On Thu, Aug 20, 2009 at 6:21 AM, Ben Goodger (Google)<b...@chromium.org> wrote: > I don't know much about the technical details at play here, but a > couple of high level notes: > > - I am sympathetic to concerns around codebase cleanliness. Many > people (like Brett) have spent very many months maintaining and > improving the hygiene of Chrome code. Sometimes it feels like an > uphill battle. Some people (like myself) tend to be more forgiving of > temporary clutter when you have an established track record of making > these changes and then swiftly sweeping up afterwards. > > - We have a growing number of ifdefs. It's getting hard to understand > when and why each is set. As someone proposing to add more, it'd be > much appreciated if you'd put together a doc on our website (and link > it up) noting when the common ones are set. If you start such a doc, > people can continue to augment/update it with others.
I'd be happy to do that. When I do, there's something that's already puzzling me, and that's OS_POSIX. I don't have a copy of the POSIX standard, at least not a recent one, so its hard to know what is or isn't POSIX, and I imagine I am not alone in that. However, various comments lead me to believe that OS_POSIX doesn't really mean "POSIX" in people's minds - it really means "UNIXish" or "not Windows" or something. How would I document this define? Is there an agreed meaning? > > So sorry if it seems like you're getting the third degree here, I just > think it's a good idea for the team at large to know what's going on > so we can all remember to follow up from time to time. I'm not complaining. > > -Ben > > On Wed, Aug 19, 2009 at 11:43 AM, Ben Laurie<b...@chromium.org> wrote: >> I've started working on a FreeBSD port. The first patch is >> here: http://codereview.chromium.org/172032. >> When looking at the patch, bear in mind a couple of things... >> 1. Added gyp lines for files like *_ar.pak are compensating for the fact >> that i18n targets are not currently being handled correctly, and this can >> break the build, particularly when -j is not used. There are TODOs to make >> them work properly. They aren't really part of the port, but because I have >> no build farm for FreeBSD, the problems show up. >> 2. There are now some directories that are called "linux" or "mac" but are >> used for FreeBSD, too. I'm hesitant to rename these at this point, because >> it may turn out later that actually FreeBSD-specific versions are needed. >> Views welcome, of course. >> Anyway, there's been some debate about how to proceed in terms of ifdefs. >> The observation is that many places that are currently: >> #if defined(OS_LINUX) >> are going to become: >> #if defined(OS_LINUX) || defined(OS_FREEBSD) >> and this is ugly. >> There's a temptation to instead say these are both POSIX, but not MACOSX, >> for example as here: >> http://codereview.chromium.org/172032/diff/3003/3013 >> but this may not always be true (to be honest, I'm not even sure if its true >> for that case). >> Does the list have a view on how this should be handled? >> >> >> >> > --~--~---------~--~----~------------~-------~--~----~ Chromium Developers mailing list: chromium-dev@googlegroups.com View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-dev -~----------~----~----~----~------~----~------~--~---
