Fred's firmware development diary comments thread

Official FreeEMS vanilla firmware development, the heart and soul of the system!
User avatar
SleepyKeys
LQFP144 - On Top Of The Game
Posts: 549
Joined: Mon Feb 11, 2008 10:52 pm
Location: Arizona
Contact:

Re: Fred's firmware development diary comments thread

Post by SleepyKeys »

Even with OEM ECUs, I have noticed that if you come off the clutch way too fast you can hear the engine knock a few times, likely to the decel effect you described.
You snooze, you lose!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Fred's firmware development diary comments thread

Post by Fred »

I never noticed that in all my years of clutch dropping, but it doesn't surprise me :-)
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
SleepyKeys
LQFP144 - On Top Of The Game
Posts: 549
Joined: Mon Feb 11, 2008 10:52 pm
Location: Arizona
Contact:

Re: Fred's firmware development diary comments thread

Post by SleepyKeys »

Let me rephrase, droping the clutch enough to come close to stalling the engine, but not hard enough to burn rubber.
You snooze, you lose!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Fred's firmware development diary comments thread

Post by Fred »

Ahh, OK, so low RPM rattle, that's fairly standard clutch dropping or otherwise :-) My clutch drops were always in the 5 - 8k region ;-)
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
SleepyKeys
LQFP144 - On Top Of The Game
Posts: 549
Joined: Mon Feb 11, 2008 10:52 pm
Location: Arizona
Contact:

Re: Fred's firmware development diary comments thread

Post by SleepyKeys »

Fred wrote:Sean, it's just you now, all other variants are tested OK. Then I'll consider that code adequately tested for general consumption and migrate it to the master branch. There's currently far too big of a gap between them... that's got to change! Ball is your court, my friend!
I feel a need to "save face" a bit here. Don't get me wrong, I WANT to get that thing back up and running, but people need to know this action is not as simple as walking out to my car and taking care of business.

1. The car is buried behind this behemoth trailer. Pulling it out and putting it back in without destroying the house takes two people and a fair amount of time.

2. One of our friends passed away and some of my garage space is taken up by his stuff(there is not enough room for my car). I suspect that by the end of July this will no longer be the case :)

Image
Image

Guys, sorry for the delay in advance, still though, feel free to razz me about it :)

-sean
You snooze, you lose!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Fred's firmware development diary comments thread

Post by Fred »

I *would* accept any excuse you handed to me, but:

1) The car ran for 30 minutes while I was there, then sat for some time without running at all, and by your own admission has run very few times at all since I helped you get it running in the first place.
2) The clutch was found to be faulty more than two years ago and no time was found to do what is, in any front-engine rear-drive vehicle that I've ever seen or touched, between an hour and a day, max.
3) The ECU wasn't maintained in a usable state ready for testing. In contrast, it was canibalised for the purposes of building another ECU for the weekend toy, thus disabling the car.
4) Due to the derelict state of the vehicle, you chose to bury it out of sight and mind behind the camper that's difficult to park there.
5) At least several of the bushings in the suspension had seen better days two years ago, too. I strongly doubt these have been replaced, which is, again, a quickish job, worst case.

The truth is, for a number of reasons, some totally legitimate, and others totally bogus and psychological, you didn't *want* to use the car, and thus felt unmotivated to bring it back to, and maintain it in, usable safe condition. Ordinarily this would be OK, as others would be able to fill the void, however you're the sole user of this code, the primary author of this code, and I will not fucking accept dead, stale, untested code laying around in the code base.

We don't need to discuss all of the reasons for not wanting to drive it, but among the bogus ones were:

1) Dislike of non-sequential running.
2) Dislike of temporary DIY hardware.
3) Desire to prove self/xg bb code.
4) Desire to not spend on fuel.

What the end result of these, and others, boils down to is the following:

1) Lack of desire to repair broken aspects of the car.
2) Lack of distaste for rendering the car unusable (even for idle/rev in the fucking driveway).
3) Lack of desire to finish off the decoder bringing it up to a world-class standard.

As a final end result of these, and the earlier list, the car is not available for testing, and all of a sudden it's too hot to work on it, and it's stuck. Reasons for it being stuck are not entirely covered above. So we'll cover them here:

1) Alternate vehicle runs mostly on free fuel, which unsurprisingly, as indicated by the name, doesn't cost you anything. JewEFI
2) Weekend toy ECU needed some FETs, at about 1usd each, you felt the need to save money in the build, and pulled them out of the other board. JewEFI
3) Normal clutch nor perceived to hold V8 turbo power, good clutch not perceived to be cheap enough. JewEFI
4) Star of David too bright while wearing orthodox amplifying glasses, blinded to fact that semi-sequential is good enough to verify firmware builds. JewEFI
5) Star of David too bright while wearing orthodox amplifying glasses, blinded to fact that temporary DIY hardware is good enough to verify firmware builds. JewEFI

Basically this comes down to two things, one of which is just what it is, and the other I find really distasteful and not in the spirit of the project.

1) Jewish tin-arse tight-arse clam-wallet squeaky-arse clenchfistedness :-) Sorry, ran out of adjectives. You might have a few, that friend of yours was WAY worse by your accounts, re flat tyres and driving on them to the gas station to fill them each morning causing lateness. This is just what it is.
2) Inability to put aside your prejudices and dislikes and personal satisfaction/gratification for the benefit of the project as a whole. This pisses me off a huge amount, even if I *am* used to it by now. You're not alone, though, Ben sequential/COP Fenner is right there with you, as are others. The difference is, you should know better.

I'm only giving you a "razz" now because you tried to make an excuse. An excuse for the last 3 months doesn't cover the previous 23, sadly.

Once this summer is gone, here's what I expect from you, and to the maximum extent possible, demand from you:

1) Fix the physical aspects of the car, the clutch, the bushings, anything else that corroded to dust in the last 2+ years, possibly even upgrading the 7" paper diff to a decent strong one.
2) Permanently put an ECU in the fucking thing, even if the wires continue to come in through the window, and even if it continues to be a TA card + shitty difficult-to-solder breadboard, and DON'T FUCK WITH IT EVER AGAIN except to, without delays, replace it with something else, including testing. If attempting to replace, and striking issues, you should be able to back out the change and roll back to the previous hw with ZERO DOWN TIME.
3) Keep some fuel in it, and run it, even if start, idle, warmup, rev up, idle, kill, every 2 weeks, or 10 commits, or when asked, which ever is longer of the first two, or without more than 48 hours delay on the second two.

Without the above, you can't reasonably hope to continue having your name in the number two position of the team page. If you want that spot, you have to act like the second biggest ambassador to the product, and that means running it where ever you reasonably can.

I'm in pain from writing this, and although you meant the excuses fairly lightly, and that's obvious, to me, it hits a nerve, because it's pretty much bullshit due to those other 23 months and two full winters. Thus I couldn't not do this.

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
SleepyKeys
LQFP144 - On Top Of The Game
Posts: 549
Joined: Mon Feb 11, 2008 10:52 pm
Location: Arizona
Contact:

Re: Fred's firmware development diary comments thread

Post by SleepyKeys »

1)Dislike of non-sequential running.
2) Dislike of temporary DIY hardware.
3) Desire to prove self/xg bb code.
4) Desire to not spend on fuel.
#1, true but that's besides the fact. I'm 100% OK with running semi-seq and when testing is needed, I can swap it back to fully sequential later. In other words #1 as an excuse is purely your bogus perception.
2) Lack of distaste for rendering the car unusable (even for idle/rev in the fucking driveway).
I really don't like it sitting. The rail had a higher priority. I cant do everything I want to do.
3) Lack of desire to finish off the decoder bringing it up to a world-class standard.
I suppose it could use some improvements, there are a couple things I don't like about it albeit it works pretty well as-is.
1) Alternate vehicle runs mostly on free fuel, which unsurprisingly, as indicated by the name, doesn't cost you anything. JewEFI
2) Weekend toy ECU needed some FETs, at about 1usd each, you felt the need to save money in the build, and pulled them out of the other board. JewEFI
3) Normal clutch nor perceived to hold V8 turbo power, good clutch not perceived to be cheap enough. JewEFI
4) Star of David too bright while wearing orthodox amplifying glasses, blinded to fact that semi-sequential is good enough to verify firmware builds. JewEFI
5) Star of David too bright while wearing orthodox amplifying glasses, blinded to fact that temporary DIY hardware is good enough to verify firmware builds. JewEFI
It's hard to get mad at you when you have a sense of humor like that. #4, thats just a flat out lie/misconception, never thought or said it was not. Not true, but that DIY board was junk. You laughed for 5 minutes when you saw it in person for the first time. You hurt my feeling :( j/k lol
2) Inability to put aside your prejudices and dislikes and personal satisfaction/gratification for the benefit of the project as a whole. This pisses me off a huge amount, even if I *am* used to it by now. You're not alone, though, Ben sequential/COP Fenner is right there with you, as are others. The difference is, you should know better.
Yes, a big motivator to get the XGATE code to a usable state was to SHUT DOWN THE MOUTHS of guys like ben, NOT TO INSTULT YOU(that's how you seem to perceive it). I got SICK of hearing "hahaha you only have 6 usable channels...." On one hand they are correct its not 1985 and we should have more, on the other in many cases you can do just fine with fewer channels.

That's not true at all, I have told you on MANY occasions how I'm willing to get my stuff to a state that allows fast vanilla code testing. I think comparing me to Ben in that respect is just flat out wrong(even though I respect most of what he does). If I can get my personal satisfaction/gratification and help test vanilla code then we BOTH win. In a sense that's my paycheck.
Once this summer is gone, here's what I expect from you, and to the maximum extent possible, demand from you:
You and my wife both!
Without the above, you can't reasonably hope to continue having your name in the number two position of the team page. If you want that spot, you have to act like the second biggest ambassador to the product, and that means running it where ever you reasonably can.
Motivating words. Sorry to hear about your pain.

because it's pretty much bullshit due to those other 23 months and two full winters
You have to remember the rail was in the garage the last 20 months and it's illegal to work on cars in my driveway. As must as I distaste that I would rather not deal with the HOA/city.
You snooze, you lose!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Fred's firmware development diary comments thread

Post by Fred »

2) Lack of distaste for rendering the car unusable (even for idle/rev in the fucking driveway).
I really don't like it sitting. The rail had a higher priority. I cant do everything I want to do.
Once again, driveway revving is enough. If you simply didn't sabotage the ECU, then it would be testable.
I suppose it could use some improvements, there are a couple things I don't like about it albeit it works pretty well as-is.
True enough, but you seem to fail to comprehend that it's developed against an API that is changing and that it will break and fall into disrepair if not tested with each API change. You're effectively either forcing me to either:

1) Rip that decoder the fuck out completely
2) Allow it to be broken in the source tree
3) Stagnate the API and not move the fw forward.

1 is fucked up because it's perfectly usable, and COULD be maintained.
2 is fucked up because someone might try, and get PREVENTABLE issues.
3 is fucked up because obviously the entire project STAGNATES without progress.
1) Alternate vehicle runs mostly on free fuel, which unsurprisingly, as indicated by the name, doesn't cost you anything. JewEFI
2) Weekend toy ECU needed some FETs, at about 1usd each, you felt the need to save money in the build, and pulled them out of the other board. JewEFI
3) Normal clutch nor perceived to hold V8 turbo power, good clutch not perceived to be cheap enough. JewEFI
5) Star of David too bright while wearing orthodox amplifying glasses, blinded to fact that temporary DIY hardware is good enough to verify firmware builds. JewEFI
It's hard to get mad at you when you have a sense of humor like that.
Good! :-p
1)Dislike of non-sequential running.
#1, true but that's besides the fact. I'm 100% OK with running semi-seq and when testing is needed, I can swap it back to fully sequential later. In other words #1 as an excuse is purely your bogus perception.
4) Star of David too bright while wearing orthodox amplifying glasses, blinded to fact that semi-sequential is good enough to verify firmware builds. JewEFI
#4, thats just a flat out lie/misconception, never thought or said it was not. Not true, but that DIY board was junk. You laughed for 5 minutes when you saw it in person for the first time. You hurt my feeling :( j/k lol
Actions speak louder than words, man. Did you run it on semi even once since you got it on seq? If so, it's so long ago that I don't remember. You've contributed VERY LITTLE testing of the vanilla fw when I desperately need that testing to make QUALITY progress. I could make progress without testing, sure, but then this project would be no better than James' various fucking dogs breakfasts. At that point, it ceases to be relevant. Quality is our middle name. Without it, we're nothing.
Yes, a big motivator to get the XGATE code to a usable state was to SHUT DOWN THE MOUTHS of guys like ben, NOT TO INSULT YOU(that's how you seem to perceive it). I got SICK of hearing "hahaha you only have 6 usable channels...." On one hand they are correct its not 1985 and we should have more, on the other in many cases you can do just fine with fewer channels.
I'm not insulted at all, it's just counter productive. On the one hand, the XGATE code may one day be used as the core of the system. On the other hand other things have suffered in your pursuit to solve that problem in that way. Simply growing a thicker skin and laughing off the ignorant is a superior solution. Especially in light of the fact that having "lots of channels" attracts the WRONG type of people who are just a BURDEN to the project at this stage. Assuming your core goal is a reliable and supported 36 channel ECU ASAP, then do what gets you there fastest:

1) Test at short notice on a real engine and provide excellent feedback.
2) Get the loader rock solid on all three OSes so that James' work is worse, not better, than yours.
3) Develop detailed requirements docs for *any* high speed IO sub-system.

IN THAT ORDER.
That's not true at all, I have told you on MANY occasions how I'm willing to get my stuff to a state that allows fast vanilla code testing.
Again, actions speak louder than words, keeping the car in a state where you can test vanilla code in the time it takes to walk out and unlock the vehicle = helpful. As mentioned above, not doing that = destructive and damaging.
If I can get my personal satisfaction/gratification and help test vanilla code then we BOTH win. In a sense that's my paycheck.
Then make it happen. PS, this isn't about me winning. It's about everyone winning, or only you winning, or both...
Once this summer is gone, here's what I expect from you, and to the maximum extent possible, demand from you:
You and my wife both!
I'll be in touch with her to make sure she gives you A) absolute hell B) no nooky. :-)
it's illegal to work on cars in my driveway.
Land of the free, right? My arse. Yuck. Flag having a driveway, build the garage out to the footpath edge.

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
SleepyKeys
LQFP144 - On Top Of The Game
Posts: 549
Joined: Mon Feb 11, 2008 10:52 pm
Location: Arizona
Contact:

Re: Fred's firmware development diary comments thread

Post by SleepyKeys »

"1) Jewish tin-arse tight-arse clam-wallet squeaky-arse clenchfistedness :-) Sorry, ran out of adjectives. "

Does living with this count as an adjective?
Image
You snooze, you lose!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Fred's firmware development diary comments thread

Post by Fred »

Is that in Rankine scale? Pretty cold, I'll give you that!
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
Post Reply