CDL Blog – ROSEdu

Hello!

We are Adrian Niţu and Andrei Datcu, two students who with the help and guidelines of our mentor,  Veaceslav Munteanu tried to make our first, small steps into the great world of open  source by working for digiKam project.
What is DigiKam?
DigiKam is a photo editing and management software. It’s motto is “Manage your photographs like a professional with the power of open source”.  Because it’s a very big and important project, digiKam is divided into two parts: digiKam core and KipiPlugins, which is a backbone constantly extended and maintained  thus offering it’s users many features. It uses the following technologies:

• C++
• Qt

Our tasks:
Our first tasks, implied working on SendImages, TimeAdjust  and KmlExport kipi-plugins. Firstly, we had to add some labels and hints in the SendImages plugin, in order to explain unclear functionalities [1]. Our next tasks was to improve the TimeAdjust plugin. We had to add the option for a user to zoom using Ctrl+Mouse Wheel and +/-[2]. Therefore, following the advice of the developers we changed the class used for image display and replaced TimeAdjust’s original class with a common, shared class. This added not only the above functionalities, but also made the code clearer and improved it’s modularity. Also on the TimeAdjust plugin, we added a new feature, consisting of a progress bar to show not only how much of the task was completed, but also that the plugin is still working. The KmlExport plugin had an file convert bug.  Although as simple as a formatting glitch, it took us a while to find exactly where the bug was found.
Our second challenge was to try to port two plugins from one threading architecture to another. ThreadWeaver is becoming a powerfull standard in parallel processing, and the Digikam software collection wanted to take advantage of it’s strenghts. It was quite difficult trying to work for it as it involved rewriting a class entirely, and we were pressured by time as we also had some partial exams during that time. However, shortly before the completion of CDL we found out that somebody else had already completed these tasks, making our progress null and void. This made us sad, but it also served as a reminder that in the Open Source world we are never truly working alone.

Conclusion:
Although we did not manage to finish all our tasks, we now have the taste of open source programming. We have learned a lot of things, especially how to work as a team, for such an important project like digiKam.
During the CDL course   we found out the importance of code understanding and team work and, more important, that weekends spent at school can be fun.
[1]: https://bugs.kde.org/show_bug.cgi?id=268315
[2]: https://bugs.kde.org/show_bug.cgi?id=268317

Hi !

We are Vladimir Cernov, Bogdan Stoian and Szabolcs Orban, three students who tried to improve some aspects of OpenSIPS project, with the help of our mentor, Razvan Crainea.

OpenSIPS (Open SIP Server) is a mature Open Source implementation of a SIP server. OpenSIPS is more than a SIP proxy/router as it includes application-level functionalities. OpenSIPS, as a SIP server, is the core component of any SIP-based VoIP solution. With a very flexible and customizable routing engine,OpenSIPS ‘unifies voice, video, IM and presence services in a highly efficient way, thanks to its scalable (modular) design.

Our objectives were:

• Implementing the XML-RPC protocol
• Adding support for mathematical operations on float numbers
• Implementing support for Sangoma cards

The first task implied the creation of a new module which would add support for MI commands from a remote computer, getting the results in XML format.   The old module used an old http library, which was not suitable for the future needs of OpenSIPS. A similar functionality (which exported results in HTML format) was already implemented, and it was used as an example. However, further work have to be done, as  the HTTP module itself needs to be improved.

The second task supposed creating a mathematical function for handling simple precision floating point expressions. The reason this project would have needed to be completed was reducing the overhead of calculating prices, which would have had a great impact on the community.  There were some difficulties regarding the understanding of the existing code-base, so we need more time to make that task functional.

The third task, implementing support for the Sangoma cards, would have also had great impact, as it would have allowed telephones (or end-points) with incompatible codecs to communicate with each other. The card is responsible for converting codecs, acting as a mediator between the end-points. The project has not been finished, as some technical problems have appeared that we did not manage to solve (the connection with the card could not be made, and we did not get a response from the manufacturers).

Even though the projects were not finished, we consider that working on it was a great learning experience. During our work, we have become familiar with understanding the code of other programmers. Additionally, we have come to acknowledge the real problems and obstacles that may appear during the development of a large-scale project. Furthermore, our team-working skills have undergone great improvement. Besides, we have been exposed to cutting-edge technology (such as Sangoma cards).

The fact that we did not manage to finish our tasks does not pose a problem, as we intend to continue our work after CDL.

Hello!

We are Adrian Bogatu, Cosmin Dragomir and Flavius Anton and together with our mentors, Valentina Manea and Mihai Maruseac, we took part in this year’s CDL course. Our main target was fixing bugs and/or making improvements in Pidgin.

Pidgin is an open source Instant Messaging client that supports a lot of protocols, thus providing access to different networks such as Yahoo, GTalk, MSN, Aim and even Facebook chat all in one application. It has a wide spread area among Linux users. It’s written in C and it uses GTK+ and libpurple libraries.

Because it’s all written in C and because the source code almost completely lacks commentaries, it was very hard to figure out how exactly we’ll make some improvements on the already existing code. We had ideas, we had our mentors’ support, but the implementation was far from easy.

The first thing we accomplished here was Ticket #2497 (Offline message should indicate that it’s a pounce [1]). We were very enthusiastic about our first improvement and we immediately send a patch to Pidgin’s developers, that also got accepted in a short time. It was cool to know that finally you’ve done something that’s visible outside your college and friends, it’s visible to all Pidgin users.

On the other side, the first thing that we tried to accomplish, but we didn’t succeed, was Ticket #121 (expiration date for buddy pounces [2]) which sucked all of us into the depths of the Pidgin pounces source code leaving us no chance to return. Therefore we lost a couple of weeks just reading and trying to understand what those C functions were doing, not being able to make something constructive. After a while we completely gave up and shifted our attention to matters not simpler, but less hard to implement.

Another thing that got our attention was Ticket #352 (Doodle Confirmation Dialog [3]). That implied modifying an already existing plug-in that assured a Doodle connection on Yahoo protocol. Although, at first glance, we thought that’s going to be very easy, we encountered a lot of problems on the way. Now, we had an almost completely working Doodle, but with small bugs here and there. That’s not a very big issue though, because we’ll continue our work, even if CDL ended.

Speaking of CDL ending, we really have to say it was a great experience:

• We had our first encounter with a big open-source project and with its community.
• We learned many new things from the presentations held during the course and now we are more confident about our skills
• We made a lot of new acquaintances that have the same interests.

Besides, one important thing that we learned was that, sometimes, waking up in the morning on Saturdays can’t always be so unpleasant.

Finally, we want to thank again our mentors Valentina Manea and Mihai Maruseac because they helped us a lot after every course, on every hackaton. For now, we’re looking forward to going on a trip with all the people that took part in the project. That is the main reason everyone should enjoy this course – people.
[1]: http://developer.pidgin.im/ticket/2497
[2]: http://developer.pidgin.im/ticket/121
[3]: http://developer.pidgin.im/ticket/352

Introduction

Hello! We are Dragoș Badea, Paul Săpunaru and Mihai Zamfirescu. During the Spring Edition of CDL v5.0, we have been working on the “World of USO for Android” project, alongside with our mentors, Irina Preșa and Alex Eftimie.

Project Description

As you might know, World of USO is a browser-based game dedicated to Computer Science students at University Politehnica of Bucharest. World of USO for Android is a project aimed to port the game to the Android platform, while retaining as many of the original features as possible.

Technologies used:

• Java
• Android SDK
• OAuth
• XML for layouts design
• Eclipse IDE with the Android Development Tools (ADT) plugin
• Git

Requirements for running the application:

• a device running Android 2.3 (Gingerbread) or newer

Story

The project started as quite a challenge, since none of us worked with Git, nor Android, before. One of the greatest difficulties encountered was implementing the client-side OAuth protocol, as part of the authentication process (OAuth is also used by companies like Google or Twitter). After a few weeks, though, we got the hang of it, and things started running smoothly.

Conclusion / Thoughts

Dragoș
How often can one willingly wake up early Saturday morning, without regretting it instantly? Not many times, trust me about it, but this was one of those occasions. The community is one of the main reasons why CDL was such a nice experience; talented and humorous people, making every course both interesting and fun.

Paul
What did the Community and Development Lab mean to me? Well, it was my first
extracurricular activity of this type and I have to say that it was an intriguing experience. I had the pleasure to come in contact with great people willing to share their knowledge and passion for a certain subject and I found each presentation (technical or about social skills) particularly interesting.
Overall, I am really glad that I was accepted to join this fun and supportive community.

Mihai
CDL is many things – you get to learn a lot about Git, Python, Vim, etc. But, what really stands out is the community – you get to play Laser Tag, hack your friends’ project, we’ll even go on a trip soon. And that’s what makes it awesome!

In the end, we would like to thank our mentors, Irina and Alex, for introducing us to this project and helping us throughout the CDL journey (and poke them about that promised beer).

What is Dex Online?

DEX Online is the bigges online Romanian dictionary, one of the most important Romanian websites.

Who are we?

We are Ion Octavian Năstasie and Alexandru Ciprian Fărcășanu and we are thrilled to present to you our achievements, the fruits of our work during CDL 2012 spring.

We received lots and lots of support from our mentor Mihai Bărbulescu and from the founder of DEX Online, Cătălin Frâncu.

What did we use?

We used web technologies like HTML, PHP, Smarty templates, CSS and JavaScript and MYSQL is used as the database engine.

Our contribution

The problem that caught our attention was Ticket #257, Linguistic games, we felt is was important for DEX Online to retain visitors for a longer time, to make the website more than a place where people searched for the meaning of a word. We wanted the visitors to engage more with the website and to teach them new things in a fun and interactive way. We choosed to do a Hangman game, because we felt it was an iconic game that would make great use of DEX Online’s database.

First we had o get ourselves familiar with the code, smarty, and the database. Neither of us had to much experience working with web technologies, so this was the hardest part, figuring out the best way to do the things we needed for our game. We had lots of help from Google, DEX’s source code and of course our mentor.

When we felt we had a decent game we asked people to test Hangman and give us feedback. we got a lot of constructive input from our fellow CDL coleagues (Android guys ) and we made improvements. we pushed our code upstream for review and in the review process we saw that there were things that we didn’t think of and that there were better ways to do some things. It was an interesting experience.

Conclusion

It has been a great time, and we have learned a lot of new and great things, especially the process of writing code for a large project.

We have also met interesting people, but the most important thing is that we made our contibution to an open source project that is used by milions of people and we hope they will enjoy our game.

The only downside  is that it is over, we still have lots of ideas and improvements for Hangman 1.X, but, well, I guess there is always a Hackday for this.