Why Ubuntu and not Debian?

June 8th, 2005 § 2 comments § permalink

My dear friend Metal took issue with my choice of Ubuntu as my new primary desktop in the comments of the post I wrote about the switch.

The matter of forks is older than open source itself, and I won’t go into the merits of the question here and now. It’s enough to say that I believe forks are essential to the open source process, and beneficial in most cases.

That said, why should I use Ubuntu, which is a fork, and not Debian, the primary trunk? Just saying I believe in forks in no reason enough for the choice.

I presented two reasons in my previous post:

First, Ubuntu is easier to install than Debian. Not only that, but it also recognized my hardware much better than any other distro I tried. It presented me with clear choices, made choices for me when they were obvious, and didn’t bother me asking how many buttons my mouse has.

Second, Ubuntu is much more polished than Debian. A lot of effort was clearly spent to make it comfortable to end users, and it shows.

Obviously, considering the relative positions of the Debian Foundation and Canonical, that is hardly a surprise. The latter has much more resources available than the former (from what we can gather from news and press releases, at least), and can use them to work in ways not readily accessible to Debian developers.

But that’s what it matters, and Debian is aware of it. I talk more about that in a while.

The first complaint Metal made is that Ubuntu isn’t giving anything back to the Debian community. I disagree. Ubuntu is giving Debian something it was losing: users. That’s more important for a distro than anything else, even code. Even if Ubuntu just repackaged Debian with more eye candy, I would see that as a valid effort from this point of view.

But Ubuntu is much more than simply Debian repackaged. As far as I know, and from what I’ve seen since I started using Linux, Ubuntu is the first coherent effort to create a Linux distro that is good enough for end users without losing the focus on open source and quality. Debian has quality and focus, but is not suitable for end users. Ubuntu solves this problem elegantly, in a way not even Mandriva, Linspire, and Suse — all notoriously for their focus on usability — were able to do. I’ve used them, and a few days with Ubuntu were enough to see that.

Of course, that’s not exactly the sense in which Metal said Ubuntu is not giving anything back. But, even when code is taken in account, the situation is not that bad.

I don’t know if and how Debian and all its forks exchange code, but I believe we are not far from seeing more integration is this area. That can be perceived in the decisions taken by the new Debian’s Project Leader. The goal of a more specific focus where supported platforms are concerned and the possibility of a quicker release cycle show that Debian is learning the lesson. After all, nobody wants to use a distro where all packages are lagging two or three years behind in relation to their more recent versions. And don’t even start talking about testing and unstable. They don’t get security updates, and I don’t want to spend my copious free time worrying about conflicts and changes in configurations that stop services from running.

Metal’s second complaint is that Ubuntu is not compatible with Debian. Again, I believe that can be seen from two different angles.

First, compatibility is a gradual process. The changes introduced by Ubuntu were interesting and necessary enough to make incompatibility, to the extent needed, acceptable. In a certain way, Debian is the distro that needs to adapt itself. Ubuntu is much more pleasant that Debian, and that a case when a fork can teach a thing or two to the primary project. After all, a fork would exist if the primary trunk was good enough.

On the other hand, I don’t care for compatibility. Ubuntu and Mandriva are equally incompatible. On Debian, the Apache configuration is located at /etc/apache. On Mandriva, it’s located in /etc/httpd. Other similar variations are quite common, and it affects everything I write for those distros. That’s why things like ./configure exists. Even if Ubuntu is a direct fork of Debian, it’s not required to follow in Debian’s steps.

To all pratical purposes, at home I’m a typical end user, and I want the latest and the greatest. Ubuntu gives me that. If Debian can do the same, even better — I will have more choice, which is what open source is all about. But, for now, Debian will remain confined to my servers until something better comes along. And that will only happen if we drop the zealot talk and start seeing each others efforts as something that adds to the community, even if those efforts are forks.

At the end of the day, when everything has been done and said, a Linux distro for end users should try to please a single public: grandparents of common users. If this public can use a distro, with no more and no less pain than a typical Windows session provides, Linux at the desktop will be real. From what I can see, Ubuntu is much closer to that than any other distro. And no amount of ideology will change that.


June 5th, 2005 § 4 comments § permalink

Sometime ago, I wrote about my first experiences with Debian.In the post, I said I had loved the distro, but that I didn’t plan to use it for anything else but serves since its long update cycles resulted in lots of old packages missing the features I needed in development or desktop systems.

Time passed, and Ubuntu appeared, a Debian fork with the state purposed of building a distro featuring the best of both worlds. Debian’s famous stability and easy of maintenance would be preserved, but Ubuntu would have shorted update cycles, resulting in a more modern distro, more like Mandrake and Linspire.

I decided to play with Ubuntu as soon as possible when I first heard about it, but I only managed to do it this week. One hour using it through the Live CD and I was done with Mandrake as my primary operating system. I installed it today, and I was very impressed with the whole process.

So far, I’m liking it. When you need to install a new package, Debian’s easy of use is quite evident. And the system feels a lot more polished, which was one of the things I liked about Mandrake. From installation to post-install configuration, everything works well.

In fact, the only problem I had with Ubuntu so far was related to my video card, whose advanced features were not supported by default. Nothing that couldn’t be solved with a quick look at the Wiki.Importing my old data into the system was a pretty simple process, with only a few complications here and there (Evolution, for example, imported my old e-mails but refused to import the accounts, which had to be recreated manually.)

I’d say Ubuntu is the first distro that proves that desktop Linux is a reality. None of the new distros hide so well what’s inside for users who know nothing about Unix systems. The few errors I experienced today, for example, would be equally baffling for Windows users in a similar position.

Some people criticized Ubuton for its use of sudo. I think that’s something Ubuntu got really right. Considering the profile of common users, the solution used keeps the system safe without sacrifing usability. It makes Ubuntu as flexible and easy to use as Windows systems when you need to change configurations and install new applications. By the way, Ubuntu wizards and control panels are really good.

Now, I have only one question: how the heck is Ubuntu pronounced?


January 20th, 2005 § Comments Off on Postfix § permalink

Color me impressed. The more I use Postfix, the more I’m grateful for it. I’m always finding things to solve problems I didn’t even know I had.

In the past week, I’ve been slowly migrating the sites I host to a new server, and I wanted to guarantee that no e-mail was lost during the transfer. The solution is simple with any MTA worth its salt: just instruct the old server to relay all e-mail to the new server. But doing that with a single SQL instruction is way to easy. Where are all those text files I have to change and make sure are properly parsed?


January 20th, 2005 § Comments Off on Gotchas § permalink

Things to be aware of when your are using:

Why Rails?

December 19th, 2004 § 2 comments § permalink

Ian Bicking is not particularly impressed by Rails. He says there’s nothing special about the framework, nothing that makes it different from other Web programming libraries. As a longtime user of this kind of frameworks, I understand his criticism; I even agree with some of them. As a recent Rails enthusiast, I disagree with his comment as a whole.

As readers of this blog probably know, there is something I value above all other things in a programming language or library: simplicity. More than any other feature, that is what defines the power of a tool to me. Simplicity, in this case, is not merely ease-of-use, but a set of characteristics that make the tool useful in the long term.

So the short answer to Bicking’s contention is: Ruby presents a kind of simplicity no other framework I’ve used (or at least know of) does. As such a short answer is not so interesting, I’ll comment on some specific aspects of Bicking’s post.

» Read the rest of this entry «

Open Books Project

December 16th, 2004 § Comments Off on Open Books Project § permalink

Via Galvez: O’Reilly Books is making a lot of titles available for free in PDF and HTML. I don’t know if this is a new site, but the selection is quite interesting, including a bunch of out-of-print books that contain good material even if part of it is outdated. Some good titles there: We the Media, Unix Text Processing, DocBook: The Definitive Guide, and Linux Network Administrator’s Guide.

Trusting Google

December 15th, 2004 § Comments Off on Trusting Google § permalink

A couple months ago, when Google Desktop was released, I wrote an entry saying I didn’t trust Google, especially because of certain dissonances in its declared policy (Don’t be evil) and its actions — for example, when Google News bowed to China’s government and censored search results.

At the time, some even called me a neurotic, but I’m not particularly bothered by this label. Where my privacy is concerned, I border on paranoia. If I don’t defend myself in this area, no one else will. For certain things, I really want to know what is happening, and how it’s happening.

Returning to the subject at hand, however, the question of whether trust Google or not is still open. With Google’s announcement of a new large project, a deal with various large universities in the US to digitize and make available online thousands of books belonging to the libraries of those universities, people are starting to consider the issue again. This time, three well-known names in the blogsphere offer their opinion.

Scott Rosenberg writes:

The public has a big interest in making sure that no one business has a chokehold on the flow of human knowledge. As long as Google’s amazing project puts more knowledge in more hands and heads, who could object? But in this area, taking the long view is not just smart — it’s ethically essential. So as details of Google’s project emerge, it will be important not just to rely on Google’s assurances but to keep an eye out for public guarantees of access, freedom of expression and limits to censorship.

Rosenberg welcomes Google’s deal as one of the most important undertakings of the company, but he doesn’t lose sight of the fact that Google is still a business, with commercial interests that may conflict with he needs of the general public. Today, he argues, we can reasonably trust Google because the company has a good record of making the right calls when it comes to pleasing its users. But that won’t last, he continues. Eventually, Google’s current leadership will give way to a new generation of executives whose priorities may not be so aligned with the user-geared policies in place today at the company.

Obviously, I agree. Google still has the special glamour of a company who always innovates in what it does, always gets it right, and keeps raising the bar on what people expect of Web applications. But, many times, that also causes people to forget some basic facts. No company aligns its commercial interests to those of their users out of generosity. Expecting Google to proceed differently is a mistake.

Dave Winer talks about that in his commentary of Rosenberg’s article:

Another way of looking at it: What if Microsoft were doing what Google is doing? Of course we wouldn’t let them do it without a very serious and probably very shrill examination. Well, I’m telling you, Google today is as dangerous as Microsoft, and I wouldn’t bet on their trustworthyness (sic), not without a lot more light having been shed on this. The technology industry is built on a foundation of arrogance and disdain for users. Google is too. You may not have seen it yet, but I have.

Winer’s commentary is interesting because Google is often compared with Microsoft, but in a completely opposite aspect — that of Microsoft’s killer, just as Microsoft killed IBM as a software company. That naïve comparison also obscures the questions that should be done about the control Google has over the data that belongs to its users. Commenting both Rosenberg’s article and Winer’s entry, Jon Udell says:

Bottom line: I think that stewardship of so much of our private as well as public information requires a lot more transparency than Google currently practices. For example, on the day that MSN Search was announced I pointed to a funny but tasteless prank that Google News played on Microsoft. Are we really supposed to believe that an algorithm chose that unflattering photo of Bill Gates? Of course it didn’t. But how can we know for sure?

Today, Google controls a enormous set of information, and I’m not talking about the documents in its search database. For Orkut, Google Mail, and Blogger, Google stores in its servers a huge amount of data on behalf of its users, many times under dubious service agreements. If we question Microsoft because of those issues, why shouldn’t we question Google? Fortunately, Google has been providing ways to export that data to others places, even though sometimes they only do it when people have already routed around initial barriers.

This stewardship, as Udell puts it, is a big responsibility, and Google still has not faced a crisis of such magnitude so as to allow us to judge how they would react if serious problems arose with its terms of service. In fact, this would be an good way to gain trust, defining explicitly how the company would handle problematic scenarios.

The issue of centralization is also important. Centralization is almost always considered an undesirable characteristic of protocols, products, and services. For the kind of products Google offers, however, centralization is desirable in some ways because it makes a better service possible. Also, most of the times, decentralization makes little sense for a business. Then again, alternative access is also important. Closed and limited access points always penalize users.

This is where I agree the most with the open source movement. Although there are still many issues to be solved where the commercial interests of a company and the need to provide assurance to users that their data won’t be lost or locked away clash, I still believe open source offers the best path to the protection of the public interests.

In short, as good as Google may be as a company, I still believe our role as users is to always watch what is going one. Otherwise, we risk the prospect of unpleasant surprises in the future. As Rosenberg also writes in his article, what happens if a new generation at Google decides it’s time to be more aggressive with its product portfolio?

So I maintain my position. I’ll use some of the products Google makes available to be, but I will make my choices understanding that the company may abruptly change its policies, as it’s its right. While the water is nice, I will keep swimming at the pool. But if push comes to shove, my options will be open. I believe that’s a sensible stance.


November 26th, 2004 § 1 comment § permalink

A friend of mine, who works in the education industry, asked me the following questions about SCORM:

Why SCORM? Why should I use it, with all the paraphernalia of files it entails, instead of using, for example, a weblog with some JavaScript to integrate the learning objects or an in-house system?

I’m asking the question in the context of small schools with a limited budget, no technical support staff, small labs with no servers, unknowledgeable professors, etc.

These are pertinent questions, and I’ve thought about them many times since I started consulting in this area, although never in the context mentioned above.

So, why choose SCORM? As it often happens when considering a new technology for adoption, people tend to rely more on hype and market considerations than on what the change really means.

» Read the rest of this entry «

I don’t trust Google Desktop

October 15th, 2004 § 1 comment § permalink

So Google launched its desktop search engine. I, for one, won’t be installing it. The reason, plain and simple, is that I don’t trust Google anymore. Not because of the IPO, but because Google keeps saying they still follow their “Don’t be evil” policy, but their deeds don’t show that. (Not that they need to be accountable to me, by the way.)

The Google Desktop EULA states that the program will collect non-personal information if you use it. You can opt out of this at install time or using the preferences panel later, but there’s a catch: Google desktop will install a unique identifier in your computer and will send that back to Google whether you like it or not. That unique identifier is not used if you don’t send data, but it’s there, and if the EULA changes suddenly, it’s ready to be utilized. And, even if you opt out of sending non-personal specific data, Google Desktop will still use the same cookie the other services like Google Search and Google Mail use, allowing them to correlate data on you as you use those services. It’s supposedly non-personal data, but Google Mail seems a bit too personal too me.

You may call me a paranoid, but I don’t like the way Google is using all this information. I use Google Mail sporadically, but not for important e-mail, be it work-related or personal. Although the Google Mail EULA changed after some complaints, it was too strange at the beginning to make me comfortable with trusting Google with my e-mail.

(Before you ask, I don’t use any kind of web-based mail server, I run my own mail server, and I use encryption whenever needed. I just happen to have a Google Mail account because people kept saying it was great, and because I also wanted to see its new-fangled JavaScript-based interface. Anyway, I doubt Google would be interesting in reading the e-mails sent or received by their users. Too much trouble for too little gain. But I still don’t trust them.)

And there’s also Orkut, which has its own strange EULA too.

I probably can block Google Desktop from sending messages back to the mother ship at the firewall level, but I’m not installing it on principle. The recent problem with the Chinese version of Google News clearly illustrates the problems I see with the company, and that’s why I don’t trust them anymore.

Anyway, Google doesn’t own me anything. More than that, they’re a public company now, with shareholders that will demand profits (whatever their claims), and they may do whatever they want. But that doesn’t mean I will blindly use any kind of product they have, even if they give it away for free, just because they used to not be evil.

But, I will have to confess: I never trusted Google. Its cookies have been blocked on my browser for an long as I can remember. It’s just that my Google trustiness level hit another low.


October 3rd, 2004 § Comments Off on Clusty § permalink

Vivisimo has launched a new search engine called Clusty.

Clusty’s differential is that it allows users to “cluster” search results by topic, source or URL. In the few tests I did, grouping by topic seems to be the most useful method. I don’t know if the results returned are really relevant (they seem to be), and I guess that will require some use to assess. Anyway, it looks really promising.

Where Am I?

You are currently browsing the Technology category at Reflective Surface.