New to Raspberry PI

The quote from the raspberry pi organization yesterday was "At the moment, we see returns in the order of about 0.02% of all units".  So yes, sure, they could have waited until returns were at a minimum to make that comment.  You would have to argue with them, not me.  In any event, supplying a device with no certified or included peripherals, having people order parts anywhere at any price or of any quality, and add to that many users with little or no knowledge of electronics, computers, programming, or troubleshooting skills, well I am not at all surprised that they are getting a very high return rate of working pis.  If it wasn't for these forums, it would be much, much higher.  Maybe their intent is to teach troubleshooting skills!

> The quote from the raspberry pi organization yesterday was "At the moment, we see returns in the order of about 0.02% of all units".

I have no idea how to parse that sentence.  As you note, "At the moment" could mean just about anything.

As could "we see".  Is the RPF claiming to see all the returns?

Could it mean that today's mail bag brought in about 0.02% of all units shipped this month?

They have always claimed that OK units were about half of all returns, but I see no logical reason

why the number of OK units should decline at exactly the same rate as the number of broken units.

If they are doing something to reduce the number of faulty units, like attaching the big capacitor better,

or using a more sturdy SD card socket, that would be nice to know, but we don't hear anything along those lines.

However, the Hynix memory incompatibility

alone reportedly accounts for as many as 5 OK units in a single day, which is not surprising

because the Debian Wheezy soft-float version on the official download page

The official distribution is the Raspian hard float version so unless you have some numbers to show that the

Debian distribution is a significant part of the problem I don't see how that is relevant.

as does the latest Fedora download,

Ditto.

And there are lots of reasons besides the Hynix issue for people to

mistakenly believe their RPi is faulty.

Real strong point you're making there.

Since Farnell manufactures all their RPi's in the well-regarded Sony plant, I would expect their

return rate to be even less than the overall rate, giving E-14 something to really brag about.  

But nothing but silence.

Which proves absolutely nothing at all unless there is some rule I am not aware of that requires

companies to brag about how well they do their jobs.

I have no idea how to parse that sentence.  As you note, "At the moment" could mean just about anything.

As could "we see".  Is the RPF claiming to see all the returns?

Could it mean that today's mail bag brought in about 0.02% of all units shipped this month?

How about these things vary month to month statistically speaking and this is what we are seeing now.

Unless of course there is some deeply hidden and sinister ulterior motive that lurks behind the statement.

If they are doing something to reduce the number of faulty units, like attaching the big capacitor better,

or using a more sturdy SD card socket, that would be nice to know, but we don't hear anything along those lines.

Again this proves nothing. They are not required to tell you or anybody else what their QC is (or is not) doing.

It would be nice if they did, but I have not seen either of the two problems you specifically mentioned on any

of the three I own. I especially like the big capcitor one as I have not heard anything about this for many months

now. Maybe they fixed it and just considered it something they should do without bragging about it. Then again,

maybe not .....

Now I remember why I stopped coming to this forum.  Sean, if you have any other thread related questions feel free to ask.

> The official distribution is the Raspian hard float version so unless you have some numbers to show that the Debian distribution is a significant part of the problem I don't see how that is relevant.

Yes, the download page says Raspbian hard float is the "recommended" OS,

but that doesn't mean it's the only one that gets downloaded.  The others are

there for good reasons, such as for use with the Oracle JVM.

> as does the latest Fedora download,

> Ditto.

Yes, Fedora has a relatively small share of the RPi market. 

But even a small share of 100K units/month is likely to be

more than single digits, especially when combined with the

other listed sources of obsolete firmware.

The RPF's own figures of 5 Hynix issues per day are inconsistent

with a total of 10 mistakenly assumed faulty returns per month.

>> And there are lots of reasons besides the Hynix issue for people to mistakenly believe their RPi is faulty.

> Real strong point you're making there.

You missed the point completely.  If 0.02% of all units are returned, and 50% of those are mistakenly

believed to be faulty, then about 10 units per month are mistakenly believed to be faulty,

including 5 units per day due to the Hynix issue.  This leaves very little room for

all other reasons why people might mistakenly believe their RPi is faulty.

Edited to add:

On June 1, 2012, Eben wrote:

"Right now the RMA rate is roughly 1 in 500, and only around half the returned units that I have seen are actually defective;"

http://www.raspberrypi.org/phpBB3/viewtopic.php?f=28&t=6822&hilit=eben&start=65

On May 08, 2013, Liz wrote:

"At the moment, we see returns in the order of about 0.02% of all units ...As usually happens when we troubleshoot returns, though, we found that around half of Jessica’s Pis weren’t faulty at all."

http://www.raspberrypi.org/archives/3866

Since Liz's quote is so similar to Eben's ("At the moment" vs. "Right now", 1 in 500 vs 0.02%, half are defective vs half not faulty),

I think it's entirely possible that Liz is quoting Eben rather than providing updated statistics.

Liz may have simply shifted the decimal point by one place by mistake

when converting 1 in 500 (0.2%) to 0.02%.

That would explain why there was no note of a ten-fold reduction in

the return rate, which otherwise would have been noteworthy.