Yesterday I had to debug a Mac OSX program and therefore I had to learn about how Mac Universal Binaries work.

Universal Binaries contain (most of the time) PowerPC code and Intel x86 code as well. While they could easily also feature code for other architectures, like your toaster or microwave, it usually looks somewhat like this: [Header|PowerPC|Intel]. The operating system decides which code it needs and executes the part of the file suitable for its architecture.

Now in order to see what kind of binary information is "sandwiched" in the file, there is a header, very nicely described in the Mac OS X ABI Mach-O File Format Reference.

It starts off with a "magical number", that funnily reads 0xCAFEBABE in hexadecimal. -- Cafe babe? Yup.

Looks like the programmers had some fun coming up with a readable magic number. :)

For a more in-depth explanation of how universal binaries look like, I recommend this blog post.

(thanks to the t-shirt model ;) )

Read more…

About a week ago, Skype came up with its beta version 2.5 for Mac, so today I gave it a shot. Ambitiously, I dared to call into a meeting, where I was expected to enter the access code using my phone's number pad (DTMF dialing is what they call it, I think).

Of course, it did not work: The numbers I dialed were not understood and after a few seconds, the computer on the other side cut off my connection.

Sadly, this funcionality is meanwhile needed for a vast majority of automatic caller systems, for banks, insurances, other businesses, even the government. And, the respective Skype bug was therefore already found a long time ago, one of the first forum topics that I could find on the issue is dated in April 2005.

By the way, on Windows this feature allegedly works without problems, so this issue is not related to the "Skype Out provider issue" they once had.

Seeing that long-known problems like these are simply ignored does not make me feel excited about their new features like SMS, the webcam stuff etc. anymore.

Instead, this just makes me wish they would spend a little more time fixing important bugs every once in a while.

Read more…

Yesterday I had an interesting conversation about the meaning(lessness) of several software project names, such as the well-known PHP, Wine and others, but also smaller projects like OSUOSL's RAIV.

As it turns out, every fancy project name is also an acronym. That's something that got really popular among Open Source projects, and even though some people argue it is kind of lame, it is nice to see how many more or less meaningful project names and acronyms have come up over time.

A very nice one I learned about yesterday is: The TWAIN Scanner API, heavily used on Mac and Windows systems to connect image scanning devices to the operating system, is, when unraveled, simply:

Technology without an interesting name

:)

But there are also projects that are missing a "useful" solution for their "acronymic" name: I recently learned about the project called Oink. Oink is a collection of C++ static analysis tools and comes with a pretty awesome, still not-yet-"acronymed" name.

Anyone want to give it a shot?

Read more…

So, a few minutes ago, the Google GMail interface changed (again ;)) and has become even more "ajaxy":

The common operations are now a dropdown list in the upper right corner of the displayed message, for example.

So far, I like it. What do you guys think?

Read more…

How you you know you just walked straight through a worm hole and reached 1985?

When you try to unmount an unreachable network drive and instead of succeeding after a little while, you get a spinning beachball of death leaving you -- after waiting naively for 15 minutes (I went to get my mail etc.) -- with no other possibility than hard-resetting your computer.

While I like my Mac a lot, this is one of the days when I want to ask:

Apple, when do you finally reach the 21st century?

Read more…

It's the second time in only a few days that I read this, so I guess I have to comment on it.

Today I had some minor problems loading my GMail account, so it showed me an error as follows:

Alright, so imagine I am a naive computer user who just got this message and I obviously believe what they are saying. Now I am going to go ahead and disable my firewall altogether and live happily ever after -- until I come across the first script kiddie that turns my workstation into a spam bot, virus nest, or both.

Similar issue: That blogger on MSDN.com who nicely suggested to switch off the Phising protection altogether when the CPU usage of your new instance of MSIE 7 spikes on some AJAX websites. -- While he meanwhile revised it to an acceptable "add these individual sites to a whitelist for which you switch off the phishing protection", his initial suggestion was just as bad as the one up there by Google.

Come on, people. Not everybody is a computer geek. People actually believe what you are writing there.

So please start thinking before you type. Having people switch all their security features off first (but burying the information that this might be a bad idea somewhere deep inside the help files) is harmful and -- sorry -- just plain stupid. People will switch it all off, they will see that "everything works" and they will stop reading about the issue right afterwards.

If you really, really, really have to have them switch off part of their software (which is not too surprising for some paranoid security products), at least spend half a minute explaining how the workaround can be done securely, and only for the page in question.

You owe this to your customers. Or, to put it differently: If you handle your customers' privacy as carelessly as you handle their web security, I sincerely hope nobody ever tells you their social security number.

Read more…

alexa -- a search engine that provides traffic rankings about web pages too -- currently lists the first five websites on their "movers and shakers" list from Mozilla.

This blog is ranked #6 with an 80,000% growth and a weekly traffic rank of 4,277.

Chances are that's the one of the biggest growths Alexa has ever recorded for a single website :) -- I'm impressed!

Update: I am number 2 now! :) The growth percentage has fallen to a less impressive 460% this week but I guess being higher in the list is nice too ;)

Read more…

So, the MySpace DNS admins seriously point one address of the round-robin record for myspace.com to localhost. Smart move.

$ host -vv myspace.com
Trying "myspace.com"
;; ->>HEADER< <- opcode: QUERY, status: NOERROR, id: 61932
;; flags: qr rd ra; QUERY: 1, ANSWER: 5, AUTHORITY: 2, ADDITIONAL: 2

;; QUESTION SECTION:
;myspace.com.                   IN      A

;; ANSWER SECTION:
myspace.com.            8282    IN      A       127.0.0.1
myspace.com.            8282    IN      A       216.178.32.48
myspace.com.            8282    IN      A       216.178.32.49
myspace.com.            8282    IN      A       216.178.32.50
myspace.com.            8282    IN      A       216.178.32.51
...

While 127.0.0.1 is most certainly "my space", I don't think many people host a copy of MySpace on their workstation...

Wonder how many people have to blog it until they fix it ;) In any case, Thinkgeek already has the right t-shirt for them.

(via OpenDNS and Elliot Back)

Read more…

The Microsoft Internet Explorer Team sent us a cake for the upcoming release of Firefox 2! <!--more--> <!-- was: /media/wp/2006/10/fromredmondwithlove.jpg -->

A big thanks to Redmond, Washington!

P.S.: No, it was not poisoned ;)

PPS: My Bandwidth bill was exploding. Thanks to everyone who helped my bandwidth bill by contributing a dollar or two. These are the friendly people who did.

Read more…

I came across a nice video this morning: 1985, Steve Ballmer sells Windows 1.0.

Un-be-lievable! It comes with Write and Reversi and runns bla-zing-ly fast on a 286! And all that for only 99 Dollars! :)

(via Google Blogoscoped)

Read more…