Dead Projects

It's quite annoying when you find a nifty open source project only to realize that, upon closer inspection, the project is quite dead.

The first example is OpenRPG. I like that one can create character record sheets which can be cross referenced in dice rolls. OpenRPG's server is a separate component and there are actually many public servers available, which indicates people are still using OpenRPG, even though the project is deader than a parrot in a Monty Python sketch. Some people are using Traipse, a fork that is attempting revive the project, but it may just be nailing the parrot to the perch. I have to play with it a bit more before drawing any conclusions.

The second example is AutoRealm, which I found just today. It looks like a very useful tool for creating maps for role playing games. Not only is the project dead, it was Windows only and written in Pascal, hardly a common language for open source projects. Well, AutoRealm was originally closed source which may explain the unusual choice. A precompiled Windows binary is available which may run in Wine, but I haven't played with Wine in years and wasn't in the mood for that today. Someone started rewriting AutoRealm in Python, but the UI is non functional and the last check in was two years ago. Not a good sign.

A Little Smug

Way back around 1998 or so when I was first deciding on which open source database to use, I choose PostgreSQL on technical merit. Later, as the ramifications of MySQL's dual licensed business model became apparent, I realize that PostgeSQL was a good choice for other non-technical reasons as well.

Also in the late 90s, I first played with Java on Linux. At the time Java on Linux sucked, as Sun did not officially support Linux, and was one big reason I was turn off by Java. My search for a new programming language led me to discover Python around 2000 or so. Again, the choice was influenced by non-technical reasons.

I've never been a fan large application suites because they are the Windows way of doing things. I came to Linux directly from the Amiga where such suites never existed. Consequently I never used OpenOffice and chose individual applications for each function instead. As before, this was not a technical choice but one based on my existing experience.

All the open source software that I avoided in the last decade or so, is now owned by Oracle after they acquired Sun. Oracle seems hell bent on destroying all the goodwill that they acquired along with software. Of course, the open source software will continue without Oracle, but you'll pardon me for feeling a little smug that I avoided this debacle (even if it was mostly by luck ^_^).

Wave Replacement

Google's decision to discontinue Wave caught a lot of people by surprise and left many looking to alternatives. Google intended from the start that it would not be the only provider for Wave so federation was designed into the protocol at an early stage. However most of the alternatives are not quite ready for production use yet.

Novell Pulse may be the best option for enterprise users. For the casual user, there is supposed to be an free version but it is not clear when that will be available. Another option is WaveLook. And no, I'm not referring to the Outlook based client. ^^ They are working on a browser based version but it is not clear if this is just a client or a full web service, or even if it will be free.

In theory it should be possible to run one's own server and still access the federated Wave network. I'm aware of three open source servers but they are all far from complete.

FedOne, Google's reference server, is the most advanced but it still lacks support for bots and persistent storage is not built in. You have to add it yourself and since this is written in Java, that won't be so easy for me. I haven't even been able to build this server yet.

PyGoWave is another open source Wave server. It is written in Python so I could hack on it more easily but the roadblock is that it requires MySQL. Since I already have PostgreSQL running, I really don't want to run another database server. So I would need to modify PyGoWave before I can even try to get it running.

The third option is RubyOnSails. I haven't investigated this one yet, so I can't say any more about it for now. For all I know, it may be the best of the bunch.

And even if I get a Wave server running, there's still the not so minor problem of a Wave client... We live in interesting times, indeed.

Building From Source

Building from source written in a language with which you are not deeply familiar, almost guarantees that the build will fail. Usually in an obscure way. Trust me on this. ^^

Now when this happens, you should ask your self, do I a) figure this out or b) just download an older prebuilt binary, even if it may not be up to date? If you picked a) then you are as crazy as I am. Nice to meet you. ^^

The sane people can leave now.

Billion-Dollar Software Companies

This article is a very realistic assessment of Redhat's growth potential. In order for Redhat to become $5 billion, it needs to displace $50 billion dollars in existing software sale. In other words the software market will actually shrink because open source saves people money. This is good thing and only the opponents of open source could see it as a negative.

Repeat after me, helping people to save money on software is a good thing.

Windows. Sigh

Sometime I wish hadn't done so much Windows development over the years. The problem is the systems are all getting old. The newest is already nine years old. The oldest is now 16 years old. Thankfully, the latter system has seen some improvements apart from OS upgrades, but all of the core functionality is still original.

The main obstacle is cost of course, even in good times never mind a recession. For me there is a bigger problem, however. These are all Microsoft shops and they'll have not yet seen the light of open source. This is a common in the smaller companies that are my usual clientele. Unfortunately, there not much I do except be there to offer advice when it is needed. I never try and persuade anyone to use open source.

So even if these ageing system are replaced, they will use Windows again. Sigh. If I'm lucky, the replacement systems may use open source databases, programing languages, and other bits, but that is the best that I can expect.

Launchpad

I had quick look at Launchpad today and just as quickly abandoned any plans to try it. You need a virtual environment to try this beast because the install modifies the Apache configuration, and expects a virgin PostgreSQL, otherwise the existing databases will be blown away.

Launchpad is seriously heavy duty and has some equally heavy duty dependencies (Zope) so Launchpad is not going to be a trivial web application to get running. It will be interesting to see who gets the first publicly accessible Launchpad running outside of Canonical.

Launchpad has some great features (like tracking bugs across projects) but it is too heavy weight for my tastes.

Canada And Open Source

A recent report ranks Canada's open source activity at number 28 out of the 75 countries that were included in the research. It is quite sad that several third world countries rank higher, but I'm hardly surprised by our mediocre position. The Federal government has only now started to study open source as an option for government projects. While this a good sign, we all know how long these studies can take so it could be at least another 5 years before conditions for open source improve in Canada. And I'm sure the lobbyist from the propriety vendors will be working overtime to ensure it doesn't happen.