[ 
https://issues.apache.org/jira/browse/TAP5-423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12693440#action_12693440
 ] 

manuel aldana commented on TAP5-423:
------------------------------------


I had a look at other asset implementations. The client url for classpath asset 
has a randomly generated number, which changes after each restart of the 
server, e.g.:
/testTAP/assets/app/ee30f2419b002faf/pages/classpathAsset.txt

What is the reason to use such a random number? I would understand this if it 
was a checksum for the resource itself, so it can be invalidated. For search 
engines this is not nice either to have changing links for same content all the 
time. Further more client-side caching won't work after each restart of server.

For my side I would tend to omit this random number for filesystem assets.

What do you think?

> Implement a file-system Asset type
> ----------------------------------
>
>                 Key: TAP5-423
>                 URL: https://issues.apache.org/jira/browse/TAP5-423
>             Project: Tapestry 5
>          Issue Type: New Feature
>          Components: tapestry-core
>    Affects Versions: 5.0.18
>            Reporter: manuel aldana
>
> currenty it is built in to retrieve assets from classpath and from webcontext.
> a file system asset type which can retrieve stuff directly from the 
> filesystem (e.g. from /srv/statics/xxxx) makes also sense. 
> something like:
> *ResourceFileSystem extends AbstractResource* and *FileSystemAssetFactory 
> extends AssetFactory*

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to