Happy coding

Now that OpenOffice.org 2 is getting more and more popular, people is starting to migrate and to write documents in ODT format. However, for people who still aren’t able to install OO2 on their machines (outdated distros, outdated hardware or whatever) it’s a pain to receive such documents by email.

It would be nice to have a tool that allows command line document conversion. So, a company can have that tool installed on a centralized host and the users would ssh into it in order to convert documents. to old formats or even to PDF.

Browsing the web, I have found an article telling exactly how to do that. I’ve tried the instructions and all worked fine.

I went beyond and move the macros to a common place (/opt/openoffice.org2.0/presets/basic/ConvertidorPDF), so they can be used by all users without the need to install them in every account. You may need to edit some script.xlb files to make it work, just have a look about how it has been done for the other macros and macro libraries.

In addition to that, I wrote a script and put it in /usr/local/bin, so all users can now convert documents to the format they like without the need to launch any window.

I stopped here, but I’ve great ideas in mind: create a kind of CGI, maybe in PHP, and publish a webpage in our intranet, so everyone could go there, submit his or her document, and receive a PDF or an older format.

During this month I’ve been using my spare time to update my Linux system administration knowlegde, that was a bit outdated. One of the basic things I couldn’t live anymore whithout was usermode linux, so I put myself at work an decided to learn about it.

Next, I’m going to explain the steps taken to put usermode linux working for me. At the end, you will find internet references that would help you in the process. Continue reading Making usermode linux work

This afternoon, listening Dape‘s comments, I’ve been curious about Python and decided to give it a try.

Looking at the official documentation I discovered that it’s much like pseudocode and indent and colons (“:”) play an important role in the language. It hasn’t semicolons (“;”) and follows the rule “a line, a sentence”. It’s a very curious language… and doesn’t seems difficult to learn.

As some of you already know, I like KDE very much, so I decided to research about if a KDE program could be made using Python. I discovered PyKDE, but it isn’t included in Sarge. I found also a tutorial on writing Qt applications, so I give it a try. It seemed interesting. 🙂

Browsing the internet today, I could read a javascript trick that kept my attention. You can type javascript sentences in the location bar and they can have efect on the current document. That is, you can type something like this (literally, in one line):

javascript:document.forms[0].getElementById('myReadOnlyEntry')
.readOnly=false;void(0);

…and get read-write an original read-only entry whose id=”myReadOnlyEntry”. NOTE: the void(0) is required in order to avoid the submission of the form.

The best of all is that you can define a new “bookmark” that doesn’t link to an ordinary URL, but to a javascript sentence one. By this way, you can do changes over the current web page.

Some other curious things you can do using this technique are to replace a previous given function already present in the document. In this example we redefine a form checking function in the current document:

javascript:formCheck=function(){return true;};void(0);

Those discoverings have frightened me. What if a hacker writes a web page, hidden frame or underlying web page that creates a javascript timer and from time to time it tries to read values from input boxes of other opened web pages? Can it be a phishing threat? Is this technique already being using by crackers? It’s frightening… 🙁