Re: HTTP-1.3.0RC1 (beta) Released.
| From: | Stefan Neufeind | Date: | Thu, 10 Jun 2004 21:24:53 +0000 |
| Subject: | Re: HTTP-1.3.0RC1 (beta) Released. | ||
| References: | 1 | Groups: | php.pear.dev |
| Request: | Send a blank email to pear-dev+get-30440@lists.php.net to get a copy of this message | ||
On 10 Jun 2004 at 17:20, Daniel Convissor wrote:
> On Thu, Jun 10, 2004 at 04:18:36PM -0400, Daniel Convissor wrote:
> >
> > Mike, can you please send me the tarball of HTTP-1.3.0RC1.tgz? I'll pass
> > it through my test server and see what happens.
>
> Sure enough. The database query is dying when processing this tag:
>
> <dep type="PHP" rel="ge" version="4.0.6"/>
>
> Under the current system, "PHP" needs to be lower case.
>
> So, Mike, remove the CVS tag, update the package.xml file to have the dep
> say "php", put the CVS tag back and re-package, then upload the new
> release.
>
> Since RC1 never made it into the database or server, go ahead and reuse
> RC1 as the release name.
>
> Now, a question for everyone, how should we handle this validation in the
> future? pearweb/include/pear-database.php line 1400 says:
>
> if ($dep['type']=='php') {
>
> So, the simplest validation process is to add the following before getting
> to that if statement:
>
> $dep['type'] = strtolower($dep['type']);
>
> Or, should we catch this and raise an error? I'm thinking it's simplest
> and not a big deal just to lowercase it.
Could we auto-lowercase it upon packaging? I don't think that the
validation should allow such files with uppercased types to path
through - because then we'd end up with files havin php, PHP and Php
in them, which would also require "flexible" parsing later when
dealing with the packages. So imho at the validation-step this
package should be rejected with an error. (Same for unknown types -
don't know how this is handled at the moment.)
Stefan