Re: [PATCH] compat: add a getpass() compatibility function

From: Erik Faye-Lund <kusmabite@gmail.com>
Date: May 19, 2011, 20:12 UTC
Message-ID: <BANLkTimG8E_Riz3rYC9PMw_2-D=Za0Ar6w@mail.gmail.com>
In-Reply-To: <74D1D7FC-A747-4F85-8B1E-7ABFC9DA70A3@gieschke.de>

On Thu, May 19, 2011 at 9:42 PM, Rafael Gieschke <rafael@gieschke.de> wrote:

Show 30 quoted lines

>
> Am 19.05.2011 um 21:19 schrieb Erik Faye-Lund:
>
>> On Thu, May 19, 2011 at 9:16 PM, Rafael Gieschke <rafael@gieschke.de> wrote:
>>>
>>> Am 19.05.2011 um 20:07 schrieb Erik Faye-Lund:
>>>> diff --git a/compat/mingw.h b/compat/mingw.h
>>>> index 62eccd3..e37d557 100644
>>>> --- a/compat/mingw.h
>>>> +++ b/compat/mingw.h
>>>> @@ -48,6 +48,8 @@ typedef int socklen_t;
>>>> #define EAFNOSUPPORT WSAEAFNOSUPPORT
>>>> #define ECONNABORTED WSAECONNABORTED
>>>>
>>>> +#define PASS_MAX 512
>>>> +
>>>
>>> It might be worth to have this in git-compat-util.h, so it can also be used by compat/getpass.c. I'm not sure about the name, however; it's a remote password and MAX_PASS is about local passwords AFAIK. Maybe PASS_MAX_REMOTE or something similiar. Like that, compat/getpass.c could also be used in case MAX_PASS has some insanely low value on a platform.
>>>
>>>
>>
>> PASS_MAX should be defined in limits.h (which is included by
>> git-compat-util.h) on POSIX platforms:
>> http://pubs.opengroup.org/onlinepubs/007908799/xsh/limits.h.html
>>
>> If it turns out it's not on some platforms, perhaps a define-guarded
>> version would make sense in git-compat-util.h instead.
>
> It is defined even on Android :-).
>

Good to hear.

> But I still wonder whether it's the right thing or if you should use it only for local passwords.
>
> If you imagine a platform with MAX_PASS=8, git couldn't connect to an IMAP server with a password of 10 chars on this platform without any proper reason.
>

Well, those platforms would currently fail, since that's the limit on the string returned from getpass. Since that hasn't happened yet, I suspect that this is a very theoretical problem.

If we created our own define we could fix such problems by setting NO_GETPASS and having a reasonable lengthy GIT_MAX_PASS. But let's leave that theoretical fix for when/if it turns out to be real, huh?