Re:Re: It seems that Firefox is now at 115.0.3esr

[chippy]

I managed to move on from previous problems and I can successfully run
./mach configure (without wasm for now).

After that I found a couple of problems that need to be solved in
makeicecat I believe:

For instance I found that in
devtools/client/netmonitor/src/connector/moz.build

before running makeicecat I have :

DevToolsModules(
    "firefox-data-provider.js",
    "har-metadata-collector.js",
    "index.js",
)

The list is alphabetically sorted, but after applying the changes, I
have 

DevToolsModules(
    "icecat-data-provider.js", 
    "har-metadata-collector.js", 
    "index.js", 
)

Which is not sorted any more and this breaks the build with
mozbuild.util.UnsortedError. This I think, requires an extra step in
makeicecat, to obtain:

DevToolsModules(
    "har-metadata-collector.js",
    "icecat-data-provider.js",
    "index.js",
)

which is ordered.

Other problems I'm finding are like the Firefox/Icecat sed
replacements/patches/etc  also  did something unexpected to the bare
syntax of some files such as adding comments to lines that contain a
function declaration leaving the closing bracket around, so basically
making the source just not working per se.
(netwerk/protocol/http/nsHttpHandler.cpp, seen in makeicecat) 

It is not clear which part of makeicecat does this, so, for the sake of
the build I have fixed this by hand in the output directory (I know it
is not sustainable, and requires changes in makeicecat).

Despite this, it still breaks on a file that does not appear to have
the described problem, which means that it is upstream on some other
file(s). 
Build log: https://paste.classictetris.net/eXzDFd 


For solving this I need to narrow down what is causing it, so right now
I'm t trying to see which files are affected with cpplint, which now
gives too much output even when using filters. 
I am now basically looking to recursively spot unmatched brackets (that
is). Probably not the most effective way, because even if I do, there
might be more.

Next step (if this fails) is to manually run the 'finds' used by
makeicecat to feed files to sed and see which ones alter the files I
alrady know being syntactically wrong. I also need to take a better
look at the patches...

I whish I could run makeicecat in chunks but the rename and sed are
tightly coupled and skipping either would result in lots of files
having the wrong name, later on, during build.

I any case, if someone is already working on this, pretty please let me
know so we can either coordinate the effort or just let this be done by
more skilled people (I'm a devops engineer).

Open to suggestions/directions/help.


Thanks,
Chippy.