[Sat Feb 27 14:56:54 CET 2010]

A friend passed me a link to the official Tron Legacy trailer:

Who doesn't remember the original Tron, a 1982 movie that made it to the list of science fiction classics. Well, Tron Legacy is the sequel to that movie with a similar plot but much better graphics, obviously. I cannot wait to show it to my kids, who already watched the old one and sort of enjoyed it. {link to this story}

[Wed Feb 24 12:52:41 CET 2010]

CIO Insight published a slideshow on the 20 IT skills that are on the rise and, together with some skills that anyone could expect (several specialized jobs related to Microsoft Exchange Server and SAP), there are also some surprises (are C, C++ and Perl truly on the rise?). Somehow I find that difficult to believe. Don't get me wrong. I fully understand all those languages are still quite important, but that's one thing and arguing that they are on the rise is quite a different thing. {link to this story}

[Wed Feb 24 12:44:25 CET 2010]

I came across a piece titled 5 Tools to Googlize Your Business published by OPEN Forum that is worth mentioning here. While I understand some people's worries about privacy concerns regarding the different Google services and cloud computing in general, the thing is that they truly provide a very cheap, accessible and flexible way to organize groups. Would I consider using them if I were in charge of a large corporation? No. Chances are that privacy concerns would be the key in that case. Besides, large corporations have their own resources. However, if I were in charge of a smal business I might consider using these tools and, definitely, I recommend them for non-profit organizations and the like. I have been using Google Calendar, Google Mail and Google Docs for one such organization for at least a year now, and am pretty glad I made that decision. It is free, flexible, always available and I don't have to worry about operating system compatibility. A true godsend, if you ask me. {link to this story}

[Mon Feb 22 12:55:26 CET 2010]

I don't know about you, but whenever I run OpenOffice on a system that doesn't have plenty of memory... well, it grinds to a halt just trying to launch the darn thing. If, on top of that, you sare trying to view a large PowerPoint presentation or spreadsheet... well, in that case, you are better off just giving up or, at the very least, leaving the computer for a while and go get some coffee and stretch out your legs. Yes, I know newer versions of the app do slightly better, but they still suck memory like there is no end of it. So, today I decided to run a quick apt-get, install Gnumeric and give it a try. All I can say is that I was very surprised. It launched very quickly, handled the same large files easily, had no problems displaying the spreadsheets I wanted to view (I have no idea if it may have problems with certain files and macros) and, above all, it didn't hog the memory. All in all, Gnumeric is a nice app. I highly recommend it to anyone who is sick to his/her stomach of OpenOffice taking forever to launch and using up all the system resources. Nice work. {link to this story}

[Fri Feb 19 10:04:19 CET 2010]

I wrote a few days ago about the overall arrogant attitude taken by a good amount of BSDers out there. My comments came as a consequence of something I had heard on the bsdtalk podcast and was accompanied, of course, by the usual disclaimers (not everyone behaves like this, the OS is still a very solid product, etc.). Well, yesterday I came across the perfect illustration of what I was saying. While perusing the Distrowatch website, I came across a piece titled GNOBSD: killed by GUI-is for wimps hacker culture. As it turned out, somebody by the name of Stefan Rinkes dared send an email to an OpenBSD mailing list announcing a live OpenBSD distro he had put together, and the reaction was quite strong. To be fair, some of the people who responded didn't like the fact that the OpenBSD project itself might lose income as a consequence of this (it's a fair comment, I think, and something that also worried those involved in the Debian project as Ubuntu grew in popularity), but others chose to deride it simply because... well, because it made it easier for people to use the OS! Now, I just don't remember this happening with any of the myriad Linux live CD distributions out there, and it does illustrate the hollier than thou attitude many BSDers adopt when discussing these topics. I know, I know, the BSDers, supposedly, don't have the same drive to evangelize that characterizes the Linux crowd (something that, incidentally, most BSDers also tend to explain while clearly showing some gesture of disdain towards the "GPL lunatics"). What can I say? First of all, I find it quite strange that a good amount of BSDers don't just avoid evangelizing but actively seem to take steps to make sure that people do not run their favorite OS, so that they can continue feeling like the chosen elite. But, second, one has to wonder what would have happened to open source (and BSD in particular) had Linux not showed up and taken the lead, expanding into markets where no open source software was seen before. My guess is that this would be a Microsoft world now. There is a good chance that projects like Samba would be dead by now, while others (like GNOME or KDE, both born from the Linux community) would have never seen the light. As a consequence of all this, there is algo a good chance that Apple would have never been able to return to the limelight as it did and, as for the BSDs, they'd have ended up like the commercial unices (dead or about to die). So, I'm glad it was the "loonie GPLers" and not the "sensible BSDers" who won this particular battle within the open source movement. Besides, this way they can continue feeling like the chosen elite, which suits them just fine. {link to this story}

[Thu Feb 18 13:45:06 CET 2010]

I had to chuckle while reading Linux Weekly News' quotes of the week from the Linux kernel mailing list, specifically when I read the one from Alan Cox:

It's really simple: overcommit off you must have enough RAM and swap to hold all allocations requested. Overcommit on —you don't need this but if you do use more than is available on the system something has to go.

It's kind of like banking overcommit off is proper banking, overcommit on is modern western banking.

It certainly puts things into perspective, doesn't it? Incidentally, the feature is not so different from IRIX's vswap parameter. There are still certain areas where commercial Unices are (or were, until very recently) ahead of Linux. {link to this story}

[Tue Feb 16 12:04:01 CET 2010]

Here is a good example of the cool stuff that Google employees develop during the 20% of company time that they have allocated to personal projects:

Clear evidence that, if you hire the right people, letting them fly freely pays off in the end. {link to this story}

[Sun Feb 14 20:28:12 CET 2010]

I have been listening to the bsdtalk podcast a lot lately. Following the BSD news is something I do every now and then. I first encountered FreeBSD back when I worked for USWest's ISP service in the late nineties and have been running that OS on a home server until pretty recently (about 2-3 years ago) in spite of my skepticism towards their community (more on this later). The fact is that BSD-derived OSes are indeed rock solid, as their fans claim. They're a nicely put together flavor of UNIX that is, nevertheless, fully open source, run nicely on most x86 boxes and doesn't try to lock you in. In other words, you get all the power of traditional UNIX in an open product. That is definitely nice.

So, what is not to like about the BSDs? Sorry to say this, because there are plenty of nice people in their community too, of course, but a good number of BSDers just come across as a bunch of smug pricks. Yes, sure this is meant to happen in any group of people you can put together but, for whatever reason, it's more present among the BSD community, just the same way there are more arrogant Linuxers than Windows users out there, let's be honest. Yes, BSD can claim to be a direct descendent from the original UNIX and its developers were behind very innovative technology, such as TCP/IP, BIND or Sendmail. And what? Arab scientists and philosophers were also behind lots of great discoveries several centuries ago and we barely pay any attention to the region these days, unless it is to hear about their contant political quarrels. The same way we have to admit that the BSDs are indeed rock solid, we also have to recognize that most innovative development is not done on their tree but rather somewhere else. To some extent, they don't have the mojo anymore, although the BSDs are definitely still a contender to take into account for certain purposes (I've never been the dogmatic type). Simply put, the Linux world is far more dynamic these days, and it also has far more money backing it up and financing new developments. It's where the action is in the UNIX and UNIX-like world, as even the top executives of Sun Microsystems ended up acknowledging. Again, that is not to say that you shouldn't consider BSD or that you shouldn't deploy it. I will repeat it once more: it does have its place, it's still actively developed, it is rock solid and it works very nicely on single-processor x86 systems. Also, if security is your priority, you should spend sometime checking out OpenBSD. No doubt about that. But if you want to scale to tens of processors, run high-performance applications for the scientific market, be sure that commercial vendors out there will develop on the OS that you are running and know for sure that every year tons of new products will be at your fingertips... then you'd better make sure you run one of the three biggies (i.e., Windows, Apple or Linux).

What else do I find wrong with the BSD community? Well, when it comes to this particular discussion, their arrogance is thinly veiled. They claim they don't care what people run out there, as long as they are allowed to run their beloved BSD. They also claim that this is something that differentiates them from the "Linux zealots" who, supposedly, go around trying to evagelize everyone in sight. And, we have to admit it, they have a point. However, the reason why they won't make much effort to evangelize their favorite OS ("like the Linuz zealots do") is not so much due to their love of freedom or their respect for other people's choices as due to their deeply seated conviction that they belong to a special elite and, by definition, elites cannot be too crowded. They're the chosen ones, you see. After all, they descend directly from the holy grail of true UNIX, unlike those Linux mongrels who dared steal their thunder. Think I'm being unfair? You just need to check out a few BSD publications out there. I'll give you just one example that I recently came across of. I was reading first issue of the BSD Magazine for 2010. Starting in page 76, you will find an article titled Living the PC-BSD Lifestyle written by James T. Nixon III where he tells us about the wonders of PC-BSD, a desktop solution based on BSD. This all fine, if it weren't because in order to praise his beloved BSD he seems to have the need to attack Linux. After telling us how much he loves games and multimedia applications, he goes on with the following statements:

I was also dual-booting random Linux distros (with much displeasure), because I got tired of performance on my Windows box. I found that I couldn't enjoy most of my computer related hobbies on Linux, and worse yet, most of the websites I was developing or visiting didn't work or look the same. So I forgot about Linux for a while too.

What? Does he even know what is talking about? First of all, there is no need whatsoever to disparage Linux in order to praise PC-BSD. However, if you are going to do it, at least do it with some data to back it up, please. The comments about the websites he was "developing or visiting" not looking the same or working the same are quite cryptic, to be honest. Go figure what he may be referring to. Yes, it is possible that you visit a website in Windows or Apple, then check it out on Linux and it doesn't render the same. However, this very same problem also affects the BSDs. It doesn't have anything to do with Linux and a lot to do with certain websites using technologies that are not open and supported on non-commercial OSes. But what truly pissed me off was to read that the "computer related hobbies" he couldn't enjoy on Linux were things like graphics editing and music apps that certainly run on Linux without any problem and, to top it all off, were developed on Linux! He even mentions apps like GIMP and Ardour in the next paragraph, for crying out loud! Does he even know what he is talking about? Sure, perhaps he didn't even know how to use the Synaptic software manager to install additional software, but that is not Linux's problem. It's his own! And what to say about his love of gaming? Not only are there more games that run on Linux than on the BSDs, but some of those that do run on BSD do so through... Linux emulation! In any case, if gaming is your priority, you're definitely better off with Windows. There is no doubt about that.

In any case, there are thousands of examples like this. BSDers who cannot say no to a cheap shot against Linux are definitely doing little service to their favorite OS, at least in my eyes. They're just coming across as a buch of elitist pricks. Hardly a way to convince more adepts. {link to this story}

[Thu Feb 11 12:41:10 CET 2010]

Somebody sent me a link to a piece published on the Web announcing that Facebook Chat may soon connect to iChat. Actually, the headline is a bit Mac-centric, since the true news is that Facebook has announced its intention to release a new interface to its chat feature that relies on Jabber's XMPP protocol. In other words, they will be using the same open protocol already used by Google Talk. In other words, the move won't only make it possible for iChat to connect to it, but also for any other instant messaging client that supports the protocol.

Now does any of this truly matter? I, for one, do find it useful. I do use Facebook quite a bit and find it very annoying that people start conversations with me using their interface, which is tied to their website. It forces me to be paying attention to their page all the time, which obviously I don't do because I have better things to spend my time on. The end result? I quite often miss people trying to talk to me about something. Sure, most of those people could use a regular instant messaging client but, for whatever reason (laziness, ignorance...?) they don't. And true, Facebook may also currently have the capability to decouple the chat window from the rest of the page, just like Google does (truth be said, they may or may not, since I didn't bother checking). In any case, I find that type of web-based interface sort of annying, especially since I already have an IM client that talks to multiple protocols. Why should I run one more window to do exactly the same? So, all in all, it's good news, I think. {link to this story}

[Thu Feb 11 12:32:08 CET 2010]

Here is a bit of geeky humor I came across of while browsing around the Dr. Dobb's Journal website. It's a short snippet of code to get tomorrow's date in Perl:

tomorrow_date {
  sleep 86_400;
  return localtime();
}

{link to this story}

[Fri Feb 5 17:11:11 CET 2010]

While listening to a recent episode of the Computer Action Show (formerly known as the Linux Action Show), the hosts mentioned a deal between Samba and Microsoft over the Windows protocols used by the Samba folks and how most of the free software community doesn't seem to be so riled up over that particular issue as they are over Mono and similar. I have to agree. It almost seems as if, to some people in the open source community, Microsoft and anything that touches it is tainted... but with exceptions, of course. In other words, Mono, Novell and Miguel de Icaza are simple pawns in Bill Gates' plans to conquer the world and destroy Linux. However, Samba can do as it pleases without unleashing the wrath of the Free Software Foundation. It just doesn't many any sense, unless one chooses to acknowledge that a good number of the FSF fans are, unfortunately, quite dogmatic and close-minded — and I say "unfortunately" because today's open source environment would certainly look very different without the work and inspiration of Richard Stallman and his cohorts. On this topic, as in the case of politics, paying as little attention as possible to highly ideologicized people is always a good thing. So, enough said. {link to this story}

[Thu Feb 4 11:45:19 CET 2010]

I recently came across of an interesting piece on C's biggest mistake written by Walter Bright and published by Dr. Dobb's Journal. Asking himself about the biggest mnistake in the design of the C programming language, the author argues that it is not null pointers, as most people would say, but rather what he describes as "conflating pointers with arrays":

I don't mean them using the same syntax, or the implicit conversion of arrays to pointers. I mean the inability to pass an array to a function as an array, even if it is declared to be an array. C will silently convert the array to be a pointer, and will rewrite the function declatation so it is semantically a pointer:

  void foo(char a[])

is exactly equivalent to:

  void foo(char *a)

This seemingly innocuous convenience feature is the root of endless evil. It means that once arrays leave the scope in which they are defined, they become pointers, and lose the information which gives the extent of the array —the array dimension. What are the consequences of losing this information?

An alternative must be used. For strings, it's the whoe reason for the 0 terminator. For other arrayd, it is inferred programmatically from the context. Naturally, every situation is different, and so an endless array (!) of bugs ensues.

The trainwreck just unfolds in slow motion from there.

The galaxy of string functions, from the unsafe strcpy() to sprintf() onwards, is a direct result. There are various attempts at fixing this, such as The Safe C Library. The there are all the buffer overflows, because functions handed a pointer have no idea what the limits are, and no array bounds checking is possible.

Bright offers a possible solution to the conundrum. All in all, it makes for an interesting read. {link to this story}

[Tue Feb 2 12:34:56 CET 2010]

While reading issue 7 of PragPub magazine, I came across of an article written by Daniel Steinberg on how to build a slide rule (it starts on page 5, in case you are interested) that got me thinking about our attitude towards technology. The thing is that not once, but several times, the author longingly refers to the "simpler life" that we use to live before calculators and cell phones. He even tells us about his love for Nero Wolf's detective novels, where the main character doesn't have the technology or the resources we see in contemporary TV shows like CSI or Law & Order. As one would expect, he says all this with a clear admiration for the "good old days", when men were men, and things were "simpler". And this is precisely what got me thinking. Are we sure things were truly simpler back then? In what sense were they simpler? Steinberg even tells us how the characters in the Nero Wolf novels have no access whatsoever to DNA tests and, to make things worse, while the fictional detective "stays in his home and uses his wits to solve crimes", his assistant "does most of the legwork" without access to a cell phone, which means that "he has to go find a pay phone or borrow one". How could any of this be considered "simple", one asks? As it turned out, the detective working from home has no way to communicate with his assistant in the field, and this one, in return, has to struggle searching for a stupid phone to call him back and report of any important discoveries. I am not going to comment on how tremendously inefficient all of this is but, on top of that, I just cannot see how anybody could consider it "simple". Quite to the contrary, it is a lot of work. And the same can be said of using Steinberg's slide rule instead of punching numbers into a calculator. The slide rule may certainly be quaint, perhaps even elegant, but definitely not simple. How the hell could it be? It requires quite a few more steps to make it to the same point, and a simple human error along the way will give us the wrong answer!

Yet, we see this attitude all the time when discussing technology. Plenty of people in today's society decry progress and technology, feeling nostalgic about the past... at least in theory. It tends to be just an image thing, a cool postmodern attitude showing how detached they are from technology, while society at large is nothing but enslaved to this new god. But, of course, don't you ever try to take it all away from them and send them back to the past. It's just an image thing. A "cooler than thou" fashion statement limited to this or that little detail of our lifes that never extends to any other field. I certainly don't think many of these end up joining the Amish communities out there. What fails, I think, is the whole premise. Society was not any simpler before we invented modern technology. Quite the opposite, it was more complex and even the simplest activities required a lot of work or resorting to convoluted mechanisms to help us carry them out efficiently. I'm not sure that getting the paper, fountain pen, ink, writing the letter, putting it inside an envelope, getting stamps and going to the post office to send it to a friend is any simpler than sending an email. It is not. It is actually more complex. Sure, the technology being used is simpler, but the process (and, therefore, the actions we perform during our daily lifes) was actually more complex. This is, I think, the key that explains the whole confusion: our tools are more complex these days, but the processes we follow to accomplish things with them are simpler; it was the opposite in the past, when the tools were simpler but the processes more complex and required more work. We'd do well realizing this, so we don't over-romanticize the past with our "cool" postmodern attitude. {link to this story}