Robot brain development
9 views
Skip to first unread message
Peter Law
Oct 11, 2014, 4:19:26 PM10/11/14
to Student Robotics
Hi everyone,
We're now only a couple of weeks away from the Kickstart, much of the
preparation for which seems to be going quite well. I'm a little
concerned about an area where I know there were a number of plans, but
which we've heard almost nothing about. [1]
This is the new kit brain, both in terms of the hardware and the
software that it will be running.
I'm currently assuming that the plan is to keep up with our previous
goal (and recent success at) of handing out full, fully working kits
at Kickstart (this is rather imperative for the micro games). Based on
that and the current lack of information, I hope it's clear that my
concern here is well founded.
Thus, please could those involved (either directly responsible for or
actively working on) making these pieces let us know where we stand
here.
If things are happening in this area, and I'm sure they must be, then
that's great. I'm greatly appreciative of those beavering away in the
dark to make things work, and I don't want to belittle any of the
efforts, I'd just like to know where we stand and how others in the
community (myself included) can help.
The specific questions I have at this point are:
* What are the specs & details of the new kit to be issued?
* We need to know this in advance, both to update the docs and to
train ourselves in their use
* Who is working on making what?
* Are there any impacts on our other services?
* Aside from the docs, mentioned above, I think there was a plan to
change the way that code was delivered to the robots, which would
impact the IDE. Is this happening? Are there any others?
* What tasks remain to be done before the kits can be shipped?
* How can we help?
Without answers to these questions soon, I think we'd need to start
making contingency plans [3] if we want to have successful Kickstarts.
Thanks,
Peter
[1] The last post in this area (and only one I can find at all on
srobo-devel within the last few months) was Rob's one about ODROID
boot [2] stuff, and even that didn't actually present anything
concrete, only some ideas.
[2] https://groups.google.com/d/topic/srobo-devel/1ce1AuHjxfc/discussion
[3] Currently I'm assuming that we can re-use the SR2014 power boards
if it comes to it, though it would be useful to know for sure if this
is viable.
We're now only a couple of weeks away from the Kickstart, much of the
preparation for which seems to be going quite well. I'm a little
concerned about an area where I know there were a number of plans, but
which we've heard almost nothing about. [1]
This is the new kit brain, both in terms of the hardware and the
software that it will be running.
I'm currently assuming that the plan is to keep up with our previous
goal (and recent success at) of handing out full, fully working kits
at Kickstart (this is rather imperative for the micro games). Based on
that and the current lack of information, I hope it's clear that my
concern here is well founded.
Thus, please could those involved (either directly responsible for or
actively working on) making these pieces let us know where we stand
here.
If things are happening in this area, and I'm sure they must be, then
that's great. I'm greatly appreciative of those beavering away in the
dark to make things work, and I don't want to belittle any of the
efforts, I'd just like to know where we stand and how others in the
community (myself included) can help.
The specific questions I have at this point are:
* What are the specs & details of the new kit to be issued?
* We need to know this in advance, both to update the docs and to
train ourselves in their use
* Who is working on making what?
* Are there any impacts on our other services?
* Aside from the docs, mentioned above, I think there was a plan to
change the way that code was delivered to the robots, which would
impact the IDE. Is this happening? Are there any others?
* What tasks remain to be done before the kits can be shipped?
* How can we help?
Without answers to these questions soon, I think we'd need to start
making contingency plans [3] if we want to have successful Kickstarts.
Thanks,
Peter
[1] The last post in this area (and only one I can find at all on
srobo-devel within the last few months) was Rob's one about ODROID
boot [2] stuff, and even that didn't actually present anything
concrete, only some ideas.
[2] https://groups.google.com/d/topic/srobo-devel/1ce1AuHjxfc/discussion
[3] Currently I'm assuming that we can re-use the SR2014 power boards
if it comes to it, though it would be useful to know for sure if this
is viable.
Rob Spanton
Oct 12, 2014, 7:27:33 AM10/12/14
to srobo...@googlegroups.com
Hi Pete,
Yes, we've been a bit silent on this front -- faces to several
grindstones and whatnot.
> * What are the specs & details of the new kit to be issued?
Same as planned: new power board + tablet + ODROID-U3 + 2 wifi dongles +
servo board.
> * We need to know this in advance, both to update the docs and to
> train ourselves in their use
Yes, our priority right now is getting it ready to go in the box. We're
planning on getting it into the boxes on the 18th and 19th -- but more
on that later. Once it's in the boxes, we'll work on the docs.
> * Who is working on making what?
I'm spending most of this coming week on the image for the odroid.
Jeremy is working on the firmware for the power and servo boards.
> * Are there any impacts on our other services?
> * How can we help?
The contents of robot.zip is getting considerably smaller: it is
basically only going to contain the user's code, and a small number of
other files (run.py, overlay config, and an overlay). The python libs
for using the hardware will be installed via RPM in the appropriate
places on the system.
The impact on the IDE is that the pyenv submodule will likely change to
point to a different repo. There'll still be a make-zip script in
there. In order to provide for the wifi situation, this will need an
additional argument that's the TLA of the team the zip is being made
for. That's the only change to the IDE that'll be required.
> * Aside from the docs, mentioned above, I think there was a plan to
> change the way that code was delivered to the robots, which would
> impact the IDE. Is this happening? Are there any others?
At the moment, the priority is to get code shipping out in the same way
that it has previously -- i.e. by robot.zip. Post-shipping, we can deal
with the more exciting code delivery arrangement.
Cheers,
Rob
Yes, we've been a bit silent on this front -- faces to several
grindstones and whatnot.
> * What are the specs & details of the new kit to be issued?
servo board.
> * We need to know this in advance, both to update the docs and to
> train ourselves in their use
planning on getting it into the boxes on the 18th and 19th -- but more
on that later. Once it's in the boxes, we'll work on the docs.
> * Who is working on making what?
Jeremy is working on the firmware for the power and servo boards.
> * Are there any impacts on our other services?
The contents of robot.zip is getting considerably smaller: it is
basically only going to contain the user's code, and a small number of
other files (run.py, overlay config, and an overlay). The python libs
for using the hardware will be installed via RPM in the appropriate
places on the system.
The impact on the IDE is that the pyenv submodule will likely change to
point to a different repo. There'll still be a make-zip script in
there. In order to provide for the wifi situation, this will need an
additional argument that's the TLA of the team the zip is being made
for. That's the only change to the IDE that'll be required.
> * Aside from the docs, mentioned above, I think there was a plan to
> change the way that code was delivered to the robots, which would
> impact the IDE. Is this happening? Are there any others?
that it has previously -- i.e. by robot.zip. Post-shipping, we can deal
with the more exciting code delivery arrangement.
Cheers,
Rob
Murray Colpman
Oct 12, 2014, 7:32:26 AM10/12/14
to srobo...@googlegroups.com
On 12/10/14 12:27, Rob Spanton wrote:
> The contents of robot.zip is getting considerably smaller: it is
> basically only going to contain the user's code, and a small number of
> other files (run.py, overlay config, and an overlay). The python libs
> for using the hardware will be installed via RPM in the appropriate
> places on the system.
How will we ship a new pyenv version then? By shipping an overlay in the
> The contents of robot.zip is getting considerably smaller: it is
> basically only going to contain the user's code, and a small number of
> other files (run.py, overlay config, and an overlay). The python libs
> for using the hardware will be installed via RPM in the appropriate
> places on the system.
robot.zip, or am I missing something?
Murray.
Rob Spanton
Oct 12, 2014, 7:34:42 AM10/12/14
to srobo...@googlegroups.com
On Sun, 2014-10-12 at 12:32 +0100, Murray Colpman wrote:
> How will we ship a new pyenv version then? By shipping an overlay in
> the robot.zip, or am I missing something?
Yes, through the overlay.
> How will we ship a new pyenv version then? By shipping an overlay in
> the robot.zip, or am I missing something?
(And it won't be called pyenv...)
Cheers,
Rob
Peter Law
Oct 12, 2014, 11:41:14 AM10/12/14
to Student Robotics
Rob wrote:
> Hi Pete,
s/Pete/Peter/, please.
>> * We need to know this in advance, both to update the docs and to
>> train ourselves in their use
>
> Yes, our priority right now is getting it ready to go in the box. We're
> planning on getting it into the boxes on the 18th and 19th -- but more
> on that later. Once it's in the boxes, we'll work on the docs.
This feels like a rather unnecessary delay on making changes to the
docs. Surely anyone should be able to write at least outlines of the
docs now, leaving only minor fixes to be made during the last week?
> The contents of robot.zip is getting considerably smaller: it is
> basically only going to contain the user's code, and a small number of
> other files (run.py, overlay config, and an overlay). The python libs
> for using the hardware will be installed via RPM in the appropriate
> places on the system.
Shiny. Do we know what specific version of python this is (so we can
update the docs), and what packages will be included?
> The impact on the IDE is that the pyenv submodule will likely change to
> point to a different repo. There'll still be a make-zip script in
> there. In order to provide for the wifi situation, this will need an
> additional argument that's the TLA of the team the zip is being made
> for. That's the only change to the IDE that'll be required.
Cool. I've just added the TLA as an (optional) parameter to the
existing script & submitted that to gerrit [1] so that I could then
update the IDE [2]. I'm expecting the new script will require it, but
didn't want to break anything which currently only passes two
parameters.
Thanks,
Peter
[1] https://www.studentrobotics.org/gerrit/1738/
[2] https://www.studentrobotics.org/gerrit/1739/, which clearly blocks on 1738
> Hi Pete,
s/Pete/Peter/, please.
>> * We need to know this in advance, both to update the docs and to
>> train ourselves in their use
>
> Yes, our priority right now is getting it ready to go in the box. We're
> planning on getting it into the boxes on the 18th and 19th -- but more
> on that later. Once it's in the boxes, we'll work on the docs.
docs. Surely anyone should be able to write at least outlines of the
docs now, leaving only minor fixes to be made during the last week?
> The contents of robot.zip is getting considerably smaller: it is
> basically only going to contain the user's code, and a small number of
> other files (run.py, overlay config, and an overlay). The python libs
> for using the hardware will be installed via RPM in the appropriate
> places on the system.
update the docs), and what packages will be included?
> The impact on the IDE is that the pyenv submodule will likely change to
> point to a different repo. There'll still be a make-zip script in
> there. In order to provide for the wifi situation, this will need an
> additional argument that's the TLA of the team the zip is being made
> for. That's the only change to the IDE that'll be required.
existing script & submitted that to gerrit [1] so that I could then
update the IDE [2]. I'm expecting the new script will require it, but
didn't want to break anything which currently only passes two
parameters.
Thanks,
Peter
[1] https://www.studentrobotics.org/gerrit/1738/
[2] https://www.studentrobotics.org/gerrit/1739/, which clearly blocks on 1738
Reply all
Reply to author
Forward
0 new messages