Blogging Roller

Hierarchical categories. <a href= "http://sourceforge.net/mailarchive/forum.php?thread_id=1420154&forum_id=9297">Matt mentioned the idea of hierarchical categories on the roller-dev list the other day. Scott Switzer mentioned this <a href= "http://sourceforge.net/mailarchive/forum.php?thread_id=1256114&forum_id=9297">back in October and <a href= "http://opensource.atlassian.com/projects/roller/ViewIssue.jspa?key=ROL-57">put it into JIRA. A hierachical categories feature seems a little complicated and maybe a little confusing. Maybe with the right UI, hierarchical catagories would work. Is it overkill?

Multiple categories. Right now, each weblog entry has one and only one category. Most blogging software allows you to assign multiple categies per entry, I think Roller needs this too.

Selective display of categories. It would be nice if the Roller page macros allowed you to specify which categories are to be displayed on each page and in each newsfeed. That way, you can have a developer oriented page that includes your Java, C#, and AOP categories; and you cab have a family oriented page that includes your Personal, Cat, and Look-At-My-Cute-Little-Baby categories. As Matt suggested, we could also use hierarchical categoes to achieve the same goal.

The French Application-Servers.com website is experimenting with using Roller-based <a href= "http://blogs.application-servers.com">weblogs to complement and possibly to replace message-board based forums, which often erupt into flame wars. Tres cool!

Cool! A new Roller-based weblog, by Matthew Porter. Matthew: you might want to turn off the new-users-allowed setting in web.xml unless you really want to host a Roller community on your site.

I just commited to the Roller 0.9.6 CVS branch a series of changes that should improve the performance of the Roller main page (index.jsp) by making better use of OSCache page caching. I hope that FreeRoller will put these changes in place, because the FreeRoller main page is VERY SLOW! Assuming that the changes work well, I'll label and release Roller 0.9.6.5 on SourceForge sometime next week.

As you may have noticed, Roller 0.9.7 development has been put on the back burner. At the moment, Matt, Lance, and I are all too involved in other side projects to do any significant work on Roller. I'll probably get back into Roller development around Christmas time with some enhancements to Roller's referer tracking.

Doc Searls mentions Roller in a post titled Rolling a new blog on the Linux Journal website.

Roller is part of homework <a href= "http://spectral.mscs.mu.edu/Ent2002/assignment/04.html">assignment #4 in Enterprise Architecture class at Marquette University. Take a look at the "ass4" links on the <a href= "http://spectral.mscs.mu.edu/Ent2002/ClassList.html">ClassList page to see what the students had to say about Roller and weblogging in general.

If you haven't checked the FreeRoller statistics lately, perhaps you should. FreeRoller responded to 702,015 requests for the month of November! I think we should break a million requests per month within a couple of months. [Anthony Eden]

Very cool! The first Roller site (mine) went live on June 8 of this year and now thanks mainly to Anthony there are over 250 Roller-based webloggers out there. If you use FreeRoller, please help Anthony pay for the bandwidth.

The link below showed up in my referer logs today. I sure would like to read that document, but, alas, it appears to be behind the iPlanet firewall. It looks like somebody is suggesting that Roller or some part of the Roller architecture could be used in the iPlanet portal. Way cool!

<a href= "http://ips.red.iplanet.com/%7Echanezon/ipsarch/suggestions_portal_architecture.html" title="...portal_architecture.html">http://ips.red.iplanet.com/~chanezon/ipsarch/suggestions_portal_architecture.html

Anthony applied the Roller 0.9.6.4 build to FreeRoller today. The RSS fix seems to have worked and I am no longer seeing old FreeRoller posts pop up in Aggie. It is nice to see Aggie report the "Channel has not changed since last read" message.

I fixed those two RSS feed bugs that I mentioned on <a href= "http://www.rollerweblogger.org/page/roller/20021121#everything_old_is_new_again">Thursday. This site should be behaving perfectly with RSS aggregators that try to track which articles are new, so please let me know if you see any misbehavior.

No, it's not just you Darren.  I use Aggie and I have noticed that posts from Roller-based blogs often show up as new when they are in fact old.  Others have reported similar problems with FeedReader and now you are seeing this problem with Radio's aggregator.  I'm hoping that by fixing the following two bugs will resolve this problem:

Dominic DaSilva and Mike Cannon Brookes both read my recent Roller persistence proposal as a proposal to escape from the Castor persistence framework.  That is really not really the case.  I want to make it easier for Roller to make full use of Castor and at the same time I want to separate the Roller business logic from from the Castor-based persistence logic.  I don't like the fact that the Roller business tier is littered with calls to Castor and I want to separate concerns.

Why can't Roller make full use of Castor now?  The answer is in the history.  Roller is constrained by it's own EJB legacy.  When I originally wrote Roller back in the Spring of 2001, I implemented the Roller manager objects as EJB Session Beans and the domain objects as EJB Entity Beans.  I quickly learned that EJB Entity Beans are way too heavy.  I started to take advantage of XDoclet's ability to generate light-weight Value Objects.  Eventually, I came up with an architecture where the presentation tier knew nothing about EJB.  The presentation tier only knows about the business tier interfaces and the Value Objects that those interfaces returned or accepted as arguments.  Once I got to that point, I was able to completely swap out EJB and switch over to Castor.

So, Roller uses Castor for all persistence now, but it still uses the EJB Value Objects generated by XDoclet.  To add a new persistent object to Roller, you create an abstract class that extends javax.ejb.EntityBean and you mark that class up with XDoclet @ejb, @castor, and @struts tags.  This new class will not actually be used by Roller at runtime.  It is used at build time as the basis  for XDoclet code generation that generates Roller's Value Objects, Castor mapping file, and Struts form beans.

As I state in the proposal, I'd like to stop using the Value Objects for persistence.  They are just dumb data buckets and we need them to be real business objects.  That said, I hate the idea of hand-maintaining code that was once generated.  I'm conflicted.

Now, on to the other issue: separating the Roller business logic from from the Castor-based persistence logic and possibly attaining persistence layer pluggability.  I feel less strongly about this because 1) if Roller moves from Value Objects to full fledged business objects then much of the business logic can reside in those objects and 2) Castor works pretty well and at this point no other persistence framework offers a compelling reason to switch.

Like Dominic, I'm on a "quest for a good Java/XML O/R mapping framework."  None of the competing persistence frameworks offer me a really compelling reason to do the work required to switch, so I hope to learn to use Castor better and and at the same time try to design away some of that work.  

One other thing.  I tried to get some design assistance from the Castor and XDoclet folks in the past, but I came up short.  If you are an XDoclet/Castor expert, please speak up and help us make Roller a best practices example of XDoclet/Castor usage.  Read the Roller <a href= "http://www.onjava.com/pub/a/onjava/2002/04/17/wblogosj2ee.html">article, read the <a href= "http://article.gmane.org/gmane.comp.java.roller.devel/335">proposal, get on your high horse, stand on your soap box, or climb the stairs into your ivory tower, and tell us what we are doing wrong and how we can do better.

I made a <a href= "http://article.gmane.org/gmane.comp.java.roller.devel/335">proposal to solve Roller's Castor problems and to make our persistence mechanism more modular. I welcome your comments.

I should have mentioned <a href= "http://www.brainopolis.com/roller/page/lance/20021116#spam_via_referer_log">Lance's ignoreHosts setting. He used it to blacklist the two spammers we had this morning.

I've read about it and now I've experienced it myself. I got hit by nasty referer log porn spam today. I just checked a couple blogs over at FreeRoller and they got hit by the very same thing.

I guess I need to get busy again. I'm working on a fix for referer spam in Roller 0.9.7. Roller will check the validity of each referer ensuring that the refering page exists and that it contains a referering link. Roller will also have the ability to ignore URLs that contains certain key words.

UPDATE: today's spam is coming in from a site called voodoomachine and another site whose name is a just little too nasty to mention. It does not appear to be a one time thing, the hits are still coming in from the spam machine. What a drag. If this continues throughout the day, I'll be taking the referer log off of my page template. What a drag.

When I lived in Jamaica, there was one TV station and on this one TV station there was one weekly Reggae show. The show was called Reggae Strong. In the canned intro to the show, the announcer would always explain that "a weekly show about Reggae could be called Reggae Week, but reggae nah weak: Reggae Strong!"

Roller is certainly going strong. At the start of the week I forked 0.9.6 off into it's own branch in case we need to do a bug fix release. Since then, Lance has checked in two great new features: support for <a href= "http://www.brainopolis.com/roller/page/lance/20021109#andy_oliver_wants">comments on weblog entries and a <a href= "http://www.brainopolis.com/roller/page/lance/20021114#s_p_e_l_l">spelling checker for the weblog editor. Great stuff. Matt Raible has kindly opened up a <a href= "http://www.raibledesigns.com/page/rd/20021114#wanna_try_roller">demo account on his Roller install so you can try the new features.

The Roller-driven weblogs over a FreeRoller are also going strong. There are some great weblogs and some real bona fide Java gurus writing over there. They are bringing in lots of readers. Roller is going strong, but I don't think it is strong enough withstand getting slashdotted. Thank goodness Rickard posted his J2EE-vs-DotNet review on a real webserver.

I've been trying to take it easy with Roller, but I have been doing a little work. I did some work reduce the per-session memory usage in Roller and I thought I had eliminated all leakage. Unfortuntely, the leak remains, but it is a much slower leak. I've also been trying to figure out why old stories show up as new in aggregators such as Aggie and Feedreader.

With the cool new comments and spell checking features checked in, I think we need to start thinking about a 0.9.7 release (after we plug the memory leak of course).

I made a couple of changes in the Roller 0.9.7-dev codebase to address memory usage:

1. There is now one and only one instance of the Roller business-tier implementation object RollerImpl instead of one per session.
2. Many calls to getSession(true) were removed and now the RssServlet no longer creates a session.
3. Velocity template caching has been turned back on, but I'm really not sure why leaving it off (apparently) ate so much memory.

#1 and #3 can be easily backported to Roller 0.9.6.X so FreeRoller can take advantage of them. My site now seems to be using memory at a much more reasonable pace now and I've survived 24 hours with max heap size set to -Xmx64m.

Lance checked in support for Weblog Entry Comments and has comments up and running on his site. I've started working on referer trackback for Roller 0.9.7, but apart from that, I'm not sure what is next.

There are lots of little problems with Roller's UI, lots of room for improvements, and lots of missing weblogging features. Make sure the issues that are bugging you get into Roller's JIRA issue tracker. Look at the list of issues that are <a href= "http://opensource.atlassian.com/projects/roller/IssueNavigator.jspa?reset=true&mode=hide&sorter/order=ASC&sorter/field=priority&resolutionIds=-1&pid=10000&fixfor=-1">not yet assigned to a release and vote on the ones that are most important to you. You can also view the <a href= "http://opensource.atlassian.com/projects/roller/BrowseProject.jspa?id=10000&report=popular">currently most popular issues.

Looks like FreeRoller may have been bitten by the OutOfMemory bug. I'm looking for possible leaks and experimenting with ways to reduce per-session memory use.

FreeRoller has been up and running fine for almost a week now, but Matt Raible and I have been having some problems. Both of our Roller sites have been crashing about once a day with OutOfMemory errors. We are using essentially the same codebase as FreeRoller so we are looking for other differences that might explain the crashes.