Closed GoogleCodeExporter closed 9 years ago
I looked at your branch and I saw a huge number of commits. I would ask you to
either collapse all of the changes with a tool or by hand.
I might actually suggest that you split this up into many smaller commits. It
is much easier for people to review smaller logical chunks.
Original comment by sandordo...@google.com
on 17 Jan 2012 at 5:07
I'll work on it after work tonight. I think that there would be two to four
logical groupings of commits. Most of the changes are in one file so the
review should not be to bad after I group up the commits. Will that be ok?
Original comment by chet.0x....@gmail.com
on 17 Jan 2012 at 5:41
That sounds fine. I might suggest submitting one or at most two commits at a
time. Please keep each logical change in it's own clone. If one change
depends on another don't send the former until the later has been approved.
Original comment by sandordo...@google.com
on 17 Jan 2012 at 5:55
Ok, I have grouped the code into three commits in a new clone:
code.google.com/r/chet0xhenry-antplus/
The first commit has almost all of the changes in the mytracks folder except
for the goldencheetah export class, however there are calls to that class so
the second commit is also required. I would have liked to split it up with the
ant+ code in one commit and the goldencheetah code in another. This however
proved to be too time consuming because the changes were altering the same
files. Most of the code however is ant+ and that should be the focus.
The last commit is the code need for the mytracks-lib. This should all be code
related to ant+.
Original comment by chet.0x....@gmail.com
on 18 Jan 2012 at 6:07
These commits are still huge and hard to review. Putting it all in one clone
makes it very hard to have address comments as I have to look several commits
later for your response.
Original comment by sandordo...@google.com
on 20 Jan 2012 at 5:21
You aslo did not allow comments on your clone. Please do this again but only
put one small initial change in that clone. We can work from there to the
final change. Also avoid pointless format changes like:
http://code.google.com/r/chet0xhenry-antplus/source/diff?format=side&path=/MyTra
cks/res/layout/sensor_state.xml&r=4e0a17ac73dd4b770cfaa6243b48413067b49f18&spec=
svn4e0a17ac73dd4b770cfaa6243b48413067b49f18
Original comment by sandordo...@google.com
on 20 Jan 2012 at 5:25
I realize that this is a large change however the changes are for one feature,
the ability to read ant+ standard power messages and use the information to
calculate power, cadence, and speed (bSpeed).
Unfortunately, this makes almost everything dependent. The above format
changes was not intended to be useless. The bSpeed is the measured speed from
ant+ standard power message it uses the event period and the wheel rpm to give
us an accurate measurement of speed and distance. This speed is also more
reliable than gps. Are you saying that if we have an ant+ power sensor we
should replace gps speed with the ant measurement? I personally would like to
record both and maybe only display one. Maybe we could use the bSpeed
measurement to improve the accuracy of the gps, along with acceleration, and
magnetic field?
The one thing that could be broken out is the golden cheetah export, but that
may take a few hours.
There is one more todo I would like to take on this weekend, currently the
bSpeed is not localized and only reports in miles per hour.
I update the clone to allow non-members to review code, I'm sorry I missed that
step.
Original comment by chet.0x....@gmail.com
on 20 Jan 2012 at 5:52
Not everything in your change is dependent. There are fairly clear lines to
separate the changes. I do like the changes that you are proposing but we have
to do this in a controlled manner. Please create a clone with 1 commit and
only one commit. In that commit put one logical change. Perferably less that 5
changed or added classes.
I would suggest that you do the GoldenCheetah export first as another
contributor is radically reworking the ANT+ sensor work. I like how he has
changed the classes and will likely take his contribution. He did not however
implement power or speed sensors which I would like to take from you. However
since his change is not complete I would encourage you to work on the export
code as that should be compartmentalized. Then once the new ANT+ changes are
in you can integrate with that framework to add power and speed.
You might want to look at:
http://code.google.com/r/ezerotven-mytracks2/source/list
It shows where I would like to take the ant sensor code. I also shows good
development style as each commit is small and atomic.
Original comment by sandordo...@google.com
on 24 Jan 2012 at 6:02
Jimmy these changes LGTM. I think it is ok to pull it into head and maybe even
the latest release.
Original comment by sandordo...@google.com
on 31 Jan 2012 at 6:14
Jimmy these changes LGTM. I think it is ok to pull it into head and maybe even
the latest release.
Original comment by sandordo...@google.com
on 31 Jan 2012 at 6:14
Oops. Jimmy ignore my previous message. Chet the other ant changes will be
committed soon. Please look at how you can add power in that framework.
Original comment by sandordo...@google.com
on 31 Jan 2012 at 6:27
Direct connection to ANT+ power sensors and speed from odometry are important
additions that I'm keen to use. The recent additions that allow direct
connection to ANT cadence sensors are not enough. I'm wondering how this is
going.
Original comment by kenep...@gmail.com
on 13 Mar 2012 at 1:58
Also can you please include the distance as calculated by the speed sensor in
the exported TCX files as described at
http://developer.garmin.com/schemas/tcx/v2/xmlspy/index.html#element_Trackpoint_
Link05E0A1B0 and in the exported CSV files.
Original comment by kenep...@gmail.com
on 13 Mar 2012 at 3:16
Currently, we support Ant+ heart rate monitors and Ant+ speed distance
monitors. Please file new feature request for add other Ant+ profiles. Going to
close this bug.
Original comment by jshih@google.com
on 5 Jun 2012 at 12:01
A new code review request with this feature was created:
http://code.google.com/p/mytracks/issues/detail?id=1023
Original comment by chet.0x....@gmail.com
on 9 Jul 2012 at 2:44
Original issue reported on code.google.com by
chet.0x....@gmail.com
on 17 Jan 2012 at 5:00