Berkeley XML Database

Next topic

Author	Message
Tom C	Friday 11 February 2005 2:07:13 pm Has anyone tried integrating eZ publish with a berkeley database? Does anyone see potential for their XML database as an open source alternative to mysql or postgre, particularly for legal firms and other clients who may require feature-rich xml-based, structured knowledge and document management? (see www.sleepycat.com) Is there any chance that mysql will develop their own XML database? These are key questions for my project, since if we have to redesign, the last thing I want to do is migrate from one database format to another. Additionally, we may wish to use a DTD that is not native to eZ publish for xml data storage or at least for export or transformation for publishing. I'm not sure that's possible. These xml capabilities were the key attraction for me to the Apache Lenya CMS (lenya.apache.org), but I'm hoping eZ publish can be made to do the job. Then, I would gain all the other benefits of eZ publish. Lenya wouldn't work at all for me right now. Another way to get at these questions is to ask, how easily could eZ publish be modified to mimic the xml pipeline architecture of the Apache Cocoon framework or use an equally-effective model? I'm heartened that eZ is gaining functionality to import/export to OpenOffice. This may satisfy our needs, although the industry sector-specific DTD is still an issue. Thanks, Tom
Gabriel Ambuehl	Saturday 12 February 2005 1:25:33 am Personally, I'm not much of a believer in the whole XML DB thing. The concept just seems weird to me. I know there are uses when it makes sense, but all in all, it's really not a general purpose solution. I think Coocoon both gains and suffers from the XML pipe. While very flexible, it's also a pretty slow way to go about assembling pages. If you simply need to import/export a custom DTD, that's probably more within the league of a custom XSLT template, I guess. If it's just one not overly complex DTD you might be able to simply write an import to ezcontentobjects anyhow. Else I'd probably store the XML as you got it and filter it through XSLT upon display. Visit http://triligon.org

Author

Message

Friday 11 February 2005 2:07:13 pm

Has anyone tried integrating eZ publish with a berkeley database?

Does anyone see potential for their XML database as an open source alternative to mysql or postgre, particularly for legal firms and other clients who may require feature-rich xml-based, structured knowledge and document management? (see www.sleepycat.com)

Is there any chance that mysql will develop their own XML database?

These are key questions for my project, since if we have to redesign, the last thing I want to do is migrate from one database format to another. Additionally, we may wish to use a DTD that is not native to eZ publish for xml data storage or at least for export or transformation for publishing. I'm not sure that's possible.

These xml capabilities were the key attraction for me to the Apache Lenya CMS (lenya.apache.org), but I'm hoping eZ publish can be made to do the job. Then, I would gain all the other benefits of eZ publish. Lenya wouldn't work at all for me right now.

Another way to get at these questions is to ask, how easily could eZ publish be modified to mimic the xml pipeline architecture of the Apache Cocoon framework or use an equally-effective model? I'm heartened that eZ is gaining functionality to import/export to OpenOffice. This may satisfy our needs, although the industry sector-specific DTD is still an issue.

Thanks,

Tom

Gabriel Ambuehl

Saturday 12 February 2005 1:25:33 am

Personally, I'm not much of a believer in the whole XML DB thing. The concept just seems weird to me. I know there are uses when it makes sense, but all in all, it's really not a general purpose solution.

I think Coocoon both gains and suffers from the XML pipe. While very flexible, it's also a pretty slow way to go about assembling pages. If you simply need to import/export a custom DTD, that's probably more within the league of a custom XSLT template, I guess. If it's just one not overly complex DTD you might be able to simply write an import to ezcontentobjects anyhow. Else I'd probably store the XML as you got it and filter it through XSLT upon display.

Visit http://triligon.org