Hi Dave,
answers below. I must admit I was not expecting this in a thread with
this title, but I spotted it because I had replied the other day. I hope
I can address all your questions.
Best,
Chris
Am 29.02.2012 15:00, schrieb cctalk-request at
classiccmp.org:
I do have a Kryoflux board too ? I got it mainly for testing purposes.
The Kryoflux team does not seem interested in working together with
interested people at the same level as Phil ?
Are we? In fact many people working
on KryoFlux are contributors.
Several people work with the source, and we recently even visited a KEEP
(
http://www.keep-project.eu/ezpub2/index.php) workshop to discuss and
exchange with others, to name just one occasion. I never got a request
or proposal from you, but if you have ideas or comments, or whatever to
share... please let us know!
If you want to browse our website... there is a ton of information we
shared (and still do) over the last decade.
their actually-useful
analysis software is very expensive and even dongle protected,
according to the FAQ at
http://forum.kryoflux.com/viewtopic.php?f=3&t=3#p1241 .
Yes it is, so are many products. It must admit I don't get how this is
related to KryoFlux, or us being interested?
I have several products from the same company, some are cheap, others
are more expensive. Some come with a dongle, others activate only
(yikes!). People that want it can contact us. Some did/do. And some use
the software as part of a special arrangement. I really don't see why
this is wrong.
The basic DTC
software is closed-source and can be rather temperamental, as is the
firmware.
Is it? I'd be happy to hear about quirks. I in fact have many posts
and
mails that report the opposite. People are in fact happy it works so
well. If anything is broken, please send in a bug report. I haven't seen
one and I really wonder about the firmware. How would one notice that
it's currently "temperamental"?
I wonder what this has to do with the status of the source. I'd say it
does say nothing about quality. I in fact know many quality products
that are not open but I think they are very well executed. Does that
mean being open but flawed would be more desireable?
Though Christian says the Kryoflux "never had
such a
problem," I have never seen an independent analysis or source code or
anything confirming that.
We don't have anything to hide. We do have eleven
years of field
experience in regard to preservation. The problem outlined by Johannes
was a design fault, it's not a glitch. You must not have ambiguity in
encoding. Apparently it was fixed lately, a year after release.
I may end up doing this myself. Without
source it's definitely more work to verify it.
Let me know how we can help
you.
It feels a bit odd to read the above when in fact you never asked for
anything of the above. How should we know you would want to audit it?
An overview of the engineering of the Kryoflux board
indicates it was
designed mainly with low cost in mind. The level buffers on it are
very ESD-sensitive and are not designed for cables (so a long floppy
cable may cause issues), and there's lack of protection overall,
especially in the power supply.
I think this is very easy to explain... it was made
to be as cheap as
possible without sacrificing ingestion quality. We like to make it
affordable, and we'd even make it cheaper if we could. We don't want it
to be more expensive. The buffers work very well, and we haven't had a
single dead one in more than 600 units sold. In regard to protection:
One dead board in 600 units supplied, the PSU was broken and just fried
the board. We exchanged it. Should I have made all 600 more expensive to
cover that one case? That's 599 vs 1. And again: We don't have to hide
anything - so I am telling you here in public.
There's no way the current Kryoflux design ever
will be able
to image high-RPM media.
That's correct. The software could, but the hardware
defnitely can't. We
never targeted harddrives. We focussed on one specific field of use. And
it's working very well. If we'd want to do harddrives or other high-RPM
media, we would do another model or product.
If these things were cars... we wanted to make something that drives,
not something that floats or flies. Just one thing... but very well
executed.
As for Linux support: I've seen several complaints
lately on the
Kryoflux forums about libusb and glibc issues.
Right, you did not want to mention
that this is more or less a
"theoretical problem"? We offer 32bit and 64bit builds. The 32bit worked
for all users (as far as I know, and also on 64bit systems), the 64bit
version needed to be refined to make it work on as many distributions as
possible. That's just the way it is with binaries on Linux. I am not
really happy with that.
But on the other hand it satisfies many customers. Just install the
software, attach the board - have fun. Buying a bread instead of always
baking your own isn't that bad.
I wonder how many users will be able to compile a source on Windows or
Mac, just to shed some light on the other side of the coin. It's not all
black and white.
Take a look:
http://forum.kryoflux.com/viewforum.php?f=3 . Even though they're
somewhat responsive, it's still relying on them for support.
I don't get
the meaning of this but... if you mean we need a bug report
to remove a bug... yes, we do. The wording "somewhat responsive" makes
me wonder... people usually get replies the next working day, and even
fixes come along pretty quickly. What is wrong with that? To make it
better, we have to listen. We do.
There's also the issue of the Kryoflux software
lagging behind.
Currently the Mac version (which is what I use) is a few months old,
and lacks many of the features I would like.
That is true, you would have to use
Windows or Linux. Today.
When I asked about this,
I got brushed away, saying that the person who usually builds that
version is busy.
Excuse me, would you mind publishing the answer I gave you as well? I
wonder where I did brush you away. I think I was very polite and
explained that the Mac port was being worked on. And yes, it's true, our
Mac devs are very busy, which is why the team was expanded and a new
release is near completion.
We welcome everyone that needs an earlier port to step forward and help.
Sometimes it's just diffing some files and applying a few platform
specific changes. Still, it needs to be done with care and by someone
who knows what he's doing. We take quality very serious.