httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Eric Covener <cove...@gmail.com>
Subject Re: svn commit: r1082196 - in /httpd/httpd/trunk: CHANGES docs/manual/mod/core.xml include/ap_mmn.h include/http_core.h include/httpd.h server/core.c server/request.c server/util.c
Date Wed, 18 May 2011 18:15:54 GMT
On Wed, May 18, 2011 at 2:06 PM, Eric Covener <covener@gmail.com> wrote:
> On Wed, May 18, 2011 at 2:04 PM, William A. Rowe Jr. <wrowe@apache.org> wrote:
>> On 5/18/2011 12:34 PM, Philip M. Gollucci wrote:
>>>> Modified: httpd/httpd/trunk/include/httpd.h
>>>> URL:
>>> http://svn.apache.org/viewvc/httpd/httpd/trunk/include/httpd.h?rev=1082196&r1=1082195&r2=1082196&view=diff
>>>>
>>> ==============================================================================
>>>> --- httpd/httpd/trunk/include/httpd.h (original)
>>>> +++ httpd/httpd/trunk/include/httpd.h Wed Mar 16 16:45:25 2011
>>>> @@ -1496,7 +1496,7 @@ AP_DECLARE(int) ap_unescape_url(char *ur
>>>>   * @param url The url to unescape
>>>>   * @return 0 on success, non-zero otherwise
>>>>   */
>>>> -AP_DECLARE(int) ap_unescape_url_keep2f(char *url);
>>>> +AP_DECLARE(int) ap_unescape_url_keep2f(char *url, int decode_slashes);
>>>
>>> Did we mean to change a public API in a patch release ? .17 -> .18 ?
>>> This has broken things in FreeBSD ports and is being reported a bit.
>>> I see the mmn bump so you can detect it at least.
>>
>> If this was backported, that's bad.  We proclaim binary compatibility within any
>> stable sub-rev release.  Sorry that I did not catch this.
>>
>> http://svn.apache.org/viewvc/httpd/httpd/branches/2.2.x/include/httpd.h?r1=1076611&r2=1082630
>>
>> I suggest we revert in 2.2.19 and consider this .18 an aberration/broken release.
>> I would offer to tag and roll today.  Those who wish to work around it, may, but
>> a quick reaction would allow folks to avoid it.
>>
>> Other thoughts?
>
> Missed this in review.  Seems we could add the old signature back in
> and grab the setting from the core dir_config instead of reverting the
> entire thing? (util.s already defines CORE_PRIVATE)
>

Actually not even this much, since the old caller of "keep2f" always
decoded.  Can we resolve with just  e.g.:

AP_DECLARE(int) ap_unescape_url_keep2f(char *url)
{
   return ap_unescape_url_keep2f(url,1);
}

?

-- 
Eric Covener
covener@gmail.com

Mime
View raw message