Created attachment 146143 [details] fcgiwrap.patch
title is misleading, it's just bumping the portrevision (to 2, not 5). Note that the diff version is broken so only half the patch is showing.
actually, this has a maintainer, ask his approval first.
In order to avoid a maintainer timeout I acknowledge having received this report and I will look at it sometime between right now and shortly after 31 August when the STAGEDIR storm dies. Thanks to the submitter for his/her contribution; it will be looked at in no more than a week from now.
Ok Fonz, and the answer is?
There appears to be a slightly (minor-minor, so to speak) new upstream version of this port available upstream. I'm currently testing whether it works properly with FreeBSD and to what extent it fixes the two pending PRs for this port.
Comment on attachment 146143 [details] fcgiwrap.patch Feel free to reject current the patch using the maintainer_approval flag, and request a status change according to your needs.
The patch as it is needs some work. I'm on it.
@submitter: Can you elaborate why you think the change to the signal handler is necessary? As far as I can tell the waitpid() call works the same on FreeBSD as it does on Linux (which I presume is the primary development platform for this software) so the question is whether you're fixing a bug or changing something the author(s) really intended as it is.
Update: I've isolated the different aspects of the patch (which are several) and contacted the maintainer about the signal handler issue. To be continued...
Maintainer -> upstream vendor - sorry!
So far I have heard from neither the submitter nor the upstream author. So I'm probably going to have to make a judgement call here or there, which I'll do this weekend unless I hear otherwise.
I just got a notification that this PR needs special attention. Update: the submitted patch is rejected and I'm working on a replacement patch as we speak.
Created attachment 147558 [details] New patch, maintainer approved. The attached new patch has maintainer approval and can be committed as far as I'm concerned. The modifications are as follows: * The changes to the Makefile are good and have been kept. Thanks! * The patch for the README has been omitted. That file is not installed anyway so there's no point. But even then, I think this is the upstream vendor's job, not ours. * The modification to the signal handler has been abandoned (at least for now) because nobody cared to convince me of its correctness. * The additional feature has been kept. I also updated the man page accordingly, which the submitter forgot. However, I do suggest the submitter send their work to the upstream vendor so it may be included in a new release if there will be one.
Please update the issue summary to reflect the changes, as latest patch is not an "update to 1.1.0"
The summary/title has never been right in the first place and doesn't reflect what the submitter's patch does either. Anyway, as far as I can see, I can't change the summary/title myself. That's probably committer's prerogative. But if you need a suggestion, I'd say something like: "Improved handling of binary stripping and addition of a new command line option that restricts what may be run" ought to cover it nicely. Or, if you like it shorter, replace the second part with simply "addition of a new feature".
fonz has suffered enough
A commit references this bug: Author: marino Date: Sun Oct 5 23:37:47 UTC 2014 New revision: 370135 URL: https://svnweb.freebsd.org/changeset/ports/370135 Log: www/fcgiwrap: Improve binary strip handling, add new feature PR: 192907 Final version: maintainer (fonz) Changes: head/www/fcgiwrap/Makefile head/www/fcgiwrap/files/patch-fcgiwrap.8 head/www/fcgiwrap/files/patch-fcgiwrap.c
Thanks!