 |
 |
Hi, I am on it. My very personal opinion is that the original mistake was to keep it outside, hence my preference would still be to cleanly incorporate the voms extractor into the main codebase (with a well made Cmakelists of course). I am anyway investigating Brian's workarounds. Not very confident on the final outcome though. f On 02/21/2017 08:22 AM, Andrew Hanushevsky wrote: > You are correct Brian. I missed that use case. Your suggestions are quite reasonable. > > Andy > > On Mon, 20 Feb 2017, Brian Bockelman wrote: > >> >>> On Feb 20, 2017, at 7:24 PM, Andrew Hanushevsky <[log in to unmask]> wrote: >>> >>> Well, the issue now is that XrdHttp/XrdHttpSecXtractor.hh is in the public >>> header section. It should have been in the private header section since >>> it's part of the compiled package and no user would actually use it. >> >> Hi, >> >> The only two plugins that utilize this interface (xrootd-lcmaps, xrdhttpvoms) are both external projects and not part of the >> compiled package -- meaning the header is correctly marked as public. >> >> In fact, I'm a bit confused by the comment as there are no concrete implementations of the interface in the xrootd codebase. >> >> (Plus, it's an extremely useful interface as it allows a single authorization plugin for both the xrootd and HTTP protocols...) >> >>> According to EPEL rules we can't move it until a major release change. Of >>> course, given the circumstances we could get dispensation but that is up >>> to Mattias. >>> >> >> Why not: >> >> - Rename InitCtx back to the original Init >> - Make InitSSL / FreeSSL not pure-virtual (put in the trivial inline implementation in the base class). >> >> That fixes the API issue. Then beg forgiveness on the ABI break, given it's a new interface and the two consumers of the >> interface both can work with this (simple coordinated RPM "Conflicts:" statement). >> >> Brian >> >>> Andy >>> >>> On Mon, 20 Feb 2017, xrootd-dev wrote: >>> >>>> Hi, >>>> >>>> I never knew that the SecXtractor was considered >>>> a public interface, for sure that was not >>>> intended. Was it ? >>>> >>>> The recent changes wanted to comply with the guidelines >>>> for xrootd plugins loading, in order to make it possible to >>>> become a public interface with version checking, etc. >>>> >>>> Andy ? What do you prefer that we do here ? >>>> >>>> f >>>> >>>> >>>> >>>> >>>> >>>> >>>> On 02/20/2017 05:02 PM, ellert wrote: >>>>> I have received additional comments on the 4.6.0 update in EPEL testing: >>>>> >>>>> https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2017-9b2cd39ee3 >>>>> >>>>> First comment: >>>>> >>>>> It appears this PR breaks ABI / API compatibility in the XrdSecXtractor interface: #444 >>>>> <https://github.com/xrootd/xrootd/pull/444> >>>>> >>>>> My plugin https://github.com/bbockelm/xrootd-lcmaps fails to compile after the update. >>>>> karma: -1 >>>>> >>>>> Second comment: >>>>> >>>>> The configuration of for the pfc.* attributes changed, meaning existing configurations that use the caching proxy fail to >>>>> start after upgrade. >>>>> >>>>> The old-style attributes should still be permitted (especially as there appears to be a simple translation between old and >>>>> new). >>>>> >>>>> PR #444 <https://github.com/xrootd/xrootd/pull/444> that is referenced does mention that there is breakage, so maybe this was >>>>> intentional? >>>>> >>>>> ÿÿ >>>>> You are receiving this because you are subscribed to this thread. >>>>> Reply to this email directly, view it on GitHub <https://github.com/xrootd/xrootd/issues/470>, or mute the thread >>>>> <https://github.com/notifications/unsubscribe-auth/AD7YjnupBw1VGCbk4ExabxslSJfCOXtnks5rebkAgaJpZM4MGWj1>. >>>>> >>>>> >>>>> -------------------------------------------------------------------------------------------------------------------------------- >>>>> >>>>> >>>>> Use REPLY-ALL to reply to list >>>>> >>>>> To unsubscribe from the XROOTD-DEV list, click the following link: >>>>> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1 >>>>> >>>> >>>> >>>> -- >>>> You are receiving this because you are subscribed to this thread. >>>> Reply to this email directly or view it on GitHub: >>>> https://github.com/xrootd/xrootd/issues/470#issuecomment-281127802 >>> ÿÿ >>> You are receiving this because you commented. >>> Reply to this email directly, view it on GitHub <https://github.com/xrootd/xrootd/issues/470#issuecomment-281219301>, or mute >>> the thread <https://github.com/notifications/unsubscribe-auth/AD7YjmmapU9GmMtZmiGKEmTlP0ngfrXeks5rejzagaJpZM4MGWj1>. >>> >>> >>> Use REPLY-ALL to reply to list >>> >>> To unsubscribe from the XROOTD-DEV list, click the following link: >>> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1 >>> <https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1> >> >> ######################################################################## >> Use REPLY-ALL to reply to list >> >> To unsubscribe from the XROOTD-DEV list, click the following link: >> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1 >> > > ######################################################################## > Use REPLY-ALL to reply to list > > To unsubscribe from the XROOTD-DEV list, click the following link: > https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1 ######################################################################## Use REPLY-ALL to reply to list To unsubscribe from the XROOTD-DEV list, click the following link: https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=XROOTD-DEV&A=1