This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Fix up LD_* vars behaviour


On Thu, Apr 12, 2012 at 1:26 PM, David Miller <davem@davemloft.net> wrote:
>>> If they are documenting the tool, and their documentation is accurate,
>>> then it's a valid reference point for what we are allowed to change on
>>> people or not.
>>
>> No, it's not. ?What the past behavior was is what matters, regardless of
>> whether or where it was ever documented.
>
> And I'm merely saying that such documentation, if accurate, is a
> viable reference for past behavior.

Agreed.

>> But in cases where there is not a clear specification such as POSIX
>> or ISO C indicating the detail should never have been relied upon,
>> then it's entirely a question of our judgment about what is better
>> for users overall.
>
> I think using such standards documentation to justify potentially
> breaking things on people would be very bad policy.
>
> If we made a mistake on an implementation and it's been out in the
> wild for some time, and people do depend upon it, we're essentially
> stuck supporting it forever whether we like it or not.

Agreed.

I think perhaps I didn't explain myself clearly enough in my first
email in this thread.

My guiding principles are:

(a) Compatibility.

- If we release it, and users use it, and can be reasonably expected
to have it keep working either by ambiguity in the standard or
documentation, then we must support it.

(b) Standards and documentation.

- If the standard or documentation says it is X, and the user used Y,
we have the ability to use common sense and good judgement to decide
if Y is too much burden for us to support when making future changes
to X.

I feel like Roland is talking about (a), and Dave is talking about
(b), and that you both actually agree when it comes down the case of
good judgement for users.

(1) The linux man pages.

Roland, unfortunately I feel that your position with respect to the
linux man pages project is not correct. As a project we have failed to
document our own interfaces and over the years the linux man pages has
become *the* canonical documentation for the GNU Libc interfaces like
LD_* vars passed to ld.so.1. Whether we agree with them or not is
irrelevant at this point, we need to think about our users, and our
users for lack of our own documentation, rely on the linux man pages.

We should still rely on (a) and (b), but now (b) includes a set of
documentation that we don't control. We can change this situation but
it takes work on our part.

Does that clarify my position?

Cheers,
Carlos.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]