Major problem with 1.31

[BarryC.]

no thats the only one I had to modify,
I think everything else is ok

[cossie1]

no thats the only one I had to modify,
I think everything else is ok

Map looks similar to the JDM version now

Just the knock sensor filter map on the JDM ecu looks scary, lol

Oh and whats with the 11000 rpm timing being mapped in ?

[burgers]

Hi

Here is an xml of the JDM VII 98640014 for version 1.31 of ECUflash. Thanks for the pointers on the edian issue

MB

[ahbao9388]

im running problems opening my rom file. its all out of place. anyone can help? i have attached both my rom n my xml

[BarryC.]

im running problems opening my rom file. its all out of place. anyone can help? i have attached both my rom n my xml

I have opened your rom and xml and they work fine for me ?????
did you download the updated base files on the previous page on this thread????
did you even read this thread????

[ahbao9388]

im running problems opening my rom file. its all out of place. anyone can help? i have attached both my rom n my xml

I have opened your rom and xml and they work fine for me ?????
did you download the updated base files on the previous page on this thread????
did you even read this thread????

sorry computer idiot here. updated and all is well now. anyway it seems to me that the new base has lesser tables defined. will there be an updated one?

[EvilevoIX]

So it would NOT be a good idea to upgrade to the latest version and keep my 1.30 Version until further notice?

[cboles]

I made the endian change in 1.31 because the previous versions had the meaning reversed from the correct use of the term:

http://en.wikipedia.org/wiki/Endian

I had known about this error for some time, but was hesitant to make the change as it would create problems for people using old definitions. The 1.31 release has a significant number of new Subaru definitions thanks to Merchgod, so I figured it was a good time to make the change.

I believe all of the definitions in the 1.31 release are correct and use the correct endian. I think the problems people are seeing are due to using old or self-made definitions that were compatible with 1.30 and earlier. If you are dealing with byte sized scalings, endian won't matter as the result with be the same. All of the SH CPU based Subarus and Mitsubishis are using big endian though, so it make no sense to use "little" endian in scalings, even if they are only 8 bit.

Most users who are upgrading from 1.30 or earlier can resolve this issue by going to File | Options | Directories | Metadata Directory and changing this option to

C:/Program Files/OpenECU/Ecuflash/rommetadata
(or the non-default location to which you installed EcuFlash)

so that you can return to using the definitions that came with the installation.

Probably an endian mixup due to the fix in 1.31 where the endian attribute reversal was fixed. Find a table that has the issue and open your base xml file (ex. evo9base). Search for the table and find the scaling attribute for the axis is question. Look at the endian attribute. If it is "big" then change to "little" or vice versa. See if that fixes the issue.

That fixes the issue for all of the definitions I have. I always was puzzled why the endian order was reversed in older versions. I had just assumed it was ignored on the older versions.

Really easy to "touch up" I have to go through all my base defs and adjust them, but all the important ones are fixed..

[FQ-Airtrek]

I have finally get 1.31 working !! Thanks all! The graphs display is excellent.

Initially was a bit confusing but with plenty of help here I managed to get it working!

I migrated from 1.29a to 1.31.

Cboles thanks!

The necessary steps:

- Map to your rommetadata directory in yr laptop
- Copy the new EVO7BASE.xml and EVO9BASE.xml and the other xmls posted here
- for those xml u have created, change the endian accordingly to the new EVO7/9 BASE.xml from "little" to "big". This will fix the data display errors. To compare the scaling just open up a working xml, click on scaling. there you will see the endian parameter and edit accordingly
- If you inherited from the EVO7/9 Base. The new EVO7/9 base will fix the endian