Just Journal Developer Blog

 I've been working on some changes for the unix client.  It now allows you to change the host (for other justjournal installs), supports subjects, and has better error handling. 

I'm also working on some new features to the blogger api and the normal journal update code.

 Tonight I noticed that there have been delays with the pinging services while posting blog entries.  When I get time, I'd like to work on making that process a background operation on a separate thread. 

This website has been slow lately.  Most of the problem is high amounts of traffic to my other open source project website.  It should slow down in a few days.

I've also got a new version of just journal almost ready to release.  It fixes bugs with RSS feeds, redundant attempts to close connections, etc.  It also add some support for Sphere.  This service is used by many sites to feature blog content related to other stories, etc.  For instance, CNN uses it.  It will only be in effect for public entries and I plan to make it opt in.  

 The links aren't showing up on jj anymore.  I just noticed that.  I'll look into it

I found a bug in the rss feed code that generates from journal entries.

  seems to screw up.  I filled a bug report on SF about it. and will look at it soon.

 New code published.  report any bugs

I've been working on just journal the last few days.  The next software update will improve the site in several ways.

The update journal page will have the following changes:

• Date and Time are separated and now have nice selection menus to make it easier to work with for users wanting to back date entries or post in the future (for some odd reason)  
• The body area was enlarged and the label was removed.  This should help with writing entries.
• The autoformatting and by default newline text is hidden if the FSCK editor is loaded because those options are not needed if the browser supports it.  This should be less confusing.
• Dojo, an ajax friendly library was added which is BSD 3 clause licensed.   

 I've updated the server to use the latest version of Apache Tomcat.  Last time I tried to upgrade, this site experienced a lot of problems.  Please report any unusual behavior to luke@justjournal.com 

So far, so good.  

We've got a new web developer helping us out with the project.  He's been added to source forge and will be working on an improved site design.  I'm trying to coax him into making some themes as well.  :)

I added an ant build script.  I've always just built binaries in Intellij, but it's a real hassle for everyone else.  I'm not that good with ant, so any suggestions for improving the build.xml are greatly appreciated.

The next steps should be to document all the libraries jj is using, and catch up the sql create scripts. 

 Just Journal has a new feature.  There is now a hit counter present on blog pages.  I'll probably allow users to turn this on/off and reset it in the future.  The counter is in the menu at the bottom.  I am considering adding some type of counter to the RSS and ATOM feeds so users can get a rough idea how popular their blog is.  I'm still determining how I'd identify subscribers.

I've also fixed the style sheet for the tag cloud in the menu on blogs.  The font is now smaller for all three sizes and  a typo was corrected in the stylesheet.  This seems to fix rendering in Safari significantly.  I've noticed problems with some themes.  I'm looking into that.

I'm working on a hit counter for blog pages.  Each page would be considered unique, although I've considered adding a page for blog owners to see "total" hits to their blog as well as popular pages.

I'm thinking about making this an optional feature that is turned off by default.  Counting the hits requires two sql queries which is not efficient.  In the classic ASP era, I used to store counters in memory and just dump them when the webapp was shutting down.  I'm a bit fuzzy on persistence in memory with servlets.  

It's important to realize this is a hit counter and not an impression counter.  I don't store cookies to track hits and make no attempt to identify the user's already been there.  I might add logic to avoid counting multiple times for users logged in though.  

I also took the time to rewrite part of the ping code for various blog tracking/searching services.  I'm testing a forth service out and I'm using common code for google and weblogs.  (public entries only)   

Just Journal now has over 620 users. 

 I was recently told that in some cases it's possible to hijack a session from any webapp, and that just journal had a problem with this.  I quickly went to work on this problem.  It has caused problems for big sites like MySpace and Facebook.  

What does session hijacking mean to me?

Session hijacking means stealing your login.  While you're logged into the account, someone could read your private blog entries and post entries to your account.  Anything you can do, they can do to.  They don't know your password, and can only do this while you're logged in.

What users can do to protect themselves:

1. Always use the secure login feature.  (SSL)  This will prevent the first type of attack on your account.  
2. Always log out of just journal when you're done.  Don't just leave the site.

Steps we're taking to minimize this attack

1. A review of just journal's code is pending.  
2. We're probably going to limit special characters allowed in titles of blog entries, journal titles, music, tags, etc further.  We may limit what can be pasted into blog entries as well.  The only other attack possible is stealing a session cookie which requires a cross site scripting (XSS) attack.  That means someone put javascript code on the site and used that to steal your session.

 I've added a Tag Cloud to justjournal.com  You can see the tags that all users use on the site and their popularity.  Eventually, they'll link to blog entry choices.

 I added the new tags code tonight.  I'm still working on it, but it's significantly more useful.  There are still some performance and usability issues I want to work on.

The development version of just journal can now display recent entries with tags filtered by the tag.  I'm working on refining the design and allowing general searches.  Once that's complete, I'll publish the new code. 

http://www.cnn.com/2008/WORLD/europe/07/14/oldest.blogger/index.html?eref=rss_topstories

A 108 year old woman died.  Since 2007, she's been blogging about her entire life.  She also posted video entries on youtube.  It gives new meaning to blogging for life.  Some people live very interesting lives.

 Just Journal uses the FCKeditor for entry.  I updated the version we're using to 2.6.2.  It's been a few years since I updated this.  It now supports safari, and I've created a custom layout.  

I may work on the layout, but users can now see the HTML source view in all browsers that support the editor, change colors, fonts, and other styles, and make alignment changes.  Let me know if you have problems using the new interface.  I'm considering allowing users choice of their editor when logged in.  

This update does not include any backend code, so the tag changes aren't up yet.  Also, the form includes a trackback field, but that's not used yet.  

I've been testing some additions and changes to the tag feature. The listing should be much better in the next version. I want to get one more problem ironed out and then I'll be updating the site.

I also added a GTK interface in cvs for the unix client. It's not the best, but it's usable.

I released a zip file with the source for the just journal server as of yesterday on source forge. It includes the intellij idea project files.