Thanks a lot for your analysis.
I had missed that the company was bought by Garmin. That fact
dwarfs any argument about revenues originating from SDK licenses,
as a reason to restrict protocol specifications.
 At a first glance, both SDK and protocol
specifications seem to be concerned with the (binary) file format
and ways to encode/decode relevant data. Lower levels, as shown
in the image beside, are not discussed. I don't think Garmin fear
that a competitor would gain an unfair advantage if they openly
published that part of the FIT protocol and the related SDK source
code. Rather, an increased availability of compatible software
packages could foster device usability. Assuming that Garmin's
core business is selling devices --not software licenses-- why do
they keep such a restrictive license? I hope not just to save
lawyers' fees...
Sic venit IoT
Ale
On Sun 22/Jan/2017 23:01:05 +0100 sandy wrote: Right, the website shows quite a few well-known
companies using their technology. Possibly we have the timetable
backwards. Perhaps, "We needed something better so we bought that
company because we thought it had a better solution." That gives
them a better solution and the ability to push changes as needed
to _keep_ it better in the future. The only downside is to users
outside their company, who can't predict changes.
-Sandy
----- Original Message -----
To:
<sandy@xxxxxxxxxx>
Cc:
"ietf@xxxxxxxx Disgust" <ietf@xxxxxxxx>
Sent:
Sun, 22 Jan 2017 14:21:07 -0500
Subject:
Re: What is Ant's Fit Protocol?
On Sun, Jan 22, 2017 at 12:27 PM, <sandy@xxxxxxxxxx>
wrote:
> The URL you provide says:
>
> "ANT/ANT+ are managed by ANT Wireless, a division of
Dynastream Innovations
> Inc.......Dynastream was established in 1998 and became a
wholly owned
> subsidiary of Garmin Ltd. in December 2006. "
>
> So, there is the most likely reason for using a new
proprietary solution.
> They OWN the company. Clearly, the company's software is
superior to any
> solution that was 'Not Invented Here'.
Sorry, but no -- that is an entirely reasonable supposition for
Garmin, but ANT/ANT+ is used by a significant number of other
companies as well -- for example, I know them from Peloton
exercise
bikes, Samsung phones, etc.
A quick glance at their "directory" shows what looks like a few
hundred brands, and >600 products.
From a *very* quick glance it seems that, for simple things, it
is
simpler to integrate than programming against bluetooth directly
(e.g
the Nordic RNF5283 is ~$2.00, and does BLE, ANT and NFC) --
there are
a number of friendly looking libraries and demos for things like
talking to bikes, etc. ANT *feels* like it provides a higher
layer /
abstraction to program against -- for example, this datasheet
https://www.thisisant.com/resources/fit2-fitness-module-datasheet/...
'tis been many years since I tried to integrate a bluetooth
module
into something, but when I did, there seemed to be so much rope
that
getting started was tricky...
I suspect that the root answer to Alessandro question is a
combination
of 1: NIH, 2: because other devices I need to talk to already do
this,
and 3: better abstractions / nicely defined and documented
stack.
[ Note: This is just from a brief skim of their website - I may
be
completely wrong, BT / BLE may have progressed and libraries
improved,
and better, clearer profiles created, etc... ]
W
>
> -Sandy
>
>
> ----- Original Message -----
> From:
> "Alessandro Vesely" <vesely@xxxxxxx>
>
> To:
> <ietf@xxxxxxxx>
> Cc:
>
> Sent:
> Sat, 21 Jan 2017 21:10:57 +0100
> Subject:
> What is Ant's Fit Protocol?
>
>
>
> Hi,
> I annoyingly see this stuff in some GPS devices. It seems
to be a
> proprietary
> protocol for wireless communication, also used as a file
format. It is
> resumed
> here:
>
> https://www.thisisant.com/company/
>
> I'm wondering why companies use that protocol instead of an
open standard.
> Is
> it because there is no suitable open standard or just
because they hate open
> standards?
>
> I would try and dissuade open source packages, e.g.
gpsbabel, to try to
> support
> it, since that protocol seems to be going to change
unpredictably and hence
> their software will never work. Opinions?
>
> TIA for any reply
> Ale
>
--
I don't think the execution is relevant when it was obviously a
bad
idea in the first place.
This is like putting rabid weasels in your pants, and later
expressing
regret at having chosen those particular rabid weasels and that
pair
of pants.
---maf
|