My test with the source code

Started by scrax, June 15, 2012, 12:33:09 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

scrax

Hi, here i'll like to post my experience in compiling and patching source code to learn how it works basically. I'm not a developer, just i like to read the code and figure out how it works, changing some value to add simple things like menu or feat.

14/06/2012
so far i've added one thing that for me was useful like rename back avi to cr2 after imported by my eyefi, so i've copied the code from cr2toavi and changed the extensions to have the revert function.
after that i've made a nice submenu for the two options and moved it to tweaks from debug looking how other menu are done.
Then there was something wrong with compilation of doc and i've made some correction for osx in makefile to have a almost working compilation of docs, still something missing but needs more investigations.
I've also tried to enable audio photo trigger at least but it's buggy (of course).

Now, since i suppose to have made some mess with hg i want to restart from start to make a working fork and maybe another one for my test and post here how it goes.

First I've installed all the things to compile ML (se other post) and then from terminal:
hg clone -r unified https://bitbucket.org/hudson/magic-lantern

then I've made some changes to Makefile.inc as suggested on the wiki for a compiling issue.
this is what i've changed to compile on osx: 657d28d4947f

adding the eyefi menu was done here: 9ec61f42a6e9

i've also added a new version of my preferred cropmark crossmeter with a trasparent dark overlay for the guide here 32ca15089015

and after reading an issue on the magiclantern page i've tried to enable intermediate iso for hdr an it worked good (at first) here: 28454caf86bb

now i'm working on prepare.Lion.sh to automate the install of all what is needed on osx to compile ML. here what we have so far: b89ec23455cf

15/06/2012
Another feat missing for me is some sort of "organization" of the card. So that's why all thing were moved except for autoexec.bin (and ROM0.BIN, BOOT0.BIN) to a ML folder on the card.
Now my card use this file structure:
autoexec.bin
DCIM/
MISC/
ML/DOC
ML/PLUGINS
ML/CROPMKS
ML/SCRIPTS
ML/SETTINGS

I'm thinking about moving ML/MAGIC.CFG ML/FONTS.DAT, ML/CROPMKS and ML/RECTILIN.LUT to ML/SETTINGS/ or ML/REQUIRED/ and all the ML generated files to ML/LOGS/ done

For now i've tested cropmarks, defishing, format etc. and all is working ok. here the changes made: f833a8a9917c and here: 3703530d7a52

IMHO a "standard" well defined file structure is better to make scripts for managing ML (for example something that checks the ML/LOGS/ folder to post the log automatically). And when this script thing are done a ML/TOOLS folder could be added for distribuite them.

17/6/2012
I've found some wrong link missed in the folder structure, now fixed here: 71290d361dc1
Also plugins compilation now is active, to start testing it on 600D, so far all went good after patching it to the different folder.

25/6/2012
I've finally moved all ML stuff into an appropriate folder (and it's now on the unified branch), also then found a way to hide EyeFi trick menu when the card is not EyeFi.
Now, docs compilation works, and MLTools is next to be almost all working. Not all is yet well tested but most of it now works like expected.
I've also started adding some options for ubuntu, but i'm not sure if they are needed or working.

01/07/2012
Alex introduced a new menu system that is great, I've just enables some submenu he left commented out and resized Mics key shortcuts because it was too small for 600D options. Also moved EyeFi trick to Shoot from Movie because it's for transfer RAW pic to PC.
Another small change is adding Dark Green to display color schemes.
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-