naranjacompleta.com / blog

We finally launched the ETL website. Yesterday I spent some time trying to secure the website. We decided not to allow new members to sign up, and to have all authenticated users work on a secure connection.

We used a wildcard SSL certificate to set up secure sockets with apache and redirect all authenticated requests to https. Here’s a great howto on how to do this.

However for Plone 2.1, modifying the logged_in script no longer works. Instead go to /portal_skins/plone_login/login_next

came_from = REQUEST.get('came_from', None)

if came_from.split(':')[0]=='http':
    came_from = came_from.replace('http:','https:')

I also disabled the join link since we don’t want new users to register. To do this go to /portal_registration on the ZMI, and then uncheck the visible checkbox of the join action.

Then go to / and click the security tab, and look for the Add Portal Member permission. Uncheck acquire and make sure that only Manager is checked in that row. That way only managers can add new users.

After reading plasta88’s comment about my New Year’s resolution, I suddenly decided it’s time to to start posting in this blog.

For a while George and I have been talking about launching a website for the lab, which would encompass the work and research that goes on at the CS department, as well as the weekly meetings we are having.

Deciding on the platform for the website was easy. George officially classifies Plone as software that doesn’t suck, and three years ago I worked on a banking project whose main development platform was Zope. At the time I was very impressed with how powerful Zope and python were. The undo feature, version management, page templates (we really didn’t use DTML that much), collaborative features, and workflow system.

So after some time off from Zope (but not from python), I installed Plone on the ETL server and started hacking away.

And once again I was very impressed. At first I thought customizing Plone’s look would talke a lot more effort (since I’ve seen so many Plone sites with the same look). But after reading Andy McKay’s excellent Definitive Guide to Plone (specifically Chapter 7) and having previous knowledge of TAL, I had a completely customized plone site in a matter of hours. Douglas Bowman’s Sliding Doors technique was especially helpful for customizing the navigation tabs.

I like how Plone’s templating system is composed in layers. Each layer can contain elements, and elements in the upper layer will always override elements in the lower layer (for example, having a ploneCustom.css in a user-defined layer which is on top of all the other layers will always override ploneCustom.css in the plone_styles layer).

CSS files, javascript, portlets, templates can be overriden this way with the very convenient Customize button. Overall customizing the look the way I wanted took around 80% CSS and 20% templates and portlets.
I might do a how-to and a small talk for ETL on this when I’m done with the website.

A link to the website? It will probably launch next Monday, and I’ll announce it on this blog as well.

Well, I am one more of the probably thousands of people whose New Year’s purpose was to start their own blog. Hopefully I will be constant enough to make it worth for someone other than myself to read this.

I am a graduate student in the Computer Science program at Loyola University Chicago. I also work as a research assistant for Dr. Thiruvathukal, the graduate program director at the CS department. Currently we are doing a lot of work on distributed filesystems.

My interests span: python, java, web development, linux, C++, artificial intelligence, databases and distributed systems. I also enjoy philosophy, literature, and photograpy. I’m also really interested in learning about business and finance.

Mostly you will find posts about technology, software development and business, but ocasionally you’ll find random thoughts on Chicago, literature, and other things. Hope you enjoy it as much as I will.