I wonder if we could use annotations to help with the documentation generation? I'm not that famliar with annotations though so maybe I am way off here.
Sean On 2/16/06, Matthias Wessendorf <[EMAIL PROTECTED]> wrote: > Hi Adam, > > that stuff looks really great! > > -Matthias > > On 2/16/06, Adam Winer <[EMAIL PROTECTED]> wrote: > > It's not a TLD file, since TLDs simply don't have any > > of this information. Instead, it's a faces-config file. > > > > Our approach is that TLDs and JSP tags in general > > are secondary artifacts; faces-config is the primary > > artifact. > > > > > > -- Adam > > > > > > > > On 2/16/06, Bill Dudney <[EMAIL PROTECTED]> wrote: > > > That is great Adam, > > > > > > Is the source of the docs in the link below a standard tld file? if > > > so that is great! > > > > > > TTFN, > > > > > > Bill Dudney > > > MyFaces - myfaces.apache.org > > > Wadi - incubator.apache.org/wadi > > > > > > > > > > > > On Feb 16, 2006, at 9:42 AM, Adam Winer wrote: > > > > > > > FWIW, one of the tools that will be coming from the ADF Faces side > > > > of things is a Maven 2 report that goes from our metadata to > > > > tagdocs that are a big improvement over the tlddoc generated > > > > docs. Tlddoc is pretty awful for JSF - every type is String, > > > > nothing is "request time", there's no list of facets or events, > > > > etc. The docs we will generate have a structure more like: > > > > > > > > http://tinyurl.com/bkayl > > > > > > > > I say "will" because I tried rewriting the tool a couple of weeks > > > > ago using the AbstractMultiPageReport base class in Maven 2.0.2 > > > > and ended with conclusion that this bit of Maven 2 is > > > > currently completely broken. Grrrr.... > > > > > > > > -- Adam > > > > > > > > > > > > On 2/15/06, Sean Schofield <[EMAIL PROTECTED]> wrote: > > > >>> I disagree with the removal of the usage section. The usage section > > > >>> shows how to use the component in context which is not always > > > >>> obvious. > > > >>> I don't see a problem with making it optional for trivial > > > >>> components, > > > >>> however. > > > >> > > > >> Yes well some usage sections are better then others. I took a look > > > >> again after reading your comment and some of those are decent. I > > > >> guess we can keep/port the existing ones for now. Perhaps we can > > > >> consider dropping again when we have the simple examples hosted on > > > >> the > > > >> zone. The examples themselves show usage and the source code servlet > > > >> allows you to see the JSF. No sense maintaining two copies at that > > > >> point. > > > >> > > > >>> I'm also not thrilled with the removal of the syntax section, but I > > > >>> agree that the TLD docs could be a substitute. My preference would > > > >>> be to see the syntax section generated from the same source as > > > >>> the TLD > > > >>> docs, but I'm not volunteering to do the work at this time. :) > > > >> > > > >> I agree that the automatic generation would be excellent. I'm not > > > >> volunteering either. The TLD docs are done automatically and I > > > >> bet if > > > >> we looked carefully we would see that many of the components are > > > >> already out of sync since its hard to keep the documentation up to > > > >> date. > > > >> > > > >>> At minimum, the link to the TLD section should point directly to > > > >>> the TLD > > > >>> document for the component in question rather than to the TLD index. > > > >> > > > >> I agree. That was what I was thinking. > > > >> > > > >>> Also, before the syntax section is removed, the TLD docs must be > > > >>> updated to contain the same information. That's not currently the > > > >>> case (I used dataList as a test of this theory). > > > >> > > > >> Good point. This could be done as each component is migrated to APT. > > > >> > > > >>> -Mike > > > >> > > > >> Sean > > > >> > > > > > > > > > > > -- > Matthias Wessendorf > Zülpicher Wall 12, 239 > 50674 Köln > http://www.wessendorf.net > mwessendorf-at-gmail-dot-com >
