Finding and Identifying Content in WCM

[sturner]

Here's an issue I'd like to just post for discussion and to see if anyone else has thought about it.

Part of what we'd like to use Alfresco for is to house what amounts to a knowledge base - a large number of FAQ-type articles each with a fairly simple structure:

- Question
- Answer
- Keyword list
- Category/Topic

We currently have about 1,800 such articles and the number is increasing all the time. With this many items, some problems in managing the content become apparent in Alfresco WCM:

1. It's not easy to find an existing article when you need to make a change. Paging through thousands of documents shouldn't be the only way.

2. When creating a new article, the user must enter a unique name, and there's no easy way for the user to know what names have already been used. In fact, there's no need for the user to have to enter a document name for this type of content.

3. We would really to be able to associate an immutable unique identifier with a document - an identifier other than the document name. This would allow us to change a document's name without breaking web app's reference to the document.

I realize the upcoming search capability may help with the first issue - but only if it's well integrated with the current WCM authoring UI. Will the search feature allow users to say "get me the document with name element = "Dreamweaver" ?

The second part could be addressed by having the ability to let Alfresco auto-generate document names. This could be based on a sequential integer or a guid-type string. A web form could be allowed to have an output-name specifier something like "${autoname}.xml" and the auto-generated name would be placed in the name field when the use first selected "create web content".

The third issue could also be addressed by the availability of a unique-id-generation feature. A unique value could be generated and stored in an "ID" element in the xml file. It would be nice if this element could be made read-only when it appeared on the input forms. We have already experimented with auto-generating ID element values using a custom type & JSP, but we aren't able to make the field read-only on the web forms.

I realize this is all about giving "database" kind of functionality to the WCM - I'll be interested to hear people's thoughts.

Thanks,
Steve

[spust]

Same problem for us.

We are evaluating Alfresco as possible electronic archive system for long time storage of documents like contracts, customer correspondence or invoices with an estimated volume of 200.000 documents. The filenames are often the same, which generates exactly the same problem as described above with the unique name field.

Other Document Management Systems change the filename to a guid and store the original filename in the database (Alfresco also stores the original filename in the DB, but does not change the filename itself if I´m right). This makes it possible to have various files with the same original filename stored in the same content store.

On the other hand would the modification of the filename according to a guid number result in showing crypted filenames when using webdav. This might confuse users searching for their files, and does also implicate to automatically change filenames of files added through webdav protocol.

Furthermore would it be irrelevant for the user to show the document name consisting of a unique id in the content view (icons view & list view), because it would have no connection to the documents metadata. In this case the document name field should be left out of the default view and should be replaced by showing a field like the already existing description field or another field like document category, customer name or anything else that might give an idea of the underlying document (Which might be handled by using different custom view templates).

So far my thoughts about this issue. Thank you for the post Steve.

Simon

[kvc]

1. It's not easy to find an existing article when you need to make a change. Paging through thousands of documents shouldn't be the only way.

Agreed.  We will look to solve this with our upcoming support for search in 2.1.0 this June.

2. When creating a new article, the user must enter a unique name, and there's no easy way for the user to know what names have already been used. In fact, there's no need for the user to have to enter a document name for this type of content.

Agreed.  A couple of potential enhancements here, but today what you can do is two things:  (a) using a dynamic in-line callout to generate a name and use as a default value in the form (b) use this form element in your output path to generate the name of the outputed XML.

Now, you've no way in 2.0.0 or 2.0.1 to specify this auto-generated field as read-only, and so there's a bit of a gap there.  If this soln works for you, this is a minor enhancment we could possibly look into for 2.1.0.

3. We would really to be able to associate an immutable unique identifier with a document - an identifier other than the document name. This would allow us to change a document's name without breaking web app's reference to the document.

Makes sense.  This is something for us to look into.   Also, wrt to broken references, this is an item that we are currently addressing:

http://wiki.alfresco.com/wiki/Link_Management

I realize the upcoming search capability may help with the first issue - but only if it's well integrated with the current WCM authoring UI. Will the search feature allow users to say "get me the document with name element = "Dreamweaver" ?

Looking into …

The second part could be addressed by having the ability to let Alfresco auto-generate document names. This could be based on a sequential integer or a guid-type string. A web form could be allowed to have an output-name specifier something like "${autoname}.xml" and the auto-generated name would be placed in the name field when the use first selected "create web content".

Ah, you were thinking along the same lines as myself above …

The third issue could also be addressed by the availability of a unique-id-generation feature. A unique value could be generated and stored in an "ID" element in the xml file. It would be nice if this element could be made read-only when it appeared on the input forms. We have already experimented with auto-generating ID element values using a custom type & JSP, but we aren't able to make the field read-only on the web forms.

Yup, we need to add a configurable read-only flag.  Will look into for 2.1.
http://issues.alfresco.com/browse/WCM-461

I realize this is all about giving "database" kind of functionality to the WCM - I'll be interested to hear people's thoughts.

All reasonable and all something that should be supported (WCM should be ideal for creating a KB due to its support for XML, code and content, and deployment.


Kevin

[kvc]

Same problem for us.

We are evaluating Alfresco as possible electronic archive system for long time storage of documents like contracts, customer correspondence or invoices with an estimated volume of 200.000 documents. The filenames are often the same, which generates exactly the same problem as described above with the unique name field.

Simon, for your use case you are probably best served using a normal Alfresco space and our web client as a view into that archive.  The earlier case centering around a KB where you want to capture structured XML and potential build and manage an entire self-service app around it (not necessarily using our own OOTB web client) is much more in-line with a Web Project (which introduces notions of structure doc authoring, multi-channel pub, staging, site archival, and deployment).  Not sure if those capabilities add much value in your scenario …

Kevin