?

Log in

Previous Entry | Next Entry

So it seems that all these new services (hating only since I use them) like Twitter, Dopplr, Highrise, Digg, etc all do 95% of what I want and often 10% of what I don't. The problem is that none of them have open code bases, meaning that I'm stuck between getting new features as they role them out, not using them, or rolling my own. In the end, I'm unable (being the good geek that I am) to download the source and submit a patch to make it have the missing piece that I want. In most cases when I go and talk to the owners of these sites, everything I want to see is already on their road map, just not prioritized to happen *now*.

So what happened between the older services like LiveJournal and Slashdot, each based on Open Source cores, to these Web 2.0 services which have shut their repos? Come on guys, anyone can write the code to do what you've done, but it is the overall service which is hard to duplicate and where you should be differentiating yourselves. An open code base only benefits everyone, including yourselves!

Tags:

Comments

( 11 comments — Leave a comment )
mart
Jul. 7th, 2007 06:09 pm (UTC)

I don't think open source is really that important for network-based apps. What's important is open standards and protocols that allow you to combine a bunch of services that do one thing well to do what you like. Some of those services might even be things you've written yourself.

daveman692
Jul. 7th, 2007 06:30 pm (UTC)
I buy that argument to a certain extent, but why should I need to go write a second app if I want to for example show current weather next to my upcoming trips in Dopplr or be able to classify someone's mobile number as a "work mobile" in HighRise? Those aren't things I can do, usefully, via APIs.
mart
Jul. 8th, 2007 09:39 am (UTC)

Even if (for example) Dopplr were open source, you'd still have to convince them that showing the weather is a good idea. Unlike desktop apps where you can run your own copy if you like, you share network apps with other users and so local customizations make less sense.

However, if Dopplr gave API access to your upcoming trips it'd be easy for you to create a separate app which does what you describe without affecting any other Dopplr users and without Dopplr's approval.

daveman692
Jul. 8th, 2007 10:26 am (UTC)
Of course, though I'm not arguing that service owners should accept every patch or feature proposed. I think it is a fairly common case, and certainly is true for this feature example with Dopplr, that it is a lack of time versus lack of desire for many features.

Dopplr does provide an API, but why would I want to create a clone of a large percentage of Dopplr just to add this one small feature? I mean certainly GreaseMonkey is one route which is more middle-ground, but still not ideal.

I'm not trying to argue that an open code base solves all problems, just that there are many cases where a desired feature isn't added since the developer team doesn't have time. This is where if the code base is open, users who feel strongly can take it upon themselves to contribute.
(Deleted comment)
stuffyouknow
Jul. 7th, 2007 08:51 pm (UTC)
An open code base only benefits everyone, including yourselves!
That's what tech geeks think, not what BAs and VCs think.
daveman692
Jul. 7th, 2007 08:55 pm (UTC)
I agree with that, but how many things come together with BAs and VCs saying "this would be a good business lets go find tech geeks" versus the other way around? I would argue that most services are started by the geeks who have a problem that they want to solve; though not saying that the opposite doesn't happen too. Sure the business side always plays into decisions (can I get paid for building this), but the burden should be on the geeks to explain to the business people why open code bases can be beneficial.

Now don't get me wrong, I'm not arguing that every line of code for every service should be Open Source. Just that when the core of a service is, it can allow others to submit patches while balancing that with your own private secret sauce.
anildash
Jul. 8th, 2007 05:11 am (UTC)
An open code base only benefits everyone, including yourselves!

For somebody like Rupert Murdoch, I think he sees "benefits everyone" as a *negative*, compared to "only benefits MEEE".

And if you look at the numbers on Facebook or some of the others, how much leverage did they get from LJ making all of its code open? I mean, I think LJ's easily generated a billion dollars in value for other companies. That's actually one of the things that makes me happy about 6A, everyone sees those other successes as a *good* thing.
desh
Jul. 8th, 2007 05:05 am (UTC)
Insecurity? With LJ, Brad knew that he knew what he was doing, and that he could do it again if necessary. So opening things up just makes the code better. On the other hand, maybe the Twitter folks just stumbled across something popular, and they're scared that if they share it, they'll lose it, and along with it their only chance to be in the Web 2.0 A-list.

But it's probably not that. It's probably just the money.
ext_56311
Jul. 20th, 2007 08:59 pm (UTC)
Tasty code
I love fixing other people's code too so I can get features I want.

With webapps, its still awesome that someone else is paying for hosting and maintenance; so adding a few features might be a way to extend an app on a limited development budget.
codetoad
Aug. 9th, 2007 10:46 pm (UTC)
I'm with you 100%. I don't understand why so many geeks tolerate this stuff, too. Somehow lock-in and closed source services are cool?
( 11 comments — Leave a comment )