Eds Status On The Pandora Os - Disheartening

[Grench]

WiFi - We -need- to get 802.11g working at a reasonable speed. This is one of those -core- components of the platform. It's like saying that the CPU only works on Thursdays. In my very humble opinion, without wireless working, the device isn't.

I know that there are high hopes of the 'community' fixing all of these 'software' issues - but this one merits OpenPandora hiring someone to get it solved.

This is the exact scenario where developer fund $$$ should be spent. Something that we -must- have for the device to be properly functional.

That developer fund wasn't supposed to be a party/slush fund - right? It wasn't a 'donation' to the OpenPandora team. If getting wireless working isn't a qualified project for it, I don't see what would be.

 

[WizardStan]

I still wonder how slow, slow is, are we talking 11mbs apposed to 54mbs or are you talking about 1mbs speeds?

ED answered that in the other thread-

Sorry, my fault... it should be 1200bps approx
So REALLY slow.

I suspect that 1200bps is not correct. He went on to say that MWeston was web surfing. 1200bps is slow enough to give someone an aneurysm on todays web. Trust me, I surf at 14400bps when I visit my parents, and it's enough to make me not want to ever visit them again. At 1000bps, the index page of this forum would take between 20 and 30 minutes to load. Forget about looking at any posts, with all these avatars flying around.

 

[PhonicUK]

I don't care if there are OS issues really, getting the hardware in my hands is the important bit at this point. The OS and software can take time and I suspect will greatly accelerate once the Pandora is in our hands.

 

[VRAndy]

This doesn't bother me, never has, and never will! THIS IS AN OPEN SOURCE COMMUNITY PROJECT!

It's a for-profit business manufacturing a proprietary piece of hardware.

 

[TrashyMG]

This doesn't bother me, never has, and never will! THIS IS AN OPEN SOURCE COMMUNITY PROJECT!

It's a for-profit business manufacturing a proprietary piece of hardware.

Hardware maybe, software is all in the hands of the community.

 

[MDave]

I still wonder how slow, slow is, are we talking 11mbs apposed to 54mbs or are you talking about 1mbs speeds?

ED answered that in the other thread-

Sorry, my fault... it should be 1200bps approx
So REALLY slow.

I suspect that 1200bps is not correct. He went on to say that MWeston was web surfing. 1200bps is slow enough to give someone an aneurysm on todays web. Trust me, I surf at 14400bps when I visit my parents, and it's enough to make me not want to ever visit them again. At 1000bps, the index page of this forum would take between 20 and 30 minutes to load. Forget about looking at any posts, with all these avatars flying around.

If people are going to be doing firmware updates directly on the Pandora (for convenience sake) then the first firmware update will probably take a very long time Perhaps in that case, it would be better to release just the wifi update first so its a nice tiny download compared to a full firmware flash, which comes after the wifi update.

But hey, we have SD and USB to update the firmware too so this doesn't matter.

 

[atomicthumbs]

1200kbps is fine for me since my DSL gives me about 1200kbps

 

[MWeston]

The speed is about 50KB/s (400kb/s) which is slow but the real problem is that the driver loses packets constantly which means even loading a simple page like Google can be a chore. I've loaded the GP32X page successfully but then trying to click into a topic would stall forever and I'd just give up.

My understanding is that the kernel needs to move to a newer version where a better SDIO driver exists and access to an IRQ driver so that data won't get lost. DJWillis also tells me he is trying to backpedal that code into the kernel we have now in order to get it working better. He's also the guy building in new GL drivers and bug fixes and PND fixes and all this other stuff so it's a lengthy process.

 

[MonkeyChops]

Thanks for the info, MWeston, sounds like we all need to buy DJWillis a beer for all his hard work.

 

[notaz]

Im not here to blame any developers itself (who probably do it all for free in their own free-time), only the OP company.

Which is composed of a bunch of forum dudes and some devs themselves?

My understanding is that the kernel needs to move to a newer version where a better SDIO driver exists and access to an IRQ driver so that data won't get lost.

SDIO doesn't work on latest OMAP kernels too. There are some patches being posted for that but there are reports of basic SD card functionality breaking because of them. It's all still under heavy development..

 

[urjaman]

The speed is about 50KB/s (400kb/s) which is slow but the real problem is that the driver loses packets constantly which means even loading a simple page like Google can be a chore. I've loaded the GP32X page successfully but then trying to click into a topic would stall forever and I'd just give up.

My understanding is that the kernel needs to move to a newer version where a better SDIO driver exists and access to an IRQ driver so that data won't get lost. DJWillis also tells me he is trying to backpedal that code into the kernel we have now in order to get it working better. He's also the guy building in new GL drivers and bug fixes and PND fixes and all this other stuff so it's a lengthy process.

Just a quick question; How well does the bluetooth work? Has it been benchmarked/tested properly? (I'm asking this because IIRC bluetooth is in the same chip as WiFi and so it should also be over that SDIO link. Correct?)

EDIT: typo

 

[MrBlais]

^This..

I've been wondering this too

 

[notaz]

Just a quick question; How well does the bluetooth work? Has it been benchmarked/tested properly? (I'm asking this because IIRC bluetooth is in the same chip as WiFi and so it should alse be over that SDIO link. Correct?)

Wrong - BT is using one of UARTs. I've only tested basic OBEX transfers and they do work, no benchmarks yet.

 

[WoD]

DJWillis deserves far more than just a beer. He's taking on the workload of the tens, if not hundreds of contributors and testers who would otherwise have worked on the operating system. The delay of the Pandora delayed the OS, too, why do people find it so hard to understand that this is a community project that requires the work of unpaid, interested volunteers?

I lament that the Pandora has attracted far too many people who "don't get it," instead of whining on the forums why not just cancel your pre-order and go away? Or, here's a novel idea, why not learn to program and HELP get the system up to scratch when it's launched?

That is the very idea that the Pandora is built upon, yet you come here expecting everything to be finished and ready for you to run off and play, without so much as a nod toward the people who put in days of their time, with little reward but their own sense of satisfaction at a job well done, to get it to a usable state.

I hate to bring the tone of the thread down, but this sort of attitude really gets my goat. I read that latest post and saw it as overwhelmingly *good* news. The system is stable enough to keep all the emulator leeching whiners at bay, at least for a month or two.

 

[urjaman]

Wrong - BT is using one of UARTs. I've only tested basic OBEX transfers and they do work, no benchmarks yet.

OK. Thanks for the info. UART bluetooth should atleast work properly (from what i've heard of them in general). Any pointers to info on the TI chip's bluetooth system? (I just realized that i know _nothing_ about it more than "bluetooth" ... Is it 1.1 or 2.0? EDR?)

 

[mali]

- integrated Bluetooth 2.0 + EDR (3Mbps) (Class 2, + 4dBm)

 

[urjaman]

- integrated Bluetooth 2.0 + EDR (3Mbps) (Class 2, + 4dBm)

Oh, openpandora.ca (I checked the .org site because i thought i had seen it somewhere..) So, we have an UART set at how high baud rate? (atleast 3Mbps?).

 

[Sphinxter]

Sorry but never being able to get an actual status, estimation of time, level of effort or cost before now and a big list left at crunch time are all symptoms of a lack of proper project management.

 

[NLS]

Yes, I've been thinking a few times "What if the pandora was released in late 2008 or early 2009? Would there be any good functionality in the Pandora even if it was finished hardware wise?"
Oh well, now it's taken some time, so the software/OS side o things is probably better now than it could have been 1 year ago.

 

[MWeston]

- integrated Bluetooth 2.0 + EDR (3Mbps) (Class 2, + 4dBm)

Oh, openpandora.ca (I checked the .org site because i thought i had seen it somewhere..) So, we have an UART set at how high baud rate? (atleast 3Mbps?).

The BRF6300 chip allows up to 4Mbps transfers but OMAP3 unfortunately only goes to 3Mbps so that is the rate. There is also a McBSP port connected to the BT chip but Notaz didn't have any luck streaming audio across it. The BlueZ driver has changed a lot since then but no one has gone back to try again as far as I know.

BTW, I tried BT last week on a recent image. My laptop found it and I tried to connect. I got a message on the Pandora LCD that my computer was trying to access it so I chose to allow it. My laptop then asked for a passcode so I typed 0000 which is the only default one I know. Pandora didn't like that code and shut down the link and my laptop told me it was denied. That's as close as I got to testing it but I probably just don't know what I'm doing.