Newsgroups: php.internals Path: news.php.net Xref: news.php.net php.internals:1574 Return-Path: Mailing-List: contact internals-help@lists.php.net; run by ezmlm Delivered-To: mailing list internals@lists.php.net Received: (qmail 3285 invoked from network); 16 May 2003 15:03:42 -0000 Received: from unknown (HELO mail-2.dinet.de) (212.8.6.1) by pb1.pair.com with SMTP; 16 May 2003 15:03:42 -0000 Received: from ots-2 (carbon.gonicus.de [212.8.6.6]) by mail-2.dinet.de (Postfix) with ESMTP id 8EE392CC009 for ; Fri, 16 May 2003 17:03:40 +0200 (CEST) Organization: GONICUS GmbH To: PHP Internals Date: Fri, 16 May 2003 17:03:40 +0200 User-Agent: KMail/1.5.1 References: <200305161525.04359.holger.burbach@gonicus.de> <20030516133408.GA11228@lxr> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-ID: <200305161703.40465.holger.burbach@gonicus.de> Subject: Re: [PHP-DEV] PECL Package Proposal: kadm5 - REPOST From: holger.burbach@gonicus.de (Holger Burbach) On Freitag, 16. Mai 2003 15:46, Derick Rethans wrote: > On Fri, 16 May 2003, Sander Roobol wrote: > > On Fri, May 16, 2003 at 03:25:04PM +0200, Holger Burbach wrote: > > > There is a new release of kadm5 with some minor changes. The sources > > > can be found at: > > > > > > ftp://oss.gonicus.de/pub/php-kadm5/php-kadm5-0.2.2/ > > > > Looks fine, BUT (Derick, still reading? :) > > 1. The biggest problem I see is that the extension is throwing E_ERRORs > which extensions really should not do if they do not leave *PHP* in > an unstable state (there is also no \n required at the end of the text, > and in the best case php_error_docref should be used). So throwing E_WARNING would be better? > 2. Check our CODING STANDARDS file again :) Okay, I'm working on that. > 3. I would recommend to use this naming (As example): > kadm5_principal_create > kadm5_principal_modify > > instead of: > kadm5_create_principal > kadm5_modify_principal Sounds reasonable. > 4. There are no prototypes for the PHP_FUNCTIONs Okay, I'm working on that. > 5. There should be atleast one test per function, if possible. Hmm, to test the functions I have to connect to an Kerberos V administration server... > 6. The package file has some unescapted entities (the & in "Cleanup in > kadm5.c: Removal of confirm_kadm5_compiled & RINIT/RSHUTDOWN" for > example) Well, it seems that I have some things to do... Regards, Holger