Logistics

[Keeper-of-the-Keys]

As discussed by mail I'm going to do piece meal PRs for the FTDI stuff.

Kicking it off with a few logistical parts of the big PR:

• .gitignore
• travis lint blacklist of autogenerated files
• Fix of escaping for README files into C headers.

[peternewman]

Some comments.

You could add a bit more code into the review if you wanted, like some of the timing changes?

[peternewman]

A few more minor comments, partly not actually your code sorry.

[peternewman]

Fix the typo

[peternewman]

Please can you fix the issues codespell caught, I appreciate they aren't yours (sorry). I'm genuinely not sure if the OutputStream one should be struct-based or structured, I'm beginning to suspect the latter.

https://travis-ci.org/OpenLightingProject/ola/jobs/605241223#L1061-L1062

[Keeper-of-the-Keys]

@peternewman Is anything still blocking this (the compile failures are I assume due to the race conditions since there is no reason to assume compilation would be affected by a commit that changes no code)?

[peternewman]

@peternewman Is anything still blocking this (the compile failures are I assume due to the race conditions since there is no reason to assume compilation would be affected by a commit that changes no code)?

I'm hoping to fix the Linux builds at least in #1589 just to check this passes that, it's probably a bit belt and braces.

[Keeper-of-the-Keys]

So nothing is blocking this from being merged anymore?

[peternewman]

Blocking on #1589 to ensure at least some builds complete with these changes, but the changes LGTM.

[peternewman]

So nothing is blocking this from being merged anymore?

Correct, well not of yours anyway.

[Keeper-of-the-Keys]

@peternewman I took the liberty of purging .travis.yml of all legacy tests to get an idea if this would lead to more successful CI builds and also speed up things...

If you oppose I will push another commit reinstating these checks however as I stated in #1589 I believe that gcc-8 should not be part of legacy tests since it is not part of their toolchains, if we want to test gcc-8 we can either add bionic tests to the existing matrix or we can say that "hey master deals with something that will be released and has no relevance on Ubuntu Xenial or OS X from 3 years ago lets treat it that way"

[peternewman]

@peternewman I took the liberty of purging .travis.yml of all legacy tests to get an idea if this would lead to more successful CI builds and also speed up things...

If you oppose I will push another commit reinstating these checks however as I stated in #1589 I believe that gcc-8 should not be part of legacy tests since it is not part of their toolchains, if we want to test gcc-8 we can either add bionic tests to the existing matrix or we can say that "hey master deals with something that will be released and has no relevance on Ubuntu Xenial or OS X from 3 years ago lets treat it that way"

Can you revert them for now (at least in this PR, as it's essentially ready to be merged). I'll try and find some time soon to respond more fully to your comments in the other PR, although an issue might be a more effective place to do it, currently the discussion is scattered around a number of PRs and other places (probably some on the mailing list too).

[peternewman]

@Keeper-of-the-Keys if you can revert the last three commits and either pull in the changes from #1590 or wait for @nomis52 to approve them and me to merge them then we can merge this in.

[Keeper-of-the-Keys]

Done and done

[peternewman]

Can you remove some errant merge stuff please.

[Keeper-of-the-Keys]

Whoops 1 second forgot why I was explicitly defining files instead of commit -a because I had that debug stuff and hadn't thrown it out yet.

[Keeper-of-the-Keys]

Done

[Keeper-of-the-Keys]

@peternewman the functionality you requested was not present as is and would have actually broken the C++ string...
I now added it for Linux but have no way to check it for OS X.

[Keeper-of-the-Keys]

reverted.
I love how it needs to be tested on every OS out there but Linux.... (because until I pointed out that on Linux it's not inserting newlines no one knew or cared)

[Keeper-of-the-Keys]

See #1593 for the other improvement.