Re: [Zope-dev] Re: Post-mortem [Was: Zope 2.7.0 b3 regressions]
From: Evan Simpson [EMAIL PROTECTED] Lennart proposed additional methods, but I'm going to work on that in the post-2.7 branches, since I really want to clean up this mess properly, and beta 3 isn't the time. I totally agree. This is new features, introducing that in a beta3 seems very bad. I proposed introducing them in 2.7 only to: 1. Get things moving by scaring people. :-) 2. A way of fixing absolute_url by making it's bugs less important. If absolute_url can be fixed so it doesn't break backwards compatibility, then number 2 is of no importance. //Lennart ___ Zope-Dev maillist - [EMAIL PROTECTED] http://mail.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope )
Re: [Zope-dev] Re: Post-mortem [Was: Zope 2.7.0 b3 regressions]
Brian Lloyd wrote: Good call. I think it would be best to make sure the docstring of the new method is clear on its reason for being. I think somewhere there is an interface file that is used to generate some of the api docs - ideally that can get updated too. Done and done! Cheerios, Evan ___ Zope-Dev maillist - [EMAIL PROTECTED] http://mail.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope )
Re: [Zope-dev] Re: Post-mortem [Was: Zope 2.7.0 b3 regressions]
On Tuesday 09 December 2003 16:24, Evan Simpson wrote: Summary: absolute_url(1) Looks good. -- Toby Dickenson ___ Zope-Dev maillist - [EMAIL PROTECTED] http://mail.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope )
Re: [Zope-dev] Re: Post-mortem [Was: Zope 2.7.0 b3 regressions]
On Tue, Dec 09, 2003 at 10:24:47AM -0600, Evan Simpson wrote: def absolute_url_path(self): '''Return the entire path of the absolute URL for this object. This includes the leading slash, and can be used as an absolute-path reference as defined in RFC 2396. ''' OK. But maybe the docstrings could be more clear about BASEPATH1? IIUC, absolute_url(1) will leave out BASEPATH1, continuing the behavior from zope 2.7, but absolute_url_path() will include BASEPATH1? I propose this, in OFSP/help/ObjectManagerItem.py: def absolute_url_path(self): ''' Return the entire path portion of the absolute URL for this object. This includes the leading slash and BASEPATH1, and can be used as an absolute-path reference as defined in RFC 2396. Thus, this method is always safe to use with VirtualHostMonster. ''' def absolute_url(relative=None): Return the absolute url to the object. If the relative argument is provided with a true value, then the URL returned is relative to the site object. Does not include a leading slash or BASEPATH1, thus is not always safe to use with VirtualHostMonster. Code that uses '/'+absolute_url(1) should be changed to use absolute_url_path instead. Permission -- Always available I removed the logical rather than physical note from absolute_url because I don't think it was actually true before Evan's change, which IIUC is being reverted. Maybe I misunderstood. Lennart proposed additional methods, but I'm going to work on that in the post-2.7 branches, since I really want to clean up this mess properly, and beta 3 isn't the time. Great. -- Paul Winkler http://www.slinkp.com Look! Up in the sky! It's HYPO-SUPERCALIFRAGILISTICEXPIALADOCIOUS THIGH MAN! (random hero from isometric.spaceninja.com) ___ Zope-Dev maillist - [EMAIL PROTECTED] http://mail.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope )
RE: [Zope-dev] Re: Post-mortem [Was: Zope 2.7.0 b3 regressions]
Summary: absolute_url(1) didn't include the path to the virtual root, which broke code that assumed that it could just prepend / in all cases. I changed it to include the base path, and broke code that prepends BASEPATH1. Since the old behavior existed for two years (including part of the 2.7 beta), is arguably easier to work with, and works with existing CMF code, I will revert to it. This will leave /-prepending code broken. In line with Stefan's original suggestion, I will add a new method with alternate behavior. It's not hard to grep for broken code and use the new method to make it correct -- Product authors should do so. Good call. I think it would be best to make sure the docstring of the new method is clear on its reason for being. I think somewhere there is an interface file that is used to generate some of the api docs - ideally that can get updated too. Brian Lloyd[EMAIL PROTECTED] V.P. Engineering 540.361.1716 Zope Corporation http://www.zope.com ___ Zope-Dev maillist - [EMAIL PROTECTED] http://mail.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope )
