The following comment has been added to this issue:
Author: Brian Ewins
Created: Wed, 6 Aug 2003 5:30 AM
Body:
>I too would like to see more documentation on how dependencies are >resolved. If we
>see a directory structure on one of the urls in our >maven.repo.remote list, what
>"exactly" is going on with that?
[answer to the original question, which should be in main docs]
maven.repo.remote is a comma separated list or URLs, and a download attempt for a
dependency is made for each of the URLs in turn:
${url}/${groupId}/${type}s/${artifactId}-${version}.${type}
Note the extra slashes. A common problem is specifying a remote repository with a
trailing slash, and attempting to download (eg)
http://www.ibiblio.org/maven//ant/jars/ant-1.5.1.jar
as a result (see second double slash, which causes download to fail)
[/answer]
Most likely this remains undocumented because maven is supposed to be relatively
independent of repository structure, so you shouldn't rely on the structure remaining
the same - though there is no way to plug in non-url-based repositories right now, and
the current structure has first-mover advantage.
Having to derive all kinds of id information from urls which you have to know in the
first place in order for maven to figure out the url again is a bit of a weakness in
maven...
>Is it possible to have this kind of dependency download be
>smart enough to check for the latest timestamped version.
This is what snapshot dependencies are for.
>Is it possible for it to download the entire directory of jars, >instead of just one
No. To do this would require that the download client was smart enough to understand
how to list directories across all url schemes - http, https, file, ftp, ... for
starters the http directory listings are only possible with the PROPFIND DAV
extension. Additional metadata to support this could be stored in the repository, but
previous consensus has been against this because it would lead to problems with edit
conflicts. Similarly, downloading '*.jar' requires '*' to be expanded on the server.
I'm only commenting here because you asked multiple questions here, but questions (as
opposed to bug reports, requests for improvements etc) would be better off posted to
the mailing lists, most of this is off-topic for the original bug.
---------------------------------------------------------------------
View the issue:
http://jira.codehaus.org/secure/ViewIssue.jspa?key=MAVEN-637
Here is an overview of the issue:
---------------------------------------------------------------------
Key: MAVEN-637
Summary: usage of groupId and artifactId is not properly documented on website
Type: Bug
Status: Unassigned
Priority: Major
Time Spent: Unknown
Remaining: 4 hours
Project: maven
Components:
documentation
Fix Fors:
1.0-rc1
Assignee:
Reporter: Oliver Noelle
Created: Thu, 31 Jul 2003 7:46 AM
Updated: Thu, 31 Jul 2003 9:19 PM
Description:
The pattern to use groupId / artifactId for referencing dependencies is not documented
properly (especially on the website about project descriptor).
Looks like a very basic thing to me, which should not be extracted from the mailing
list or wiki...
---------------------------------------------------------------------
JIRA INFORMATION:
This message is automatically generated by JIRA.
If you think it was sent incorrectly contact one of the administrators:
http://jira.codehaus.org/secure/Administrators.jspa
If you want more information on JIRA, or have a bug to report see:
http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
