openSpeak - openSpeak

I am happy to announce that our little break is over and we’ve started working on openSpeak again. One big help was our new member kubrick who joined us lately and is currently working on an audio mixer. This will be the last big piece of work to be done before we can announce a feature freeze for the upcoming 0.1 release. Everything after that will only contain smaller or bigger bug fixes but should already be quite usable.

Audio is generally the section where the most work has and had to be done. Every other sections are mostly stable and only need smaller fixes.
Other big updates have been done for the documentation. I am currently adding doxygen comments to our source code to ease reading the code and to generate high quality source code documentation. For more informations on doxygen take a look at their homepage. Also I’ve done a overhaul for the wiki, especially the front page and the building instructions separately for Windows and Linux. Shortly after the 0.1 we also hope to be able to release a new homepage design which unifies all three big parts (news, wiki, bugtracker) under one layout. It will also contain our new logo created by CryptoQuick.

If there are some Windows programmers out there reading this, they might also check our bugtracker cause we still got some really annoying Windows related problems. Most members of the team are running linux so we havent got that many chances to test openSpeak on Windows.

Note for French readers : you’ll find news, useful and useless informations about openSpeak on Kubrick’s web site.

Good news, everyone.

Yesterday we made our first successful test with two people on the server (murray and me) and it worked like a charm. There would only be some problems with more than two people at the moment, but I try to get this issue solved fast. When you take a look at our new bugtracker, you can see that most big bugs for the upcoming 0.1 release are fixed and only some smaller, probably easy to fix bugs are left.
So there is only one thing left that keeps us from releasing and thats the relocation of murray (because of that he’ll have no internet) and I got my last exams in a few weeks on which I should learn a little bit. Because of that development will be a little bit slower for the next two or three weeks, but we promise you: There will be a release soon :D
If want to help us while we got no time for openSpeak, get the latest version from svn, build it, tell us what you think about it and maybe write a bug report/future request if you find something.

So long
rmbl

With every day, we’re coming closer to our first release. But while moving forward we’re always spotting new things to fix or to add.

Last week murray finished up the Speex wrapper, so we got a encoding/decoding completely working now. But after testing the client with it for a while we saw that the GUI was heavily lagging. The only fix we could imagine was to add a new thread for all non gui classes. Implementing the thread wasn’t that hard with the help of wxWidgets which provides a great thread class you can easily extend. The only blocking bug we got now, are segmentation faults when executing functions from one thread to another.

After this bug gets fixed release is REALLY coming near :) We expect a first intern release candidat next week (but unexpected things might hold us back). So if you like to test it as one of the first, send us a mail or talk to us on irc or just get the code from svn. I’ll also make a post on the openspeak-release Mailinglist after the rc1 is on svn.

So stay tuned for a few more days/weeks.

After little break, we are going with big steps to our first release 0.1!

We are not sure, when this version will be released, but we hope soon ;)

Perhaps you know, we have had a chat function in our program any longer and we missed a voice input/output function as a soundwrapper. Now, I am proud of to say, that the sound wrapper seems to work. Of course we are missing some fuctions, yet. But the basics, thus to record the voice over a microphone and the playback, works pretty well now in our testprogram.

Also some helpfull functions have been added to the chat and a daemon function for the linux server have been included.

One of our next steps will be the implementation of the soundwrapper into our main program including speex en-/decoding.

Be patient, speak freely.

As stated, here comes the second weekly update :)
This week we made large improvements on both security and rights management.

For security we added a SHA-256 algorithm to encrypt passwords sent over the internet, especially the user passwords. The only unencrypted password on the server is now the servers password if it’s a protected one. We also added several checks to drop corrupt or empty packets to protect the server from crashing. We’re also thinking about implementing an two-way algorithm to protect chat messages or voice transmission later on. That would keep everything secured but if we add it won’t be in our first release version.

On the rights and user management site we added functions to change rights of users on the server. These functions are only available if you got a decent level on the server. What’s still missing is a possibility to add yourself as an owner if you’re the first one to register. This will probably be added soon.

But the most changes where made on the invisible part, the openSpeak lib, and the functions using it. Especially the new packet creation process will speed up our coding as we need to type less chars (lazy coders ;) ).

Thats it for this weeks issue, see you next sunday :D

Welcome to our first entry into the openSpeak development blog. I hope that I or someone else from the team is able to post an update of what we’re currently doing at least once a week. We’ll see if we’re able to do this :)

So lets start with what we’ve done so far. The GUI is working pretty well so far and after adding Sizers to all elements even on Windows. Windows was a little problem cause murray and I are using Linux and got no windows computers to try to compile it and even no experience with Windows API functions to update the GUI, etc. Thanks to Levia this is working now and we solved one major problem on our huge list :)
Another big problem is the Audiowrapper. I took a look at some already available wrappers at the beginning but none seemed to suit our needs, so we probably will write our own little wrapper with all the functionality we need. This will of course limit the platforms openSpeak is able to run on, but thats only a matter of time. We’ll still keep our eyes open and if we spot a good cross-platform wrapper we would probably switch cause this would leave us more time to get the really important things done.

On the networking side we’ve made some huge improvements. I rewrote the ping/timeout code several times during the last weeks and it’s working more stable now. During this work, i also recognized another problem with our dialogs. These paused our own mainloop what would’ve caused the networking to pause. The next step will be some user management like server rights, registration and logging in. The login is already working but still needs some improvement.

At the moment it looks like we might be able to deliver a first stable version in march, but this is just a speculation ;)

Stay tuned until next weeks update!