Skip to main content

JSF 2.0 Endgame

Posted by edburns on October 21, 2008 at 10:48 AM PDT

Issues before sorting

Before sorting

Issues after sorting

After sorting

The JSF 2.0 project has reached the culmination of the href="http://weblogs.java.net/blog/edburns/archive/2008/08/want_to_help_wi.html">XSLT
contest and the href="http://weblogs.java.net/blog/edburns/archive/2008/08/jsf_20_eg_issue.html">scheduling
exercises the JSF 2.0 Expert Group has been doing since August. I
took the 88 issues in the href="https://javaserverfaces-spec-public.dev.java.net/issues/buglist.cgi?component=javaserverfaces-spec-public&issue_status=UNCONFIRMED&issue_status=NEW&issue_status=STARTED&issue_status=REOPENED&target_milestone=2.0&target_milestone=2.0_ea1&target_milestone=2.0_ea2&target_milestone=2.0_ea3&target_milestone=2.0_pd1&target_milestone=2.0_pd2&target_milestone=2.0_pd3&target_milestone=2.0_rc1&target_milestone=2.0_rc2&target_milestone=2.0_rc3&email1=&emailtype1=exact&emailassigned_to1=1&email2=&emailtype2=exact&emailreporter2=1&issueidtype=include&issue_id=&changedin=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&short_desc=&short_desc_type=fulltext&long_desc=&long_desc_type=fulltext&issue_file_loc=&issue_file_loc_type=fulltext&status_whiteboard=&status_whiteboard_type=fulltext&field0-0-0=status_whiteboard&type0-0-0=substring&value0-0-0=EGTop5&field0-0-1=status_whiteboard&type0-0-1=substring&value0-0-1=EGEasy5&cmdtype=doit&namedcmd=IssuesResolvedEDRtoPR&newqueryname=&order=Reuse+same+sort+as+last+time&Submit+query=Submit+query">query
of issues we declared we’d try to do in the remainder of JSF
2.0, and used href="http://weblogs.java.net/blog/edburns/archive/2008/08/and_the_winner.html">qoob’s
XSLT sheet to produce an 88 page document, which I printed at Kinkos
($9.42 expense report forthcoming). I laid them all out in my living
room, one deep, as shown in the “before” picture.

I walked around the issues and sorted them into 22 categories, see
the “after” picture. The categories generally represent
functional areas, with a few exceptions that can be considered href="http://en.wikipedia.org/wiki/Cross-cutting_concern">cross-cutting
concerns. The cross-cutting concerns include: javadocs, attributes
and properties, class hierarchy, and value handling. The number of
issues in each category is visually represented by the size of the pile.
You can see that “Attributes and Properties” and
“Validation and Conversion” have a lot of issues. I color
coded the categories as well. Categories in style="color:red;">red generally contain issues relating to the
original charter of the JSR, and its href="https://javaserverfaces-spec-public.dev.java.net/#EDR_goals">Early
Draft Review Goals. Categories in style="color:blue;">blue generally contain issues I think are
important to improving the developer experience of using JSF to build
solutions. Categories in yellow
generally contain issues of importance to EG members and/or the
companies they represent. Categories in black contain miscellaneous
issues the EG and the public have determined are important to try to
address in JSF 2.0.

Even with all this, no issue tracking system, no matter how awesome,
is immune to href="http://en.wikipedia.org/wiki/Garbage_in_garbage_out">GIGO.
Therefore, it’s important that we use these 22 categories to
ensure that the major tasks that remain to be done are captured. On the
JSF IRC
channel
, we had a meeting yesterday where we tried to reduce the
level of garbage going into the list and came up with href="http://wiki.java.net/bin/view/Projects/Jsf2EndGameCategories">this
wiki page. This page, and the query of href="https://javaserverfaces-spec-public.dev.java.net/issues/buglist.cgi?component=javaserverfaces-spec-public&issue_status=UNCONFIRMED&issue_status=NEW&issue_status=STARTED&issue_status=REOPENED&target_milestone=2.0&target_milestone=2.0_ea1&target_milestone=2.0_ea2&target_milestone=2.0_ea3&target_milestone=2.0_pd1&target_milestone=2.0_pd2&target_milestone=2.0_pd3&target_milestone=2.0_rc1&target_milestone=2.0_rc2&target_milestone=2.0_rc3&email1=&emailtype1=exact&emailassigned_to1=1&email2=&emailtype2=exact&emailreporter2=1&issueidtype=include&issue_id=&changedin=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&short_desc=&short_desc_type=fulltext&long_desc=&long_desc_type=fulltext&issue_file_loc=&issue_file_loc_type=fulltext&status_whiteboard=&status_whiteboard_type=fulltext&field0-0-0=status_whiteboard&type0-0-0=substring&value0-0-0=EGTop5&field0-0-1=status_whiteboard&type0-0-1=substring&value0-0-1=EGEasy5&cmdtype=doit&namedcmd=IssuesResolvedEDRtoPR&newqueryname=&order=Reuse+same+sort+as+last+time&Submit+query=Submit+query">issues
targeted for 2.0 will be our guide for the remaining seven weeks of
time we have to define JSF 2.0.

To help us out, please, please, please review the href="http://weblogs.java.net/blog/edburns/archive/2008/09/jsf_20_early_dr.html">Early
Draft Review 2 draft of the spec and send comments to <

href="jsr-314-comments@jcp.org">jsr-314-comments@jcp.org
>

Technorati Tags:

Related Topics >>

Comments

Is JSF 2.0 going to favor REQUEST scope more than we currently see with 1.2? Colleagues and I find JSF to be quite geared towards SESSION scope and it causes a lot of issues in regard to life cycles, we almost always need to obtain values directly from the request rather than relying on setters - which has to be there though in a dummy fashion, in order to satisfy JSF. And how come there is no "EMBEDDED" scope, which would work much like SESSION scope but serialize (and encrypt) context out onto an inputHidden field and get deserialized again upon each REQUEST transparently? Anyway, just something we've been wondering about.