Import libusb 1.0.9-rc3 into third_party

Import libusb 1.0.9-rc3 into third_party.

BUG=none
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=129069

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=131622

[Daniel Berlin, 2012-03-22 20:27:33]

lgtm

[commit-bot: I haz the power, 2012-03-23 19:29:41]

No LGTM from a valid reviewer yet. Only full committers are accepted.
Even if an LGTM may have been provided, it was from a non-committer or
a lowly provisional committer, _not_ a full super star committer.
See http://www.chromium.org/getting-involved/become-a-committer
Note that this has nothing to do with OWNERS files.

[bryeung, 2012-03-23 19:42:58]

LGTM

On 2012/03/23 19:29:41, I haz the power (commit-bot) wrote:
> No LGTM from a valid reviewer yet. Only full committers are accepted.
> Even if an LGTM may have been provided, it was from a non-committer or
> a lowly provisional committer, _not_ a full super star committer.
> See http://www.chromium.org/getting-involved/become-a-committer
> Note that this has nothing to do with OWNERS files.

[commit-bot: I haz the power, 2012-03-23 19:56:50]

No LGTM from a valid reviewer yet. Only full committers are accepted.
Even if an LGTM may have been provided, it was from a non-committer or
a lowly provisional committer, _not_ a full super star committer.
See http://www.chromium.org/getting-involved/become-a-committer
Note that this has nothing to do with OWNERS files.

[commit-bot: I haz the power, 2012-03-23 20:34:04]

No LGTM from a valid reviewer yet. Only full committers are accepted.
Even if an LGTM may have been provided, it was from a non-committer or
a lowly provisional committer, _not_ a full super star committer.
See http://www.chromium.org/getting-involved/become-a-committer
Note that this has nothing to do with OWNERS files.

[Garret Kelly, 2012-03-26 18:57:08]

Adding miket as a reviewer.

[miket_OOO, 2012-03-26 19:43:40]

On 2012/03/26 18:57:08, Garret Kelly wrote:
> Adding miket as a reviewer.

LGTM (semi-rubber-stamp). I downloaded
http://git.libusb.org/?p=libusb.git;a=commit;h=f07a4a78533b44d124dfe06cbf42af...
and compared shasums for common files.
third_party/libusb/libusb/os/poll_windows.h is different but the change appears
Good And Proper.

[Garret Kelly, 2012-03-26 20:26:08]

> LGTM (semi-rubber-stamp). I downloaded
>
http://git.libusb.org/?p=libusb.git;a=commit;h=f07a4a78533b44d124dfe06cbf42af...
> and compared shasums for common files.
> third_party/libusb/libusb/os/poll_windows.h is different but the change
appears
> Good And Proper.

The poll_windows.h changes are due to how we compile under MSVC. Thanks for
checking against the original commit!

[commit-bot: I haz the power, 2012-03-26 20:31:25]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gdk@chromium.org/9826025/8001

[commit-bot: I haz the power, 2012-03-27 00:36:25]

Change committed as 129069

[commit-bot: I haz the power, 2012-04-10 18:47:30]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gdk@chromium.org/9826025/29001

[commit-bot: I haz the power, 2012-04-10 20:24:39]

Change committed as 131622

[Paweł Hajdan Jr., 2012-04-23 16:23:47]

Please add support for using system libusb. You can use e.g. use_system_speex,
use_system_webp and use_system_xml as examples how to do that, and I'd be happy
to answer your questions about that.

By the way, I'm not very happy that you've said you're going to included that
option and didn't (sorry if I just didn't notice it).

[Garret Kelly, 2012-04-23 18:08:58]

> Please add support for using system libusb. You can use e.g.
> use_system_speex,
> use_system_webp and use_system_xml as examples how to do that, and I'd be
> happy
> to answer your questions about that.

Will do. I'm currently in the middle of a fairly important sprint, so
it's going to have to wait until after.

> By the way, I'm not very happy that you've said you're going to included
> that
> option and didn't (sorry if I just didn't notice it).

It wasn't clear from our discussion that you expected that to be in
this change. I'll add it after this sprint is over.

Garret

[Paweł Hajdan Jr., 2012-04-24 06:58:20]

On Mon, Apr 23, 2012 at 20:08, Garret Kelly <gdk@chromium.org> wrote:

> > Please add support for using system libusb. You can use e.g.
> > use_system_speex,
> > use_system_webp and use_system_xml as examples how to do that, and I'd be
> > happy
> > to answer your questions about that.
>
> Will do. I'm currently in the middle of a fairly important sprint, so
> it's going to have to wait until after.


What's so hard about doing it? I've heard this "I'm doing ... important
thing so this can wait" way too often, and most of the time it's used to
justify some kind of quality compromise.


> > By the way, I'm not very happy that you've said you're going to included
> > that
> > option and didn't (sorry if I just didn't notice it).
>
> It wasn't clear from our discussion that you expected that to be in
> this change. I'll add it after this sprint is over.


I wonder why you didn't ask then. :-/ Here's the relevant e-mail exchange
(you've said nothing you plan to do that later, so it seems reasonable to
assume it's going to be in the same CL; generally it's better to do things
"right" the first time):

On Fri, Mar 30, 2012 at 21:39, Paweł Hajdan, Jr. <phajdan.jr@chromium.org>
 wrote:

> On Thu, Mar 29, 2012 at 00:52, Garret Kelly <gdk@google.com> wrote:
>
>> > By the way, on Linux we should most likely use the system libusb, or at
>> > least have an option to do so at build time (-Duse_system_libusb). This
>> is
>> > important to be a good Linux distro citizen (the option; Google Chrome
>> can
>> > bundle anything it wants, the thing is about Chromium the open source
>> > project). Feel free to ask any questions about that, and I can also
>> show you
>> > how to do that in gyp.
>>
>> That's quite true, I will add options to allow for this. Thanks for
>> pointing it out.
>
>
> Cool, feel free to add me to the review of that option then (there are
> several pitfalls too; e.g. don't create automagic dependencies, see
> http://www.gentoo.org/proj/en/qa/automagic.xml).
>

[bryeung, 2012-04-24 14:43:17]

Hello,

Sorry we've run into this confusion.

On Tue, Apr 24, 2012 at 2:57 AM, Paweł Hajdan, Jr.
<phajdan.jr@chromium.org> wrote:
>> Will do. I'm currently in the middle of a fairly important sprint, so
>> it's going to have to wait until after.
>
> What's so hard about doing it? I've heard this "I'm doing ... important
> thing so this can wait" way too often, and most of the time it's used to
> justify some kind of quality compromise.

We're not trying to justify any kind of quality compromise here.
We're right at the end of a sprint targeting an internal ChromeOS team
meeting and Garret is working on an important piece.  The demo should
be code complete by Friday, but Garret's piece looks like it will get
in early.  So this should get taken care of late this week, or early
next at the very latest.  It'll be our top priority to get this
cleaned up as soon as we've landed the code for the demo.

Bryan

[Garret Kelly, 2012-04-24 14:47:14]

Hi Paweł,

> What's so hard about doing it? I've heard this "I'm doing ... important
> thing so this can wait" way too often, and most of the time it's used to
> justify some kind of quality compromise.

Nothing's hard about doing it, nor did I say that I wasn't going to do
it. I said I'd do it after this sprint, which is later this week or
early next week. I'm glad you're so committed to code quality, but
this code isn't even used yet. I'm not attempting to indefinitely
postpone this feature, the timeline is very clear and is less than 7
days.

> I wonder why you didn't ask then. :-/ Here's the relevant e-mail exchange
> (you've said nothing you plan to do that later, so it seems reasonable to
> assume it's going to be in the same CL; generally it's better to do things
> "right" the first time):

I didn't ask for the same reason you didn't ask. We both assumed that
the other had different expectations. I'm very sorry that this
occurred, as it's clearly caused you some consternation. This was
never my intent, and I'm very much committed to resolving this issue.
I've created http://crbug.com/124803 and CC'd you on it so that you'll
know the moment it's done.

>> Cool, feel free to add me to the review of that option then (there are
>> several pitfalls too; e.g. don't create automagic dependencies,
>> see http://www.gentoo.org/proj/en/qa/automagic.xml).

You explicitly mention the feature being in a different review here,
which could have been one reason that I assumed that you didn't need
it in the same change. Again, I'm very sorry that wires got crossed
here. I'm honestly not attempting to skirt around this issue.

Garret

[Paweł Hajdan Jr., 2012-04-25 11:16:49]

On Tue, Apr 24, 2012 at 16:46, Garret Kelly <gdk@chromium.org> wrote:

> Nothing's hard about doing it, nor did I say that I wasn't going to do
> it. I said I'd do it after this sprint, which is later this week or
> early next week. I'm glad you're so committed to code quality, but
> this code isn't even used yet. I'm not attempting to indefinitely
> postpone this feature, the timeline is very clear and is less than 7
> days.


Thank you, the timeline looks good to me. Just a note (to clarify), this is
now shipping in 20.0.1115.1 dev channel release, and it does affect Linux
distros that package it (e.g. Gentoo), as a slight inconvenience (only
because it's in dev; in beta and stable we really should have options to
use system libraries,
http://fedoraproject.org/wiki/Packaging:No_Bundled_Libraries).

I've written some general plan how to add the option, see
http://code.google.com/p/chromium/issues/detail?id=124803 . If at some
point you're confused about what's needed to support the system libraries,
I'm happy to help. It's a common issue people run into when checking in
third party libraries into Chromium tree.

Just to clarify (if someone else reads this later), I'm much more happy
when the first patch adding a library to third_party supports
use_system_... option. Sorry I didn't communicate that clearly before; your
response looks totally good to me. :-D