DIYEFI.org Forum
http://forum.diyefi.org/

Fred's firmware development diary
http://forum.diyefi.org/viewtopic.php?f=8&t=57
Page 97 of 99

Author:  Fred [ Thu Jun 11, 2015 11:09 am ]
Post subject:  Re: Fred's firmware development diary

Code has been getting written, and it's been getting tested, too. Good progress. A few more small issues to solve for various people, and I'm going to get stuck into the two big issues that have been holding back the project for years despite various amateurs trying to solve them. It's time to bring in some pros and knock this shit into shape, rapidly.

We also might have a cross-platform tuner that doesn't crash, lie to you, corrupt your data, etc, soon. That'd be a nice change! :-D

A brief history of FreeEMS tuners:

  • Python tuner - By Aaron barnes, the first, used to tune the first FreeEMS car in Florida, too. Rudimentary functionality, but got the job done. Made the mistake of interacting with Aaron's wife. She banned him from talking to me. Project stalled, firmware moved on to more advanced places.
  • MegaTunix - By David Andruczyk, abandon-ware now, but dropped by FreeEMS some time ago due to the author being impossible to work with despite his good heart that is usually in the right place. Often crashed, FreeEMS support only ever thinly hacked in, strong M$ flavours present, best gauge cluster setup, by far, partly courtesy of MrOnion, Dave was reluctant to make changes to move it forward toward a better place, though. And wordy about it, too.
  • EMStudio - By Michael Carpenter, showed promise, smashed together to a basic working form in just a month. Never stable functionality wise, somewhat crashy, but not as bad as MTX, stopped working on it several times, sporadic updates, half finished changes, buggy result, inconsistent behaviour, wrong cell tracing, wrong gauge readings, wrong data export, etc. Frustrating, as it's the closest in terms of how it works.
  • New tools in dev right here in NZ courtesy of mother Russia. More to be revealed once they are suitable for general use. Cross platform. Crash-free, correctly architected? Time will tell. I'll have a close hand in guiding this one, should be good.
  • New U* tools on the horizon. Vapourware for now. Needs to happen, though, and soon. Many things depend on this. Many very very good things. Differentiators from all of the competition, including the pro gear.

Onward and upward.

Fred.

Author:  Fred [ Sun Jun 21, 2015 6:10 am ]
Post subject:  Re: Fred's firmware development diary

New code has been tested and is working as designed. More tweaks to do before it can be rolled out, but this pretty much clears the way for the serious work. 2 big hitters that will KO the opposition with ease. Never fear, Fred is here. Back on the job in earnest. Mu ha ha ha.

Author:  Fred [ Sat Jun 27, 2015 12:15 pm ]
Post subject:  Re: Fred's firmware development diary

A new firmware hacker started on the job today! :-) He's got two tasks lined up, one pretty easy (fixing some crap code from an ex-contributor), and one pretty important (replacing some crap code from an ex-contributor). I'll help with both as I've done in the past with other contributors. Today was the start of that, walking through the code and documentation for the easy task and suggesting some reading material to assist in coming up to speed quickly.

The other day I tested some new tools by a long time contributor to open sauce EFI tooling and was impressed with what I found. I gave some detailed feedback and am super keen to see some new versions with some fixes/changes applied. Already very exciting stuff, though. I'll keep that mostly to myself for the time being, however once a few key things are changed I'll be pleased to be the first poster in the release thread with one of these:

Image



Between these things and the tuner I mentioned last time, this almost frees me up to work on the other really really important project. Just a little ground work to do first and then I'll get properly stuck in whilst trying to keep people happy by rolling out easy feature and structure stuff as often as necessary. Things are pretty positive and active right now, bloody good to see. Two new vehicles today, neither of which posted their thread to the vehicles section, tsk tsk. Exciting, though! :-)

Fred.

Author:  Fred [ Fri Jul 10, 2015 9:09 am ]
Post subject:  Re: Fred's firmware development diary

Fred wrote:
A new firmware hacker started on the job today! :-) He's got two tasks lined up, one pretty easy (fixing some crap code from an ex-contributor), and one pretty important (replacing some crap code from an ex-contributor).

This fresh blood is doing bloody well and has come up with a couple of nice fixes and ideas already. Lovin' it. :-) Good on ya, mate! :-)

Author:  Fred [ Sat Aug 08, 2015 1:08 pm ]
Post subject:  Re: Fred's firmware development diary

The results of some new code, working RB support in FreeEMS! :-D

viewtopic.php?f=55&t=2610

So epic.

Author:  Fred [ Sun Aug 23, 2015 10:51 am ]
Post subject:  Re: Fred's firmware development diary

PWM open loop idle control support was finalised tonight with a clue from Aitor! Thanks mate! :-D Can't believe I missed this for so long. Too obvious/subtle. Will roll it out to users soon. Onward and upward! :-D

Author:  Fred [ Thu Nov 26, 2015 9:59 am ]
Post subject:  Re: Fred's firmware development diary

Someone commented that there have been no pushes to github since sometime last year. This is true. What's not true is that there has been no work done. Just wanted to make that perfectly clear. If you are a bonafide owner of FreeEMS hardware, much newer/better code will available to you in some form, depending on your status. That's just how it has to be for a little while. Hopefully before COB 2016 we can see the whole shooting match back in full public eye, once again. Until then, lots of work is going on with firmware commits from at least 4 people in recent times. Some of this is great stuff, and I'm really looking forward to sharing it all again. Please be patient with us during this awkward time. Cheers.

Author:  Fred [ Sat Apr 30, 2016 4:59 pm ]
Post subject:  Re: Fred's firmware development diary

Well, here I am at 3am/4am posting in the diary thread, just like old times! :-p Except I was not up all night coding, instead I was up part of the night suffering and had to remedy a few things before I could contemplate sleeping. So I figured I'd sit here and type up what I got done this morning and late this evening pre midnight before I went back to bed...

Clean up!

  • I went through my repo nuking old junk branches and migrating good branches of various types into the future.
  • I fixed a badly fucked up branch of one of my favourite users, too.
  • I put some working but crappy code in the hands of a good man to see if he can think of ways to make it more epic and less average.
  • The layer beneath that I need to finish by rolling it out to all appropriate modules in the firmware. Only some implemented right now. Cleared a path for that to happen, too.

I'm going to finish doing these chores while I have breakfast in a few hours which will put me in a good place to make some progress soon. Yes, you've heard it before. Sorry. But:

One of my FreeEMS cars came out of hibernation the other day, and another one went in for a while. When the other one that went in comes out again, it'll be a slightly different animal. How different, not sure yet, but different enough! :-D

IE, I'll be using this crap again soon. And that'll mean progress. Speaking of which:

1 decoder hugely upgraded
1 decoder getting fine tuned using FreeEMS' fully configurable logging features and ULV's fully dynamic log reading
1 cold start algorithm being developed by a good man
1 epicly good feature (as mentioned above) well tested on 4+ cars and ready to roll out with a little more integration work
1 auto-calibrate tool being developed by another good man
1 tuner being developed by me, slowly, very slowly, but properly
1 air cooled VW engine approaching readiness for running and stealing 40th place from me and my blue 740 sedan

Good times! :-)

Fred.

Author:  Fred [ Sun May 01, 2016 6:11 am ]
Post subject:  Re: Fred's firmware development diary

All branches nuked/archived/rebased and ready for action. 56 current branches all up, each with something valuable on it. only maybe 6 to 10 or so of those are config branches for users. The rest are genuine features/improvements/fixes/etc. Some of them pretty epic. Speaking of which, this showed up in my garage today: https://twitter.com/FredCookeNZ/status/ ... 1981456384

I may have given away the purpose with too strong a clue, but... :-) I'll have to clear my desk and get it setup soon, just for giggles. It's big enough to use, too, so it'll likely end up on the gold Volvo with boost.

Author:  Fred [ Thu Jul 07, 2016 3:45 am ]
Post subject:  Re: Fred's firmware development diary

Some pair programming on the firmware occurred last night. 99% of the credit goes to the man at the keyboard (not me!). I was just offering humour and the occasional pedantic detail to what was already exquisitely crafted to a very high standard. We got a decent amount done, and today I put some more time into the research to back further progression of this top secret new feature. Why so vague? So as not to give a free lunch to those with no imagination and no ideas. Sorry about that. This work is on the direct path toward ending the silent game by squashing the problem(s) convincingly.

In recent times there have been at least 3 or 4 people actively using the system and pushing on me for improvements and so forth. It's been really good with quite a few good changes falling out of those needs. I need to rejoin that list and push on myself, too...

Some solid work behind the scenes has been done in all of the core areas over the last couple of months. Really encouraging. I'm absolutely looking forward to some of this stuff coming together in a way that nothing else does. Onward and upward!

Page 97 of 99 All times are UTC [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/