To register for an Internet.com membership to receive newsletters and white papers, use the Register button ABOVE.
To participate in the message forums BELOW, click here
General IT ManagementDiscussion of challenges facing IT management including articles published throughout the Earthweb IT Management network at Datamation, eSecurityPlanet and CIO Update.
I feel that the comparisons that are drawn are unfair. As you build an interface that appeals to more and more people, you are most certainly have to address usability concerns. In any case this means make things simpler and in every case thus far it means lower security. This is the case in all OS's faced with walking this path whether we are talking about Windows, Ubuntu or OS X.
The difference between Ubuntu and Windows is that under the hood Ubuntu is still a full Linux OS. The guru can still play with all aspects of security and configuration utilizing a strong knowledge of the command line and vi. Trying to implement this level of detail in the GUI would be nearly impossible and at best confusing to the average user that we are trying to draw into Linux.
Now under the hood, Windows is, well.... Windows. It has always been a low powered OS that started on the desktop and hasn't ventured far from its roots. It has been contorted to fit the needs of the enterprise services market but is still approached from the perspective of "How can we control this from the GUI?"
Ubuntu is making great strides to make linux accessible to the general public while making it a viable platform for the guru's of the world as well.
This seems like an odd duck of an article. Maybe it's just me, but I was expecting an article that addressed changes between 7.04 and 7.10, not complaints about 7.04 features that were continued with 7.10.
The excellent GUI installer in 7.04 certainly used default partitioning for a full install (you must use the Alternative distribution for fine-grained control and other advanced install options that geeks like us crave) and provided pre-selected packages as default. I well-remember the old approach of killing the user with choices; trust me, I do not miss slogging through hundreds of pages of applications to find the ones that Ubuntu provides by default!
The author also trashes the long-standing Applications --> Add / Remove approach to installing applications, because "After all, apt-get serves the same purpose as all of them". The clean, graphical install of Ubuntu was one of the primary features that attracted me to Breezy in the first place - do we really want to require users to use the command line just to install an application? And the popularity ratings arrived ahead of the web 2.0 revolution - but the author appears to be a web 1.0 kind of guy, and doesn't understand how the consensus opinion gives insight into a package's potential.
Oddly, the author avoids mentioning the actual changes in application management included in 7.10. Support for CNR isn't universally popular with Linspire's weird deal with Microsoft, but it does nicely replicate the Firefox extension experience at the application level (yes, I know Linspire had it before Firefox ;-).
Also strange was the praise for Debian's "dpkg-reconfigure xserver-xorg" command ("accommodating all levels of users and helping them learn as they go"). Yet, the author mentions that Ubuntu supports precisely the same command, yet "Ubuntu seems to forget these goals" (user education and choice)" in favor of a quick fix that keeps users ignorant and unaware of alternatives". Providing a GUI alternative to an obscure command line strikes me as broadening the range of users supported, since the geek way has been preserved as an alternative. Requiring a command line tool to adjust monitor resolution is perhaps one reason Debian has such a small following (i've seen it mentioned many times in forums).
I could go on, but you get the idea. Despite the misleading title, the article is basically a monologue on why the previous version of Ubuntu doesn't live up to the author's rather unique view of how an OS should interact with the user. Ubuntu is far from perfect, but thankfully, it's changing in a much more forward-looking direction.
The executive summary of my response to your article is this: "You are not in Ubuntu's target audience, but the things you want to do are possible if you look a little closer."
I'll go through section by section:
Installation
=======
All the things you want to do can be done with the Alternate installer. You can have all the control you want with it. That's what it is for. The "normal" installer is designed for people who don't know enough to make a decision regarding packages or partitioning, or who don't want to be bothered with deciding. It's not a bug, it's a feature. If you want different features, use the Alt installer.
As far as your complaints about keyboard layout, well, either you know you need another one or you don't. I've never run into someone who was unwittingly using a non-standard keyboard. Even if that were to happen, a text box where one can experiment seems to be the ideal way for one to ensure that their keyboard layout is what they expect it to be. This argument seems completely baseless to me.
This statement made me laugh: "Users cannot even choose the initial software to install. This lack is not only frustrating, but violates a main principle of security. After all, you can hardly secure a system if you do not know what is going on it."
If that's how you feel, you really should be using LFS. Again, the people Ubuntu is targeting would be unable to make a choice here, and being asked to make that choice would turn them away. Further, one of the biggest reasons to use a distro at all is so that you can have a pre-determined and well understood starting point without having to hand-pick every component. A basic install of Ubuntu is about as secure as it is likely to get and still be usable. If I remember correctly, there are no open ports in a default install. How would you improve on that? Further, the components that are included are well understood and easily discoverable. If you want to remove some, you can. If you can't afford to have something on a network for the time it takes to "harden" a system to your liking, it sounds like you have bigger problems that need addressing.
Boot and Desktop
===========
The Grub menu not appearing is a feature, not a bug. Even so, there is the "hit escae to see menu" prompt. The process for getting to the menu is easily discoverable if you need it. However, the target audience for Ubuntu just doesn't care about that. They want their computer to go from "off" to "desktop" as quickly as possible, with no intervention required on their part. The same thing goes for boot progress messages. Those are unintelligible to most people, and will in all likelihood just scare them. If they learn what they mean and want to see them, they can by changing boot options for the kernel.
I assume when you talk about terminal font you are talking about the VT? Well, of course it displays jagged at high res, it's not actually high-res. Again, the normal user doesn't care what the VT font looks like, they will almost never see it. If you care about this, making that change is trivial for you. Just to illustrate that this is subjective though, I run several Ubuntu machines strictly from VTs in 1024x768, and I've never had a problem with the font. It's perfectly readable to me.
If you are talking about the font used in Gnome Terminal, again, I think this is subjective. I think the default looks fine.
Software Installation
=============
You complain about lack of options, but then when you get them, you complain still? What do you want, options or not? Each of the installation tools you mention meets the needs of different niches:
apt-get or aptitude - command-line users
synaptic - people with enough knowledge to know what they want, but want to use a graphical tool.
add-remove applications tool - People who just know they want "something". The popularity rating gives them some level of guidance to help them choose among several seemingly similar options. It may not be what it was intended for, but it is useful for this purpose. "I don't know which of these is good, but that oine is popular, I'll try it out.".
Updater - People who only care about updates and don't want to install anything else.
Your myopia here is saddening. The planning is not "uncharacteristically muddy". Rather than trying to build one mega tool for everything, they build or use specific tools for specific needs. Your desire for one tool to rule them all is more "Windows Thinking" than anything else I've seen here.
Security
=====
Forcing users to have a separate login for administering their own systems is abusive. It is too inconvenient, and will drive them away. Eliminating the root user and encouraging the use of sudo closes one well-known attack vector. If a user is inclined to abuse sudo, they would be just as likely to abuse a root account, because of this your particular argument here is baseless. The broad "Sudo Vs Root" argument is an old one, and is not likely to ever be completely settled though. There is no clear "right" answer. Allowing anyone to make changes to a system implies a certain level of insecurity, regardless of how it is achieved. The approach that Ubuntu (and OSX) takes is a good one, and arguably the best choice for the people Ubuntu is targetting. If you disagree with it, enable root, and disable sudo, or use a different distro.
Ubuntu Vs Debian
===========
Let's say it again, the people that Ubuntu is targeting don't care about computers, and forcing them to "learn" is abuse. All of the options and flexibility and opportunities to learn are there, and easily discovered by those who are interested. Forcing them on everyone will do nothing improve the uptake of Linux on the desktop. They will just go elsewhere. Your dpkg-reconfigure example is a lame duck from the start. When someone uses that, they haven't learned anything about X, they've just learned how to run through a confusing wizard that does a rather poor job of configuring things for them. Really learning would be editing xorg.conf by hand. But again, nobody who uses computers wants to do that. They just want it to work.
I agree that there is a lot of polishing to be done, but none of your arguments or examples touch on places that are important to the furthering of Ubuntu's goals. In fact, they seem to simply be myopic, self-centered nit-picking of areas where Ubuntu does not meet your personal expectations, or falls short of your unrealistic expectation that all users want or need to be "educated". Further, your claims that you want options and control seem contrary to your complaints. Ubuntu is increasing options. By allowing the system to "just work" in the common case, they are making it accessible, and palatable, to a larger audience. That does not imply that the control you want is not available, just that it is not the "normal" way of doing things in the distro.
I find your article to be well written. However there is one thing I would like to add. I know Ubuntu does some quick fixes and avoids showing advanced features. I believe that that does not matter at all. It is more important for it to gain popularity and become an alternative to Windows. If it succeeds in doing that people will gain a new choice and most importantly Linux will be more widely accepted. Those that need advanced functionality will have it with another flavor of Linux.
So in my opinion Ubuntu exists to popularize Linux.
It is a lot easier to switch from distribution to distribution of Linux than from Windows to Linux.
I hope I wrote this in an understandable form. It makes sense to me and I hope it makes sense to you.
You lost me at hello. Specifically when you asked for more balance between advanced features and basic features. I was wondering at that point if you had found Synaptic and the command line apt-get.
Then you kind of went sideways and complained about having 3 options for installing software on Ubuntu. One is very basic (ADD/REMOVE), one is middle of the road (Synaptic) and one is for advanced users (the CLI for apt-get). I see that as being very well balanced. I don't know what you might ask for in that department. What are you stacking Ubuntu up against so I might be able to understand your position.
My wife is a mid level Ubuntu user and she's fine with Ubuntu. My mother, in her 60's, is a very basic computer user and she can find her way around just fine. Now that's a balanced distribution!
I'm an advanced user of Ubuntu and I loved it right up until 7.04 when everything was taken right to the edge... and got pretty buggy as a result.
It also seemed to me that you spent a lot of time reviewing things that aren't even Ubuntu related, but more immediately upstream package development issues. Will KDE 4 be included in 7.10? How can you peg things like that on Ubuntu when it's up stream development that will determine that.
The installation on the default desktop has very few features. A more advanced installer is available with the "alternate" disks, and no, it's not a live CD unless you want it to be. One of the questions surrounding Gutsy was if Ubiquity would separate the OEM option away from the alternate disk.
There are plenty of Ubuntu specific advancements in 7.10 that you didn't cover, didn't research, or seem to skip over with your four page review. For example, does Beryl finally work in Gutsy? I'm sure a lot of people want to know the answer to that one.
Check out Ubuntu's own internal targets for this release and I think you'll see a different story. How many say "implemented" verses "slow development" or worse, "Deferred". See https://launchpad.net/ubuntu/gutsy/+specs
I'm sure there is a 4 or 5 page story behind that alone and I think it paints a pretty grim picture for this release.
You made no mention of some Ubuntu related packages such as Ubiquity, AppArmor and other questions left hanging following Feisty such as,
Have they fixed the network manager?
Have they updated the proprietary drivers? for Nvidia?
How about support for Flash, Java, and other third party plugins?
Where does this new version stand on virtualization?
Can I still play Tux Racer?
There are a lot of things that my wife was looking forward to with this new release of Ubuntu. Most notably she was interested in the bittorrent clients. We had a nightmare of a time finding anything remotely close to stable and one of the posts in a wiki even admitted that there were no torrent clients available for Ubuntu 7.04.
From a more advanced side, I would really like to know what happened to a couple of packages that got deferred in the last release. Did they actually make it into Gutsy? Now that's a good basis for a review of Gutsy. Comparing Ubuntu to Windows is like comparing oranges and grassy hills.
You call it windows thinking. I'm not sure how this could be the case, windows is rarely ever without complicated configuration menus. I think they are following the path cut by Gnome not Windows.
Certainly Gnome and Ubuntu seem to think that user friendliness is of such tountamount importance that they don't need to focus on anything else int he UI. The reasoning I think goes that for most users you want the simplest and cleanest UI; advanced users will use the command line and config files.
I think there is a middle ground though. I use KDE which has GUI controls for nearly everything under the sun. Admittedly these controls are often hard to interpret, but at least they are there.
I think the ideal situation would be:
Default have the simplest options available. (like the basic installer, or the gnome controls)
Always have an advanced button that when clicked on brings a set of much more powerful controls (like KDE controls, debians advanced installer, or the new user management system)
Leave things done from the command line as "customization" instead of configuration.
If I could mod up qhartman's post above, I would. Bruce, you've missed the point of Ubuntu as a distro and also missed an opportunity to compare the experience users will have with Gutsy vis a vis Feisty. Lack of choice during installation is precisely the intent of Ubuntu as a distribution, especially when installed from the Live CD desktop. And reducing the complexity that users have to deal with in order to operate their computers adequately but safely is also the point of the whole Gnome desktop environment as we see it today. Users flee for the exits (or the nearest Windows or OSX desktop) in direct ratio to the number of questions you annoy them with at install time, or the number of choices you confuse them with when they venture as far as customizing their computer's desktop a bit.
In my opinion, Ubuntu does a wonderful job of assembling a basic set of desktop applications that most non-computer-hobbyist users will want and fitting it all on one CD, showing the Gnome desktop and applications off in a a fast-booting and very usable live CD environment, and installing them with a minimum of confusion and annoyance. When a non-hobbyist has gone so far as ordering a CD from Canonical or even downloading and burning one himself, that's a huge win for Linux. Why should we throw that away by scaring him off? He's already worried (mostly needlessly these days, luckily) about messing up his computer with Linux. He may also be worried that he forgot to back up some of his stuff - or even all of it, before installing Linux. And you want to raise his stress level even further by giving him choices he won't understand? C'mon.
What's really telling in the popularity of Ubuntu these days is how many experienced users - people who sneer at other Debian-derived newbie distros like Xandros or Linspire - happily run Ubuntu on their own computers. If it was as limiting as Bruce seems to feel, I don't think those experienced users would put up with it, do you? Ubuntu is, after all, still Debian underneath. update-rc.d and dpkg-reconfigure are still there. There is still an /etc/alternatives directory and accompanying set of symlinks. aptitude and apt-get on the command line work fine. I'm positive most of the experienced crowd know they're there and use them. They know how to add the apps they like that aren't installed by default.
As for what Gutsy is really about - the article Bruce could have written but didn't - well, my opinion is that it's not all that revolutionary compared to Feisty and earlier releases. It will have Gnome 2.20 and compiz eye candy will be enabled if the open-source drivers for the computer support it. (Take a bow, Intel folks ). Or the user can use the restricted drivers interface to install non-open-source drivers like Nvidia or (God help him) ATI. Most non-technical users will be able to handle that. Some won't and they won't get the eye candy. It's an imperfect world out there.
For most of the first ten years of its existence, the GNU project (focusing mostly on command utilities) and the Linux kernel project, were primarily concerned with providing a free and open alternative to the commercial UNIX operating systems and - particularly the Linux project - about providing an inexpensive way to get a UNIX-like experience on commodity hardware.
Over time, both the GNU and Linux projects scaled up and down - up to multiprocessors, scaling all the way to supercomputers, and down to handheld devices, used in multiple forms of communication.
In that alone, you can see that neither the Linux project nor the GNU project target one audience alone, they can be expanded and contracted or customized in any way the user or developer sees fit.
The Ubuntu project is an excellent example of taking one of the models and adapting it for a specific purpose. The initial Ubuntu project set out to produce a free, multilingual desktop product that could be used by millions of people around the world.
Given that kind of an objective, why should anyone be concerned whether Ubuntu limits choices, looks like Windows, or otherwise looks like a consumer commodity? If that is the way it comes off, then Ubuntu is well on its way to succeeding at its objectives. If Ubuntu were the ONLY choice in free software, then I would be much more concerned.
Actually, to hit this target market squarely, Ubuntu needs to do an even better job at the things it has already started.
I can appreciate your desire, Bruce, to have flexibility and all the other attributes you associate with classical UNIX and Linux systems. I really think that something like Debian Sid, Sidux, or Arch Linux would suit your interests more. As for the consumers, Ubuntu is as close as anyone has gotten in free and open software to reaching a large audience. I think they need to do even more of the things you object to - Freespire has some good ideas in this area - Ubuntu and Freespire ought to collaborate more. Much more marketing and mass merchadizing is needed; this is virtually non existent.
For someone who wants more choices, this is a choice, but within this choice, laying it out is the right thing to do for the consumer. For others, who want more customization and a slew of choices, we have tons of other distributions from which to choose. We can also hack up Ubuntu ourselves just by getting under the covers. Nothing stops you from doing that.
Therefore, I contend that as enthusiasts we have plenty of choices. What is needed are more and better consumer choices, and Ubuntu is working hard to bring them, and I applaud their efforts.
My main reaction is: Finally a real review. I don't agree with most of his critique though. Generally I do think Ubuntu could use a few more 'advanced' buttons but that's it. There have been a few very nice rebuttals on this already.
But it has to be said that I finally found a review that goes beyond "I installed distroX and it worked nicely..." usually followed by a list of features that can be found in the release notes. There should be many more of these attempts to really review a system critically. Though this author seems to have been digging for problems that really are good features after all.
A really good review would actually use and stress a system for a while. I must say that many distro's looked nice, where easy to install but where lacking features I wanted that I only find out after a while. Like having small repositories, a slow/small forum and general lack of documentation. These are all outstanding features of Ubuntu. Some other distros just got slow or very unstable after a little while (ahum pclos ahum).
These things can be hard to review if you write a review with a lot of knowledge and no need for fora or documentation. Or if you don't actually use it and therefore don't find out if essential apps are in the repos or how stable the system is in daily use.
Another thing I notice on the reactions here is the notice of 'the average user' or 'audience that ubuntu targets'. Considering it's popularity it's save to say that Ubuntu at least also reaches many advanced users. And without surprise. Having the knowledge is one thing, wanting to spent a lot of time configuring and fine tuning another. I use linux for a while now and I could probably work with Debian or even Slackware too. But why would I want to? As mentioned above: pre-configuring is the whole point of distro's. Ubuntu should do everything and I should do nothing - regardless of my ability to do it. I order pizza's too, even though I'm not a bad cook.
As for ease of installation, one should know that the 'average user' never ever even considers to install an OS. They may consider to install other software if someone else told them repeatedly that it's going to be easy. So I don't think a few intelligent questions there will hurt. Have you heard of anybody installing linux that did not know what partitioning means? Doesn't happen in my world. Also I don't care if installation takes me 20 minutes and 6 clicks or 30 minutes and 15 clicks. When it's done it's done and I get to the real work of 'being an end user' - which is much more interesting, time consuming and fun in any case.
For a next release of Ubuntu I would like to see better KDE integration though.
At the risk of stating the obvious, up until now most/all of the alternative OS community (with the exception of Macintosh) has been playing "catch-up" with Windows with respect to software. What I mean is, making sure that there are alternatives to most Windows programs that will run on these alternative OS.
What Macintosh has done, and which is why I think it has the largest share of the alternative OS market, is Pioneer its own programs and areas of speciality, forcing Windows to play catch up. That's the better approach because as long as the alternative OS offers Windows users nothing more than alternatives to what they already have, you're not going to get many of them to switch over.
There must be some clearer advantage in switching, not just the promise that it will be a smooth transition and that the programs they use now all have equivalents. That's not enough motivation.
__________________
Datamation Online Community Manager
At the risk of replying the obvious... linux actually has a lot to offer that is 'new and improved' for windows users. in my own experience (a subjective field...) there is usability and less nagging. Just think about the ease of software installation via a packagemanager..More objective ofcourse there is the server part, the safety part, the freedom part.. to please the average home desktop user there is lots of eyecandy windows users can only dream of...
I agree, but that's not enough for the average user. They might not be as dissatisfied with Windows as someone in I.T.. For those people, things like Mac's superior sound handling is the type of impetus I'm referring to. Or it's superior graphics capability. To a large extent windows has now "caught up" in these areas, but that's the point. Make MS play catch up, rather than the other way round. Average users don't understand security for the most part.
__________________
Datamation Online Community Manager
You're probably right. But I think it's mainly a matter of marketing. Thinking about the famous mac vs windows adds. Mac has been able to make windows appear as sully old fashioned and stupid. It has itself an image as being used by the hip and cool. Such advantages in perception are not so easy to achieve. Especially with the way linux' communication strategy is organized (ie. very bad or not at all - main perception being that of nerds ranting against each other). Main focus until now has been development. Equal focus should come on integration in the design and pr. Take the way ooo was in the news recently. In any company one would have discussions and clashes, ups and downs. Here however it seemed as if the whole product might be discontinued. Even though with FOSS software and so many users this is the least likely thing to happen...
Precisely. Look at what the Mozilla Corporation has been able to accomplish building what amounts to a shell browser for another company's rendering engine(although they did modify it). These are just some of the things their marketing genius has managed to convince the Web surfing world of:
1) Firefox is the most secure browser- (not true, Opera is more secure, mostly because it's less popular, and in fact on secunia.com there are 143 vulnerabilities listed for Firefox, 104 for IE)
2) Firefox is the fastest browser-(no hard evidence to back this up, and in fact in some rendering tests I've seen such as this one, IE is consistently the fastest)
3) Firefox is the most standards compliant browser-(this one may be true, but they neglect to mention that if you feed it certain code it's not expecting, it will crash. Opera, also very standards compliant, rarely crashes when fed poor code)
4) Firefox is an extensible browser-(quite true, but with virtually no intrinsic features, a feature junkie needs to hunt down and install a dozen or so extensions which really aren't quality controlled, nor tested against each other for compatibility issues, nor security issues. Those extensions often need to be updated as the browser upgrades, provided the author has updated them)
All in all Firefox represents one of the more impressive marketing campaigns I've seen on the web. I think if some of the alternative OS such as Linux took a page from their book it would really help grab a slice of the pie.
__________________
Datamation Online Community Manager
Linear Mode
