Ok I'm trying
I have
version102: spec
<version: '1.0.2' imports: #('1.0-baseline') >
spec for: #common do: [
spec blessing: #release.
spec package: 'FileSystem'].
(ConfigurationOfFilesystem project version: '1.0.2') load
worked now
(MetacelloToolBox validateConfiguration: ConfigurationOfFilesystem)
inspect
Critical Warning: No version specified for the package 'Filesystem' in
version '1.0.1' { noVersionSpecified } [ #validateVersionSpec: ]
Critical Warning: The name of package 'FileSystem' does not match the
file name 'Filesystem-cwp.63' in version '1.0.1' { packageNameMismatch } [
#validateVersionSpec: ]
Critical Warning: No version specified for the package 'Filesystem' in
version '1.0.2' { noVersionSpecified } [ #validateVersionSpec: ]
I was confused why this is a critical warning?
MetacelloToolBox descriptionForValidationReasonCode: #noVersionSpecified
I got a DNU at:at:
Stef
On Feb 3, 2011, at 10:18 PM, Dale Henrichs wrote:
>
> On Feb 3, 2011, at 1:08 PM, Stéphane Ducasse wrote:
>
>> I tried to publish a new configurationOfFilesystem
>> but after 15 min I stopped because I could not make it load.
>> For me there is something wrong with the interface of metacello because I
>> can never remember it
>> version: '1.0.2' did not load it while it defined as version: #'1.0.2'
>> and of course I tried string and symobols....
>
> Stef,
>
> I recommend that you try to use the validator:
>
> (MetacelloToolBox validateConfiguration: ConfigurationOfFilesystem)
> inspect
>
> The validator should complain about using #'1.0.2' or '1.0.2' incorrectly. If
> the issue printString isn't enough you should be able to get more info from
> the #reasonCode:
>
> MetacelloToolBox descriptionForValidationReasonCode: <reasonCode from
> validation issue>
>
> If the validator doesn't provide a good enough explanation, I'm interested in
> improving it.
>
> If you run into a problem when you reconstruct the configuration and the
> validator doesn't give you good information, then please feel free to ship me
> a copy of the configuration so I can be the human validator - which in turn
> will spur me on to improve the validator:)
>
> I extend this invitation to anyone having trouble with Metacello.
>
> Dale
