Pandora, Looking Good!
- Thread starter Findingmyzen
- Start date
I am also bothered by the delays (I'm starting to look silly in front of all the people I told the pandora would be out by now), although it's not like I blame anybody. I mean, there will always be improvements and such, so I don't expect this to be some sort of super device that will never become obsolete. My attitude at this point is sorta "Just give me the thing as advertised," and I think that's exactly what the developers are doing.
Perhaps it would be better to report the expected release date as a few months different than the developers really think it will be. That way, no one will be disappointed even if it's delayed another few months, because they'll never know it caused a delay in the first place.
I hope that's what they did with this November release idea.
Another issue is that I'm just not sure what's taking so long. I feel like the whole thing is sorta done already, and the design is just sitting there, waiting for production.
Don't get me wrong, quality control is important, but maybe the devs could gain even more credibility by setting a reasonable timeline and sticking to it.
Of course, I don't think I know of a single piece of revolutionary hardware that didn't suffer from some sort of delay, so it's not like this is unexpected and unacceptable. Just a little annoying. I know it's for the best.
Perhaps it would be better to report the expected release date as a few months different than the developers really think it will be. That way, no one will be disappointed even if it's delayed another few months, because they'll never know it caused a delay in the first place.
I hope that's what they did with this November release idea.
Another issue is that I'm just not sure what's taking so long. I feel like the whole thing is sorta done already, and the design is just sitting there, waiting for production.
Don't get me wrong, quality control is important, but maybe the devs could gain even more credibility by setting a reasonable timeline and sticking to it.
Of course, I don't think I know of a single piece of revolutionary hardware that didn't suffer from some sort of delay, so it's not like this is unexpected and unacceptable. Just a little annoying. I know it's for the best.
Pleng
Well-Known Member
- Joined
- Dec 28, 2006
- Messages
- 3,030
Nice analogy, with the exception of the fact that this kind of time line for this kind of project with the level of resources they likely have is, is pretty much common. The only companies that can avoid these major delays due to major third party components ( especially the main proc ) are ones that develop their own SoC. Most other companies just have to live with it, do what they can and hope for the best, especially at this point in the game.
Even big companies can get it wrong. When Sega released the 32X add-on for the megadrive, despite the massive amounts of funding they had, the first batch was buggy and I believe had to be recalled. This, IMHO was the first nail in Sega's coffin as far as being a console manufacturer was concerned!
Interestingly, Firefox's spellchek doesn't recognise the word Megadrive, it suggest I replace it with Megadeath! (rock on..?)
Even big companies can get it wrong. When Sega released the 32X add-on for the megadrive, despite the massive amounts of funding they had, the first batch was buggy and I believe had to be recalled. This, IMHO was the first nail in Sega's coffin as far as being a console manufacturer was concerned!
Interestingly, Firefox's spellchek doesn't recognise the word Megadrive, it suggest I replace it with Megadeath! (rock on..?)
Last edited by a moderator:
WizardStan
Mega GP Mania
- Joined
- May 24, 2008
- Messages
- 16,731
Pleng said:
Last edited by a moderator:
Karel Jansens
Active Member
You may think he's boring, but no matter how you twist it, it's "thanks" to Chad that today I've learned of another potential problem the Pandora developers have averted.gibberish said:
Every project needs its dedicated devil's advocate. Heck, I should know...
Last edited by a moderator:
chad78
Active Member
nevermind.
ldesnogu
Well-Known Member
What you say is completely wrong (except for simple SoC replacementMegamixman said:
). As I hinted, this bug has an easy workaround and NEON software runs fine, though slightly slower than it should. If you don't trust that, look at BeagleBoard videos running HD; that uses NEON assembly.
Last edited by a moderator:
Kyo
Well-Known Member
Get your torches and pitchforks, we're going to the TI headquarters!
likwidoxigen
Member
jbr said:
Come on guys. You can't be that dumb. You ever tried to do something like this before? I mean seriously have you EVER tried to do something like this? Even if you work at a large company that handles this (motorola,nokia,fic, etc.) You're still in a VASTLY different setting because those contracts aren't just that one device you're looking at millions of dollars. For the Pandora? They're not looking at nearly that much, so the Pandora is gonna be the first project slighted. If you have to Press out 3,000 pandora cases by the end of the month and samsung comes to you with a new phone they need 20,000 of in 3 months you're going to drop the pandora and get on that samsung contract. Now look at how many different manufactures are responsible for each component and realize that the Pandora is gonan be near the bottom of EACH of their lists.tmcnair said:
If any of the dev's feel that my comments are terribly incorrect feel free to mod the post or tell me so. If you're not a dev involved with this aspect and you disagree, i really don't care that much and if you think you've got a valid point. Then make it.
Last edited by a moderator:
craigix
Mega GP Mania
If only we did have the money to throw at problems to make them go away without weeks of hard work .... but we don't.
Other companies might spend 3-4 million developing devices and they still make mistakes.
We have to replace money with hard grafting , ie learning to do everything ourselves because we can't afford to pay huge companies tens of thousands (sometimes hundreds of thousands) of dollars to fix things for us.
Yes it would have been nice to have had that LCD cable the right size for the German show, ah but that would have been a one off loss of over $1000 and it would still not have been right, so it would have had to have been done again later.
Yes it would have been nice to have SLA'ed a case for the show too, rather than FDM, but at $2500 minimum per SLA would it not be better to wait until the case is 100% (ie the LCD ribbon and the case live in harmony).
That is an example of saving money so that you can buy the Pandora way under the cost of devices using the same SOC (look at the archos).
You either get a $330 next gen hand held soon or you get a $500 nex gen hand held last month (with a bug in the silicon).
Other companies might spend 3-4 million developing devices and they still make mistakes.
We have to replace money with hard grafting , ie learning to do everything ourselves because we can't afford to pay huge companies tens of thousands (sometimes hundreds of thousands) of dollars to fix things for us.
Yes it would have been nice to have had that LCD cable the right size for the German show, ah but that would have been a one off loss of over $1000 and it would still not have been right, so it would have had to have been done again later.
Yes it would have been nice to have SLA'ed a case for the show too, rather than FDM, but at $2500 minimum per SLA would it not be better to wait until the case is 100% (ie the LCD ribbon and the case live in harmony).
That is an example of saving money so that you can buy the Pandora way under the cost of devices using the same SOC (look at the archos).
You either get a $330 next gen hand held soon or you get a $500 nex gen hand held last month (with a bug in the silicon).
mindlord
Notices Two Things
It's for these reasons that I am glad that those mysterious forces of nature granted my genetic make-up and environmental upbringing with a more than modest amount of patience.
FRZ
Active Member
craigix said:
Last edited by a moderator:
emil10001
Active Member
- Joined
- May 19, 2008
- Messages
- 669
I would rather deal with 'soon' as opposed to 'there's a hardware bug in your device.'
Hardware bugs are much more final, especially if it's not big enough to warrant a recall but not small enough to be unnoticeable, or some sort of a hinderence.
I'm looking forward to a pre-order when it's ready. I've got other stuff to keep me busy until then. Like my beagleboard that's supposed to arrive in a couple of days.
Edit:
@FRZ - TI makes the OMAP3 System On a Chip (SOC) on the pandora board. The bug was with the OMAP3 chip itself. I've been seeing discussions about this on the beagleboard irc channel.
Hardware bugs are much more final, especially if it's not big enough to warrant a recall but not small enough to be unnoticeable, or some sort of a hinderence.
I'm looking forward to a pre-order when it's ready. I've got other stuff to keep me busy until then. Like my beagleboard that's supposed to arrive in a couple of days.
Edit:
@FRZ - TI makes the OMAP3 System On a Chip (SOC) on the pandora board. The bug was with the OMAP3 chip itself. I've been seeing discussions about this on the beagleboard irc channel.
MWeston
Internal Development
Well, since the forum wasn't working for me all of yesterday, I'm going to reply on this bug issue today. It seems like some of you are getting a bit carried away with what it means.
Laurent set you straight and you can trust his words. He knows what he is talking about. All chips have bugs in them. It's called errata and there's probably a nice big list for the chip in the computer you are sitting in front of right now.
The NEON bug affects ONE instruction and it causes an infinite loop that hangs the chip if some memory isn't cleared before executing that instruction. I'm forgetting the specifics but basically you have to add instructions in front of the bad one that clear this memory/register/cache/whatever it was so that the instruction succeeds.
All I am saying is that it is my understanding that the next revision of silicon is close and will probably time well with Pandora's launch so I'd like to try and get that chip for the launch system. It isn't likely TI spent millions upon millions of dollars fabbing a new chip just to fix one little bug with a software workaround. There are probably going to be lots of other fixes and I am a bit of a perfectionist, always looking to make my work as perfect as possible. Usually bugs with software workarounds get left there for years! Just take a look at the i.MX31 errata (to cite recent experience).
Laurent set you straight and you can trust his words. He knows what he is talking about. All chips have bugs in them. It's called errata and there's probably a nice big list for the chip in the computer you are sitting in front of right now.
The NEON bug affects ONE instruction and it causes an infinite loop that hangs the chip if some memory isn't cleared before executing that instruction. I'm forgetting the specifics but basically you have to add instructions in front of the bad one that clear this memory/register/cache/whatever it was so that the instruction succeeds.
All I am saying is that it is my understanding that the next revision of silicon is close and will probably time well with Pandora's launch so I'd like to try and get that chip for the launch system. It isn't likely TI spent millions upon millions of dollars fabbing a new chip just to fix one little bug with a software workaround. There are probably going to be lots of other fixes and I am a bit of a perfectionist, always looking to make my work as perfect as possible. Usually bugs with software workarounds get left there for years! Just take a look at the i.MX31 errata (to cite recent experience).
Tensuke
Active Member
Thanks for the explanation MWeston, and I too agree that I'd rather wait and have hardware with less bugs than get something sooner and have to use workarounds to fix things that shouldn't be there!
anexanhume
Member
- Joined
- Nov 15, 2007
- Messages
- 412
I know this is off topic, but I'd love to see a BOM for this thing. I'm most interested in seeing the prices for the SoC and the LCD.
As regarding to criticism being shown, it's all really irrelevant. This level of "access" to the design developments is unprecedented, and this type of approach to building the device is novel, especially with regards to resources, as Craig puts it. While the creators may make what turn out to be false claims, they are just as much victims of circumstance, and are new to the challenges that such a project entails. Therefore, any criticism levied is really without base as there is no comparative previous scenario. Frustration is understandable, but as many video game developers remark "It will be done when it's done."
As regarding to criticism being shown, it's all really irrelevant. This level of "access" to the design developments is unprecedented, and this type of approach to building the device is novel, especially with regards to resources, as Craig puts it. While the creators may make what turn out to be false claims, they are just as much victims of circumstance, and are new to the challenges that such a project entails. Therefore, any criticism levied is really without base as there is no comparative previous scenario. Frustration is understandable, but as many video game developers remark "It will be done when it's done."
FRZ
Active Member
anexanhume said:
Last edited by a moderator:
Squidge
Certified Guru
I doubt TI will want Craigix disclosing the price they are paying for the SoC, and it most likely varies with quantity anyway.anexanhume said:
It's the same with most manufacturers, they'll give you a quote and expect you to treat it as confidential, as if someone else finds out they paid more, there could be a lot of upset people around.
Last edited by a moderator:
ldesnogu
Well-Known Member
anexanhume said:
Last edited by a moderator:
