Right now I’m more active on clatko.tumblr.com.

Update 6/27: I just received an email welcoming me back to Facebook. I have not done anything to initiate this, time to change the password and sanitize the account. Wow. And here is my Facebook account. Please comment if any of my privacy is showing.

It’s been a month and a half since I killed my Facebook account. Can’t really say I’ve missed it, but at the same time I feel like I’m missing out on the party. There are new things on the site, Q&A, videos, new iPhone app, location is coming, blah, blah that I want to check out but can’t.

My entire line of reasoning for leaving can be condensed into one statement (paraphrased from a conversation about Facebook owning FriendFeed):

With FriendFeed, you have to work to add information to your activity stream; with Facebook it’s the opposite, you have to work to remove information from your activity stream.

With that, I’ll be back on Facebook shortly with the following conditions:

• My account will act more as a pointer to my main identity - www.latko.org
• Personal information will kept at a minimum
• Access denied to any and all apps
• Never use Facebook Connect
• Opt out of instant personalization
• Never use the status update

Udated: 6/14 (added views on privacy changes, updated privacy settings)
Update: 5/13 (added analysis link, other links)
Update: 5/12

Many, many reasons for doing so. I’ll update this post later with a FAQ on why. This is going to be an extremely long post and will be updated quite frequently. I’ll probably make it sticky.

1) Why?
I want to control my own online identity, not leave it to a private company with questionable ethics that is accountable to very few. I also don’t have confidence in the technical/security aspects of the site. This has little to do with privacy and I’ll even post links to embarrassing photos.

2) Privacy is dead. Anyway, you have almost 200,000 Twitter followers.
Again, this has little to do with privacy. The ask-for-forgiveness later model of the Facebook system is too much for me. The only thing I care about on Facebook getting out is my email address. And we have already seen this info provided and exposed through two XSS attacks on Yelp.

I’m not going to pin this on Facebook, but the day I deactivated my account, I began to receive hundreds of spam messages a day. I think this has more to do with the poor security of the Yelp site.

3) But they upped your privacy capabilities and made them easier to control.
True, I can now hide my friends and my interests. This move does very little to actually prevent that information from being displayed (you are still listed as a member of the interest group and you are still listed from your more privacy-challenged friends). The simpler controls actually sound nice and I can blow away permissions for the 100+ apps in one fell swoop. Nice. I am contemplating creating a new, bare-bones account that will redirect people to my website. I will continue to hold out and see if there are any more slip-ups.

Bottom Line: Facebook does not respect its users.
Not just with privacy, they have a cavalier attitude that they are smarter than the average user (which is probably true given the Read Write Web spectacle). Here are a few examples:

1. Embedding originating IP addresses in emails.
2. Excessively changing privacy controls.
3. Opt-out schemes on many new features (see below analysis).
4. Tricking users into oversharing (see below analysis).
5. Using circular logic to back up the default-to-everyone.
6. Turning a blind eye to the social game offer schemes.
7. Extracting usurious fees from profitable partners (Zynga).
8. Claiming to be open by embracing and extending open source projects and protocols (see my Salmon post).

And I could go on for a while. I have seen the future and it scares me.

Read a great analysis of the recent facebook privacy changes.

When I first got onto Google Buzz, I quickly learned that it is far too easy to cross social Activity Streams. I have accounts on FriendFeed, Cliqset, and others, that aggregate your various social networks into a single activity stream. So this becomes a problem if you add cliqset and your twitter account to Buzz - content is duplicated. And with multiple streams, it can get much worse, fast.

Another problem with these streams is that sometimes they feed into each other to create a river of data. The problem here is finding a crossed stream within that river. Or pulling a single stream out when you don’t even know where it originated.

There has been much talk about the appropriately named Salmon Protocol which allows for comments on your activity to swim upstream to the originating source. This would effectively de-aggregate your content so that conversations can occur in their original silo. Much like permalinks are used when you blast your RSS feed all over the place. Though, this becomes much more important with activity.

Facebook’s Open Graph is pretty much an implementation of Salmon via widgets and toolbars. It’s surprising to me that people in the Salmon circles are calling for OpenLike and OpenDislike when those are just a subset of Salmon. If we could just integrate OAuth and Salmon, there is your viable, open-source competitive framework to Facebook.

To me, it just seems that Facebook is closely watching these Open protocols and pulling an Embrace and Extend on them. Nothing from the Open Graph is new or unique, but people are praising the “visionaries” at Facebook for making a play to take over the web. Wake up people, we’ve already had these tools in front of us for months.

I really expected Google to make this move with Buzz and am shocked that the rug was pulled out from under them. Their focus on the mobile space may be blinding them from the threat to their bread & butter, search.

image/salmon-protocol.org (you guys really need a new logo!)

Update: I’m back. Thinking this over, I want a clearer delineation between my tech life and my personal life. This blog is all about tech so will remain that way, while Tumblr will be for my personal stuff.

I’m going to take a break from this blog for a while (and twitter to a lesser extent) and move over to Tumblr, which seems like a more ideal fit for me. John Mayer, who I’m not particularly a fan of, saw the light the day after I did and wrote an interesting piece on it. I reblogged it, (that word needs to change), so you can read it in all its glory over there.

I have my opinions on the topic as well and as an extra bonus, you can see a pic of what I really look like (hah).

I’ll continue this site with tutorials and Mozilla builds, but my rants will be over on Tumblr.

So Follow Me On Tumblr!

image/flickr someone or EAVB_IVVHXTHOJJ

Ok, first the robots have something new to say

Robots have shiny metal posteriors which should not be bitten.

Now, color me confused. I’m confused by the new Mozilla naming conventions (not about metal butts). I had always believed Lorentz was a methodology rather than a physical build. Not true. Mike Beltzner said this to CNet a while back:

Next on Mozilla’s agenda is an update to Firefox 3.6 code-named Lorentz, a release that embodies a new attempt to speed up the frequency of Firefox updates.

So Lorentz is actually a release that embodies this new methodology. Oh.

Lorentz is also the first step on the road to Electrolysis. There is also JaegerMonkey, which combines the best from Apple (SquirrelFish), Adobe (nanojit), and Mozilla (TraceMonkey). I guess V8 was left out of the party. Ars, as usual, has a good article on this. I do have a 64-bit build of Electrolysis (let me know if you want it).

So now we have Firefox 3.6.3 Lorentz, Firefox 3.6.3 JaegerMonkey, Firefox 3.7 Electrolysis, etc. It makes sense from a developers perspective to separate the projects like this, especially if you’re trying to add new features into minor updates, but I don’t think it has been explained well enough to the mainstream.

I don’t think my explanation is very good either. Is there a definitive document on this methodology?

You can download the Lorentz beta on the Downloads page.

Follow me on twitter!

Update: This has nothing to do with Apple as this API was signaled for termination long ago.

Just had a hell of a time compiling Firefox 3.6.2 against the 10.6 SDK.

First, I forget to patch the sqlite3 bug (not sure why this is causing me problems as it was closed as fixed). Then after patching, a make clean didn’t get rid of enough cruft to let the build complete.

So starting from scratch and I get a new error I’ve never seen:

In file included from /Users/beetle/src/mozilla/toolkit/xre/nsSigHandlers.cpp:58:
/Developer/SDKs/MacOSX10.6.sdk/usr/include/ucontext.h:42:2: error: #error ucontext routines are deprecated, and require _XOPEN_SOURCE to be defined

And looking in /usr/include/ucontext.h, you see:

#else /* !_XOPEN_SOURCE */
#error ucontext routines are deprecated, and require _XOPEN_SOURCE to be defined
#endif /* _XOPEN_SOURCE */

This file has not been touched since Sep 30 of last year, so WTF. I usually do all my builds against 10.6, but someone asked me to do a 10.5 build so this issue JUST came up as of Firefox 3.6.2 (does not exist in 3.7, yet). Digging and diving a tad, I find that accelerometer controls are baked into 3.6.2 so this is probably what’s causing the problem.

Anyway, to fix this change the ucontext.h include to sys/ucontext.h like this (appears on lines 57-59 of toolkit/xre/nsSigHandlers.cpp):

#if defined(XP_MACOSX)
#include <ucontext.h>
#endif

to

#if defined(XP_MACOSX)
#include <sys/ucontext.h>
#endif

Things should be then be smashing, in the Austin Powers kind of way.

So who’s to blame? Mozilla takes the fall on this one, though this bug has already been fixed.

I’m on the bleeding edge here using not-publicly-available Apple dev tools (I’m in the Apple dev program, can’t say anything, NDA, sshhh).

I could make this a list of a hundred things, but these three are really getting to me:

1) Get Straight Your Bulk Rules
Sites like Twitter Karma do not allow you to bulk unfollow because it’s against your terms of service, but they are allowed to bulk block! What do you think people are going to do? A) Click on each person, use the pulldown menu, unfollow or B) Bulk block. This does not make sense.

2) State Your Following Rules
The explanation you give for why you “are not allowed to follow more people at this time” is absurd. There is a link to Follow Limits And Best Practices Page, but it’s busted. And anyway, the rules have changed again over the past week and I’m yet to pin it down. The 24hr/1,000/1.1 ratio follow limit is no longer valid. Why do we have to guess at these things?

3) Delete Accounts
If an account is deleted, umm, I probably don’t want to follow it. I also don’t want to see it following other people (check list followers for an example).

Which brings me to another point….

Most people don’t know how to use lists. I get put on these follow-back lists, people follow the list, then message me about why I didn’t follow them back. You’re doing it wrong!

I’ve been a long time fan of The Planet, the “Global IT Hosting Leader”, because they’ve always had outstanding service. They aren’t the cheapest out there, but in my experience they’re the most reliable.

Just one quick story, then I’ll get to the point. A couple years ago, a Sarasota news channel and paper were going to sit down with our startup and do an interview. The interview was scheduled for some time before I’m usually awake, no big deal, but on the day of, I get an escalation call from The Planet saying out main webserver is down. I immediately learn that our server had literally caught on fire and we were supposed to do this big interview where we planned to demo our website, IN HALF AN HOUR. To make this story short, the technicians were able to salvage our drives, swap chassis, and get new hardware back online within 15 minutes. We were able to demo our website just fine during the interview. It was an amazing coordination of effort that come together perfectly at the end.

Well, in setting up my next company, I looked to The Planet. By following their twitter stream, I found out about The Planet Sand Castle. In their words,

The first major hurdle for a software developer is finding the online space for new ideas to grow. The costs of rolling out and testing new software can be prohibitive, especially in uncertain economic times. The Planet is offering the developer community an economic stimulus of sorts to transform dreams into realities: The Planet Sand Castle.

I applied and within a week, I was in. I was initially given the mid level server, then at the last minute they bumped me up to the highest end server under the program (about a $500/mo. deal). Free! No strings attached! For a year!

The Planet also did something else pretty amazing under this program, they gave Rice University 50 servers for entrepreneurial research and development. Since I was one of the first into the program, I was asked for a quote for an article: Hosting Company Offers Free Servers to Developers.

I’m not a paid shill for The Planet, I actually purchased a year’s worth of service on my own dime for a different server days before all this happened (they wouldn’t even go beyond the lousy 6% yearly discount. I asked for a free t-shirt and I got a “lol” over the online support). I just think they offer superior services and I’m willing to pay for that.