PDA

View Full Version : audiophile cred



bill
2005-02-16, 23:45
Funny thing, I was just talking to an audiophile who owned a Squeezebox and
returned after several weeks because the digital output STILL swaps channels
on a frequent basis. This is a known and still unaddressed bug (correct me
if I'm wrong). Can't imagine any audiophile could live with it.

I own three SBs, but would jump on another product that was similarly priced
and with remote controllability and played flac and MP3 and had decent
customer support. And buy that, I don't mean just a good user forum. I've
actually seen SlimServer's quality go DOWN over the past year or so.
Regression testing and stability have gone out the window in favor of adding
every feature conceivable. I obviously more fun to code new features than do
regression testing and refactoring to make a more stable and user friendly
product, but SB is a commercial product and a certain minimum level of
functionality needs to be maintained. It would be nice if Slim would decide
what that is and then guarantee their products to that level by regression
testing every Server and Firmware release. The correct orientation of
channels on the digital out should be on that list.

And replying to these issues with "It's open source, shut up and contribute"
doesn't wash. I paid good money for the product and it should function
consistently. If I'd paid nothing, I'd have nothing to bitch about.

All IMO of course,
-bill

Daniel Cohen
2005-02-17, 00:13
On 16/2/05 at 11:45 pm -0700, bill wrote
>Funny thing, I was just talking to an audiophile who owned a Squeezebox and
>returned after several weeks because the digital output STILL swaps channels
>on a frequent basis. This is a known and still unaddressed bug (correct me
>if I'm wrong). Can't imagine any audiophile could live with it.

This bug can, as you say, make some people refuse to own a
Squeezebox, and will annoy many others.

Unfortunately,as has been discussed, it isn't something Slim Devices
can deal with easily. The problem is in hardware, in the main chip
IIRC, not software. They have indicated that they are talking to the
hardware manufacturer. For all we know, they may be investigating
other chips, but that would need a considerable time to evaluate.
--
Daniel Cohen

Mark Bennett
2005-02-17, 01:07
On Wed, 2005-02-16 at 23:45 -0700, bill wrote:
> Funny thing, I was just talking to an audiophile who owned a Squeezebox and
> returned after several weeks because the digital output STILL swaps channels
> on a frequent basis. This is a known and still unaddressed bug (correct me
> if I'm wrong). Can't imagine any audiophile could live with it.

This bug only comes into play when switching between MP3 and PCM mode.
As long as you stay in either MP3 or PCM, then everything is correct.

Any audiophile worth his speakers will be streaming exclusively in
PCM mode, so the bug is unlikely to become evident.

--
"The biggest problem encountered while trying to design a system that
was completely foolproof, was, that people tended to underestimate the
ingenuity of complete fools." (Douglas Adams)

kdf
2005-02-17, 01:31
Quoting bill <wildbill14 (AT) mindspring (DOT) com>:

> Funny thing, I was just talking to an audiophile who owned a Squeezebox and
> returned after several weeks because the digital output STILL swaps channels
> on a frequent basis. This is a known and still unaddressed bug (correct me
> if I'm wrong). Can't imagine any audiophile could live with it.
>
> I own three SBs, but would jump on another product that was similarly priced
> and with remote controllability and played flac and MP3 and had decent
> customer support. And buy that, I don't mean just a good user forum. I've
> actually seen SlimServer's quality go DOWN over the past year or so.
> Regression testing and stability have gone out the window in favor of adding
> every feature conceivable. I obviously more fun to code new features than do
> regression testing and refactoring to make a more stable and user friendly

refactoring is one of the major priorities of 6.0. The DB backend is long
overdue and is solving a lot of ongoing issues. Memory optimisation is also
something under scrutiny during this process. By the time 6.0 hits official
release status, I hope you will be pleased with the improvement.

> product, but SB is a commercial product and a certain minimum level of
> functionality needs to be maintained. It would be nice if Slim would decide
> what that is and then guarantee their products to that level by regression
> testing every Server and Firmware release. The correct orientation of
> channels on the digital out should be on that list.

What makes you feel this issue has somehow been considered proper performance?
I believe that the source of this problem lies in the handes of the chipset
provider at this point, so its simply not something that can be turned over
with the same speed as a software fix.

I personally think that the openly available nightly builds are in fact shooting
themselves in the foot because so many seem to treat these as 'official'
releases and get very upset when one thing or another breaks. So, what appears
on this list is often a discussion on the problems of the bleeding edge or
difficulties with a very complicated plugin.

What I'd like to see here is what you think should be officially guaranteed for
function of squeezebox and slimserver as far as regression testing, and what
timeframe you expect these tests to take. Bear in mind that a known bug
that is not yet fixed isn't something that regression testing solves in any
way. Given this list, lets see how others want to respond and add to it or
even argue over what the focus should be.

I think that the process is improving fairly well, but there is always room for
improvement. There is now a 5.4.1 branch
that has been fixing 5.4.0 problems while maintaining something stable. yes,
certain bugs still exist, but sometimes a fix isn't always easy. That doesn't
make it any less annoying for those who suffer greatest from it, but its an
unfortunate fact.

> And replying to these issues with "It's open source, shut up and contribute"
> doesn't wash.

nor does blaming that response on the people who supply you with the Squeezebox.
Slim Devices does not support this policy (I know from experience ;).

cheers,
kdf

Roy Owen
2005-02-17, 09:30
<Rant>
Sheeze,
Cut SilmDevices some slack here. First of all SB is a fairly new
product. Secondly, having a problem with a main component like the
Digital Out circuit is not something that can be fixed overnight. My
programming background started in Digital Circuit Design, so I do know
what I'm Talking about. The IC Mfg. may not even know why their chip
is acting like it does, it will most likely take months to figure out
why the defect exists, several more months to come up with a proposed
solution, several more months to create the test chips... You see
where I'm going with this. I don't think that the chip is flashable
so you need the time to make the new masks for the fabrication
process. My educated guess is that if the mfg announced today that
they knew why the defect exists and they were going to start the work
of re-doing the chip it would be close to a year before the chip was
on the market. Digital Circuit Design IS darn near Rocket Science.
</Rant>

Now, I agree with Carl Maskelyne. I got my SqueezeBox so I could have
Internet Radio, and all my tunes any where my WiFi reaches. Back in
the early '80s I was an audiofile, but then I had kids. Now I just
like to listen to music other than the cookie cutter stuff provided by
Clear Channel et al.

To Slim Devices:
Keep up the exceptional work, I love my SB and I look forward to the
release of SlimServer 6.

Roy Owen - Satisfied Customer, Software Developer, and Dragon incognito
--
Do meddle in the affairs of Dragons,
for you are crunchy and good with Tabasco.

Mark Bennett
2005-02-18, 01:22
On Thu, 2005-02-17 at 07:13 +0000, Daniel Cohen wrote:
> On 16/2/05 at 11:45 pm -0700, bill wrote
> >Funny thing, I was just talking to an audiophile who owned a Squeezebox and
> >returned after several weeks because the digital output STILL swaps channels
> >on a frequent basis. This is a known and still unaddressed bug (correct me
> >if I'm wrong). Can't imagine any audiophile could live with it.
>
> This bug can, as you say, make some people refuse to own a
> Squeezebox, and will annoy many others.
>
> Unfortunately,as has been discussed, it isn't something Slim Devices
> can deal with easily. The problem is in hardware, in the main chip
> IIRC, not software. They have indicated that they are talking to the
> hardware manufacturer. For all we know, they may be investigating
> other chips, but that would need a considerable time to evaluate.

Are we sure that this is a hardware problem? I understood that the
chip provider also provided some of the firmware, and in particular
the low level hardware drivers. I haven't seen anything from Slim
that says it's a hardware problem, just that it is a problem that
is in the chip providers hands.

If it is low level firmware then Slim need to keep applying pressure
to get a resolution. If it's hardware, then there a good chance that
there'll never be a fix unless they come up with a firmware
workaround.

As others have said, making chip revisions, even when you know the bug,
is a time consuming and expensive business.

--
"The biggest problem encountered while trying to design a system that
was completely foolproof, was, that people tended to underestimate the
ingenuity of complete fools." (Douglas Adams)