[Blackpool] Saturday 15th August 2015

Arthur Garlick arthur_garlick at hotmail.com
Fri Aug 14 12:04:59 UTC 2015




Hi,Thanks Mike.  
I am helping at the Lets make it Digital BBC thing on the prom opposite the Tower which I think means I can't come (I don't have the start time yet but I assume it's before 10am).
Last Week
After Maker Fest last week I went round this guys: http://davidbuckley.net/ house, we talked solidly and without any pause for over 8 hours!  It was brilliant, I had a really good time, he had scores of working biped robots and I learnt loads about biped locomotion.  His approach to robot controlling software goes against everything else I've seen on robot forums and how I've learnt to put code together in my job.  It uses GoTo statements very heavily to model behaviour.  I have never been so sceptical about something going in to it, but I listened and watched him demo his approach and very surprisingly to me, he won me over.  
We've swapped a few very lengthy emails since and I've laid out how I'd do what he does without using GoTo commands but essentially I'll be copying his approach because it is far superior to the other code bases I've seen on the net and it's delivers what I was searching for at the back end of last year.  I always thought that the published code from the amateur robot guys on the net was not very good but I tried and couldn't come up with the better solution myself.  I have now got a better pattern of how to do it - I am quite excited about this.  The Kalman filters and ESP8266 can all play a part in the approach, and it will work in a model boat too, so a few things finally coming together.

GoTos Considered Harmful
It's worth a look at the debate about the GoTo statement in higher level languages if you aren't aware of it. Most developers will invoke the name Dijkstra who wrote a letter entitled 'Go To Statement Considered Harmful'. The phrase 'Considered Harmful' became a bit of a meme itself in technology which I've always thought amusing: https://en.wikipedia.org/wiki/Considered_harmful.  I am not religious about not using a GoTo, I use break and exit, I've never actually used a GoTo in production code though and it's more to do with peer pressure, I think they can be used sensibly. 

Maths - Kalman filters
I have started constructing the hardware to test the Kalman filter code, it is a single servo arm that can hold a gyro/accelerometer either: still, moving slowly or oscillating (sloshing) which is the model we are using to learn Kalman filters.  This single armed robot will be my first robot written using the new robot coding approach that I have learnt from David Buckley.

Halt and Catch FireI have been binge watching Halt and Catch Fire (on Amazon prime) whilst writing the Kalman robot code:  http://www.imdb.com/title/tt2543312/ Really enjoying it.  In the time when a 286 was the future.  It's a bit like Mad Men but with a PC clone tech start up.   The origin of HCF assembler command discussed here:  https://en.wikipedia.org/wiki/Halt_and_Catch_Fire  
We had the ABEND (abnormal end) error come up with Novell back in the day and we did the old joke about it being from the German Guten Abend, which I think means good evening or for us: nothing we can do now, we can all go home - good night...!
RegardsA  


> Date: Thu, 13 Aug 2015 19:09:36 +0100
> From: tonyhughes1212 at blueyonder.co.uk
> To: mikehull2008 at hotmail.co.uk
> CC: blackpool at mailman.lug.org.uk
> Subject: Re: [Blackpool] Saturday 15th August 2015
> 
> Sorry Guy's but other things on that s weekend so will not be with you. See you next week.
> 
> Tony
> 
> On 13 Aug 2015 18:38, Mike Hull <mikehull2008 at hotmail.co.uk> wrote:
> >
> > Tyldesley Road open from 10am Saturday 15th August. 
> >
> > See you all there. 
> >
> > Mike Hull 
> >      
> > _______________________________________________ 
> > Blackpool mailing list 
> > Blackpool at mailman.lug.org.uk 
> > https://mailman.lug.org.uk/mailman/listinfo/blackpool 
> _______________________________________________
> Blackpool mailing list
> Blackpool at mailman.lug.org.uk
> https://mailman.lug.org.uk/mailman/listinfo/blackpool

 		 	   		  


More information about the Blackpool mailing list