We're narrowing the issue more and more... can't tell how fast we'll get it fully working, could be today, could be tomorrow, could be in a month... but basically the whole team is working for nearly 24/7 on this... we're trying the best we can.

Please don't ask me for case / board / etc. updates, as I'm also fully concentrating on WiFi right now (tweaking, testing, etc.) so I didn't have time to follow the case production process recently... Fatih will take good care of that, I'm sure :)

Okay, back to WiFi :)
craigix said:
It's probably better to cancel your order now rather than having to pay to return your Pandora for any slight imperfections you may be unhappy with.

It's not a PSP or an iPhone.

I'll empty my PM box so you can send me your order numbers.
Dear Community, if you like this thread, please continue posting quotes and new infos. I've got some other stuff todo.
 
Last edited by a moderator:
Case quality-
craigix said:
I don't know how good the final cases will be, but I'm pretty sure they will not be perfect. Now I have a deep understanding of the moulding process I can see errors on many cases, even the DS.

What I do know though is that the Pandora case works and is nice. We were never trying to make an iPhone like flashy case. It would have cost about $500,000. What we have made is a functional case, and I like it.

So, if you want a flashy flawless case you are going to be disappointed. If you can't accept anything but a perfect case just cancel your order and buy something else.
 
Last edited by a moderator:
Regarding the current status of the manufactured cases:

craigix said:
The courier isn't picking them up until the 23rd, but they say they will be ready and waiting.
 
Last edited by a moderator:
Back
Top