>>Was Digiboard approached about advertising in your magazine? Are they
>>advertising in it? Is Cyclades advertising in it?
>Your implication here is the Mike is slamming Digiboard for not
>advertising in LJ. You've used Digiboard kit (and Cyclades for that
>matter) - surely it's obvious that he's telling it like it is...?!
We've had problems with Cyclades during the teething stage of driver
development. Problems still crop up as the driver undergoes further
development, just like anything else. I can give a bad benchmark by using
the driver of my choice, if I wanted to. It's not obvious he's telling it
"like it is" at all.
We're getting good results with Digiboard PC/8e boards using the 1.3 driver.
The Cyclades 8Y is ok, too, but sometimes you need full modem control and
so the Digiboard works for those installations. There are lots of Digiboards
out there already, that were just waiting for Linux support!
Quote:>>The Linux Journal's performing a valuable service by being published. But
>>don't ruin your credibility by acting like so many second rate computer
>>rags which link the way they treat products editorially to the amount spent in
>>the advertising department.
>Hey, just stop and think about it! People into Linux are into it
>because it interests them, not because of advertising royalty!
Nonsense. The Linux Journal is a commercial publication and they depend upon
advertising and sales for their livelihood. I'm sure some of them are also
interested in what they do for a living. That's great.
Quote:>>You make this sound as though Digiboard is responsible for the driver.
>Lack of supported driver - certainly.
This is really, really wrong.
The person who wrote the driver is a Digiboard employee. That is more direct
support than Cyclades offered. Troy (the author) has solicited assistance
from the Linux developer community in order to improve it and to give
feedback so he can fix any problems. So far he's put out 3 revisions.
Another Digi employee, Bob Lipes, has worked on it and offered patches.
Quote:>>You keep plugging Cyclades; they *have* bought advertising in your magazine,
>They have - although they may not have paid to have the driver
>developed, they do send developers excessive amounts of technical
>information FREE OF CHARGE!
They are glad to distribute and advertise the driver in order to sell their
boards, as long as Randy volunteers his time and efforts for free. But they
won't support it with their own employees. They are not responsible for it,
and all fixes are up to Randy. If he or some other volunteer doesn't do it
for free, it won't be done and they won't support it.
Take a look at the Cyclades Cyclom 8Y. There are two Cirrus serial chips on
it. There is some glue logic, and a couple of PALs. No custom BIOS or
downloadable RAM. 12 byte FIFO. The design is right out of the Cirrus
technical manual. Nice product, using high performance VLSI off the shelf
chips. But a minimal amount of development in-house. The "excessive amount"
of technical information is available to anyone who requests the chip manual
from Cirrus. That from Cyclades is limited. I understand that Randy even
went to Cyclades and gave them seminars on the board and how to exploit its
features, greatly to his credit. The other drivers Cyclades pays to have
developed owe a lot to his free Linux efforts, I am given to understand.
Now, look at the Digiboard PX/8e. Intel CPU, EPROMS with custom
programs, static RAM for downloadable software FEP which is upgradable,
dual-ported RAM seen by both board and motherboard, allowing memory access
to large buffers, no interrupts necessary. Lots of custom inhouse software
development, upgrades available, driver can load latest FEP code right into
board. And Digi's made newer FEP's that they develop available, even in this
"unsupported" Linux driver. That's pretty direct support for an unsupported
product. They even make the code available on their own ftp.digibd.com.
Quote:>>Here's the upshot folks: to be treated more than fairly in the Linux
>>Journal, purchase advertising. It'll at least keep your product from being
This is *so* common in the publishing world that it is an anomaly if it
doesn't happen. One has to have very strong and enforced policies in place
ahead of time for it not to happen. A number of years ago, BYTE even had to
run a whole article on why they could not do this even when everybody else
in the industry was doing it. The pressure on a struggling small publication
is almost overwhelming. Who doesn't want to give a boost to someone else in
a small company who's helping you survive when they "deserve" it?
This isn't just about a review. It's about the obvious ill-feeling and
disparaging remarks made about Digiboard that were expressed by the reviewer
here on the net, who was following up a post about Digiboard and instead of
discussing that had to keep talking about how wonderful Cyclades is and how
you should not use Digiboard. And we don't even get to see this review
because it's proprietary and details can't be released except commercially.
This is very unprofessional to say the least.
Quote:>Or alternatively, join the spirit of the human race (or at the very
>least the Linux community) and share!
It's always easy to tell others to "share" when your own livelihood's not at
I guess you think Digi should tell all its programmers that they are now
going to work without pay in the spirit of "sharing". Some of those
programmers have given generously of their own time to the Linux effort.
How about posting all the printed info in the Linux Journal online, and
maybe even revealing exactly how the magazine is published so everyone else
can "share" in the success by duplicating its efforts without having to go
through the difficulties you had to in launching a new venture? That would
not be fair to your families or employees.
It seems to me that it's not the place of the LSJ (if it really is saying
this - it seems quite a number of people are claiming to speak for them) to
accuse others of being selfish and lay guilt trips on when LSJ's own profits
are made indirectly possible by all that unpaid programming labor.
There are ump*-zillion serial board manufacturers out there. I can tell
you plenty of stories about the large majority who simply told us to get
lost when it came to providing any help at all on writing a Linux driver.
Not even info under a NDA was available from these folks. That's the
Now, we have enlightened board manufacturers that you can count on the
fingers of one hand offering help for free in an untried and unproven market
that's a tiny, tiny slice of the DOS/Windows monopoly. And one of them
happens to be Digiboard. Digiboard deserves praise, not insults.
Please. If you want better drivers for Digiboard, then test Troy's code,
offer suggestions for improvement, hack on it yourself. At least *two*
Digiboard engineers have publicly asked for assistance in supporting the
driver in this way.