You're saying when running a charity, Microsoft should be "allowed to compete"?!? This is the same Microsoft that has been repeatedly convicted of undermining fair competition through criminal antitrust abuses? This is the same Microsoft that is still in the process of being prosecuted for ongoing antitrust abuses? This is the same Microsoft that is being investigated for bribing government officials and standards bodies?
I say let them, they are going to have to up the hardware costs of the machine to get an XP port running, and it will inevitably be significantly less functional than what is already available for less money.
Great, then the OLPC brand is poisoned and there is confusion about what the capabilities of the different versions are. Then MS can undercut others on price in order to lock in a new market early and then bleed them for the next twenty years like they have been other markets. I'm less than impressed with this idea.
I sincerely believe that a foray into M$ for the OLPC will bring to light the inherent advantages of free software.
You're assuming people will try both and objectively compare them on a level playing field, then choose what is best for the kids. Given MS's history, their piles of cash, and their incentive here, why do you think that?
How many countries subsidize telcos with tax dollars to create their infrastructure? I'm curious. I know we are a spread-out nation here in the US, but there is no reason why cities with people living on top of each other (LA, Boston, New York, etc) can't easily have the infrastructure that the rest of the world has.
We've paid more per person in tax subsidies than many other nations. Take Sweden, for example. Their population density and median population density are both about the same as the US. Their subsidies, however, had legal teeth that required the telcos to actually provide something in exchange. They also had a huge embezzling scandal where much of the money was stolen. They still have significantly faster internet at significantly lower prices than the US, in exchange for a smaller per person tax.
The high speed internet problem comes down to pretty much the same thing as many other problems in the US. Politicians are willing to give private companies billions is subsidies, in exchange for hundreds of thousands being returned as campaign contributions. So long as this legalized bribery is allowed, companies will simply pay off politicos in exchange for subsidies or for not having to fulfill the agreements they made when the subsidies were given.
One has a major deal with Dell, but no Open Drivers, then the other announces they suddenly have Open Drivers. Is it anything on Dell's head to tell the first company to either cough up some open drivers or come and pick up their unused parts before they get discarded?
Yes. Dell needs to have multiple vendors in their pipeline so they can get them to compete where they really care, on price, as Dell usually buys the cheapest part X available on a given day from whomever. At least for consumer desktops, that is how they operate. I seriously doubt open source drivers are going to be a consideration on the low end and they sure haven't been in the past (speaking as someone who ordered hundreds of the same model from them only to find a wide variety of parts actually inside, some of which had OSS drivers like the test model we bought and some of which did not). More likely, Dell will push for this on midrange and higher end software where they actually offer Linux as a supported option and probably only then for machines targeted at business.
All it takes is for one company to start releasing open drivers and the rest will have to fall in line or risk loosing[sic] a lot of business practically overnight. In the end, everyone will benefit.
Except it actually does cost money to release drivers as OSS if you have to clean up the code or get rid of the hacks you use to work around your hardware flaws. In which case, it potentially amounts to costs raising for the first movers and people who don't plan to use OSS drivers still have to cover that cost (although it might be negligible on such a scale).
I always wondered why device drivers are not open source. As they make their money on the hardware they're not losing anything by giving the driver piece to the open source community to enhance.
Reasons include: they don't like providing anything they do for free because a competitor might use it, they don't want to expose their embarrassingly poorly written code, they're afraid their poorly written code will expose their security flaws, they don't want consumers to know about the hacks they use to work around hardware flaws or which compromise quality for speed.
It's a big assumption to assume that your computer is turned off when seized.
It won't necessarily be, however it does present a situation that falsifies the assertion I was addressing. Physical access does not necessarily mean access to the data.
It's a bigger assumption that your password or encryption key hasn't been paged out to virtual memory and then never cleared...
Those were issues with early versions of hard disk encryption built into OS's, but has been addressed by Filevault (OS X), Bitlocker (Vista), and TrueCrypt (Linux). Basically, unless the machine is in hibernate mode, or was shut down less than 5 minutes ago, no that isn't really a concern.
...and that you haven't reused your password for a less-secure purpose.
No. That isn't an assumption for me.
Anyways, you seem to be missing the point. The OP said if you have physical access, then you have access to the data anyway. That is not necessarily true, even when not including any third-party security features.
I'd imagine that if your "average cop shop" as you put it finds out they can't read the files they'll send it either to their state lab or to some other higher agency that has expert knowledge in the subject of computer forensics. If they have reason to suspect your hard drive has data they want they'll try to get it.
Sure they will, but assuming you picked a strong password, they may not be able to because it would take too much computing power to break the encryption. My point was physical access != ownage, even with stock systems using only the built in tools.
Side note: I recently built a new rig for gaming and have no issues with Vista. In fact I kinda like it.
Vista has some interesting new features as well as some anti-features. For the most part, it seems okay, but as of my last test still had too many issues with old software (including 1 show stopper for my work) and it seemed prudent to wait for the ecosystem to stabilize. I, also, don't see its lack of functionality preventing it from taking over from XP in the next 5 years, although the bad mainstream press and lack of benefits for enterprise customers may slow it down.
He said that if one is going to use one's computer as an aid to their criminal career, use a Mac. The RCMP and all the rest were completely ignorant when it came to the Mac OS as well as everything else not Windows.
This is interesting. An FBI guy told me they shipped all the Macs they seized to the RCMP, who had staff experienced in analyzing them, whereas the FBI did not.
Dude. Even Windows (Vista) supports encrypting your disk these days. Assuming it was turned off when seized, that does not mean your data has been compromised or is realistically recoverable, especially by your average cop shop.
This article poses a question I've always wondered about. Do most criminal investigations of the computer-related nature have experts that are well-versed in multiple operating systems?
From what I've seen, no. According to an FBI guy I know, as of a few years ago when the FBI found a Mac during an investigation, they shipped it to the RCMP (canadian mounties) for analysis. There is also a fairly well known computer forensics program at the university nearby (one of the largest of such programs in the country). They do cover Linux and NTFS but very sparsely. Most of the Linux stuff is about setting up a and using a Linux box as an investigative tool, not investigating other Linux machines. They don't even mention OS X, *BSD, Solaris, etc. in any of their class descriptions. They do mention DOS.
i don't see the usability flaw in the ubuntu software package manager.
Okay, I'm happy to go over it.
you say it takes longer because i have to use a different application to install the software than to find it
This is correct, it adds another series of steps to the installation process.
...quite apart from the fact that many webpages offer ubuntu compatible debs for download.
Some do, but many don't. Many offer package formats used by other package managers or tar.gz files. Ubuntu's package manager does not handle either. Because of this, most users won't know which link to click if there are multiple ones and won't know if it will work if there is just one. On top of this, many pages offer stand alone binary installers. The usability problem is that there are many different mechanisms, most of which are not as usable as just clicking the one and only binary installer for Windows.
This problem can be mostly solved by making sure the Web browser hands off all such packages to the package manager and that the package manager handles all the formats. Alternately, all the Linux distros could standardize on one format that is easily linkable from a Web page and make it easy for developers offering software to just offer one version for Linux instead of having to try to make a bunch of different packages every time.
...the advantages of apt are legion...
There are many advantages to using a package manager in general, but those advantages don't remove the usability problems noted here. The goal should be to remove all the usability problems, those that are solved by apt today as well as those that are not solved. Ignoring the latter by claiming you have the former, misses the point and is just a way of ignoring the problem.
...while as the mac "software installer" is a huge security risk.
The security risk of running a binary installer is a nother problem to be solved, not only on the Mac, but also on Linux and Windows. Almost all commercial software on Linux and Windows ships as a binary installer which is a significant security risk. Linux mitigates this problem for non-commercial software using repositories and package managers, but have not made it easy to solve the problem for commercial developers. Commercial developers need more than just a bare bones installer. They need to do licensing and registration of the software, updates, and process payments. None of the package managers I've used on Linux have taken on the challenge of providing that functionality for commercial developers, so said developers tend to use existing tools to build binary installers that also happen to ship with solutions to those problems. The end result is Linux has to deal with the same security problem as OS's that don't have package managers at all.
I'd note that OS X has partially solved this problem as well with drag and drop software bundles that don't require installation and thus remove that security risk for software that uses that (and pretty much anything not adding a kernel module can use it). OS X fails to provide an official service for licensing, registration, updates and payments as well, so many developers don't use their drag and drop bundles for installation either, and use binary installers like everyone else.
In short, there is a lot of room for improvement from all vendors. MS and Apple should be introducing package managers to solve some of the usability problems that are more or less solved on Linux, but at the same time Linux distros need to beef up their package managers to solve problems introduced by the current generation of package managers, as well as problems solved by other vendors (like Apple) with their OpenStep style packages.
I see little or no value in looking to see where your favorite OS is better than competing offerings. I see a lot of value in seeing where it is worse and borrowing solutions from other vendors and I see a lot of value in seeing where no one gets it right, recognizing that there is room to improve, and proposing solutions.
Or, she could have just opened up add/remove programs and installed it with one click.
Her task was to view YouTube, not install Flash. Given that, she was already on the Adobe Web page with one click. Install via a link seems like a reasonable expectation of functionality. Ubuntu failed at that functionality.
The problem isn't that Ubuntu is hard to use, it's just different from Windows.
The problem is that Ubuntu is hard to use for certain workflows... and those workflows are different from the ones Windows is good at.
I think that complete computer newbies are better off with Ubuntu than Windows "power users" for that very reason.
She wasn't a computer newbie though, she was an average user, you know the main group Linux on the desktop advocates need to target to grow market share. It seems like a valid failing, and one that should be fixed if people can get over being all defensive of their favorite OS and what it implies about them.
Not the topic of discussion? What's the freaking subject line of these posts? "Why MS...MUST CONTROL...".
I think you mean, "Why MS and textbook publishers must control OLPC." OLPC is not media. MS uses DRM to lock people into their OS. Content publishers use DRM to lock people into their control channels and for planned obsolescence. MS needs to control OLPC for the same reason they implement DRM to keep people locked into their OS, partly through having DRM'd content that will only work on Windows.
Large-scale developers of proprietary software know precisely where their bread is buttered, and they attend to all this as a matter of course. yes! just look at the most recent smashing usability success put out by one of those proprietary companies. VISTA!
MS is an exception to the rule. They don't compete based upon features the way other software has to. Most major software developers certainly do make use of usability testing to some degree. That said, some Linux developers do usability testing as well, just not for the entire distro that I've seen.
in context, ubuntu is doing a *fantastic* job of making ubuntu usable for ordinary people.
The Ubuntu distro is facing a serious uphill battle. They have made strides, but it is clear they don't do significant usability testing or even apply usability guidelines to software before it is included.
I'm not writing this to belittle Ubuntu. I use it daily and it is fairly functional. It does have a lot of flaws, however, and it is a very good thing to expose them and discuss them and hopefully fix them.
I think I'm going to have to sit my boyfriend down in front of the computer for the same test.
Good idea.
Why is it always, "to test the usability of X, we will use the lowest common denometer of computer knowledge- the woman"?
This was an anecdote. Would you prefer they remove references to the gender of their significant other and they try to awkwardly work around the lack of a gender neutral pronoun for the article? I can understand feeling threatened by the lack of women in technology as represented by Slashdot, but that doesn't make it sexist. No one was claiming she was technically non-proficient because of her gender.
I'm sure there are other women lined up to complain about the periodical "Look, I tested usability on my girlfriend" Slashdot story.
Why? I see plenty such anecdotes referring to their father or grandparent as well. I've seen anecdotes in this very discussion about testing usability on non-proficient business executives and lawyers.
When my bf or family/friends have computer trouble they call me. All these stories have a thick sexist undercurrent.
While their certainly are comments here with more than a sexist undercurrent, I don't see how the article itself is sexist. You're assuming they chose their girlfriend because they prejudged that girlfriend as not technically savvy, but that is your assumption, probably based upon your being overly sensitive to the idea. Not many women get into computing these days. That is simply the objective facts of the matter. Implying this is because they are predisposed directly by their gender or that any given woman will not be into computing is what would be sexist.
there are a number of things missing from your mac software installation, for example:
search around the internet site of the company offering the software until you find a download button...
I actually performed this exercise so I would not miss any steps. The download link was right on the main page for the software, and certainly looked at the main page for both the Linux and OS X programs.
...download a file 3 times as large as it should be because it needs to contain versions of every library used by the application...
The download of the application for OS X was slightly slower than the Linux one, but only by a second or so and that step was by no means the most time consuming. Given that disk space and download speeds are a pretty small concern for this workflow, I'd also argue the benefits of the OS X package format are a big win. For example, it is portable and I can use it from a network drive on both my Intel based laptop and my old PPC based media server.
these libraries and the application will not be automatically updated so turning the software into a security risk
This is not exactly true due to the dynamic binding OS X uses, but keeping software up to date is a real win of package managers (when an application is in the repository). But that isn't the issue. The issue is that Ubuntu is inferior for the workflow described. Sure it is better at other workflows, including keeping many of the userspace applications up to date. That in no way mitigates the original point. As I said, this isn't a "which OS is best" contest. It is addressing real usability flaws in all OS's. Ignoring the usability flaws present while talking about other usability wins is not at all productive.
should i carry on?
If you can stick to the point, please do. Tell me, why you feel this usability flaw that was a problem in the rather informal test, is not really a flaw and should be ignored... Your rationalizations should be quite amusing.
I absolutely understand that Microsoft profits from DRM, but they wouldn't profit from DRM if there wasn't anybody more willing to sell encumbered media than they were to sell unencumbered media.
In some cases Microsoft is doing the selling. In other cases MS has profited from adding DRM to content themselves, such as adding DRM by default to songs users ripped themselves from their CD collection or providing adding DRM to office documents.
I'm not saying that they are the good guys, I'm saying that the interpretation that they want to control all media isn't supported by their behavior so far.
Perhaps so, perhaps not, but that was not the original topic of discussion here. MS pushed DRM for profit and works hand in glove with those content producers even when it hurts their customers. Whether their long term goals are the same as the current media producers or not, is simply speculation.
Agnostic to DRM simply means that they are perfectly willing to provide the technology, but they are also perfectly willing to allow playback of unmanaged files.
Agnostic normally means you neither support nor hinder any given DRM. MS certainly does support and advocate particular DRM.
Name a Microsoft DRM technology that is not allowed to be used on a device or platform that supports playback of unencumbered files.
How about playing PlaysForSure files on Linux. MS is about making suer that if you're using any device, they make sure it gives you motivation to use Windows on your computer... this profits them.
I understand that people see the potential for bait and switch, but guess what, that's why I don't buy media with onerous DRM(I buy a DVD once in a while).
It isn't even a matter of bait and switch. When you sell music from a store you partly own, using DRM you created called "playsforsure" and then you intentionally make sure it doesn't work if the person tries to move away from Windows that is misleading people from the start. When you then cancel support for that DRM service in favor of your new, incompatible one and don't let people move their music to it, that is bait and switch and there is more than potential of that.
I couldn't care less if the devices and software I own support DRM or not, but I won't pay very much for DRM media.
Nonetheless you have to understand that MS is directly profiting from DRM and advocating it. It's not like they're being forced into it by content producers. They have the leverage to stop HDCP, and they have intentionally gone ahead with it.
Microsoft's behavior has been DRM agnostic much of the time.
I don't think this is a supportable assertion. They created several of the most popular DRM schemes in use.
I'm pretty sure that they see a formidable business case for cow-towing[sic] to big content producers
I disagree. I think they see a business case for kowtowing to the big content producers, but also see an opportunity to protect their monopolies by further making it more difficult for consumers to migrate away, since their data is locked into MS proprietary DRM and formats.
(and given that HDCP is an industry wide attack on the consumer, it's hard to argue for singling out Microsoft for supporting it
Two points here, so far Apple has resisted implementing HDCP, although they may have to comply eventually for their OS, if not for their iPod. Microsoft, however, is the one player with the ability to stand up and refuse HDCP as being bad for their customers, and be able to pull it off. It is not like content producers can afford to lose Windows users, let alone Windows and xBox users combined. MS has chosen, however, to go along with it because it benefits MS as well (at the expense of MS's customers).
In a world where most consumers don't seem to care about their rights, I'm not surprised that they are failing to use their market position as a lever to support consumer rights.
Consumers care about their abilities. They care about their music not working when they switch players or computers. Usually, however, they find that it is too late by the time they discover these limitations. There have been numerous lawsuits after the fact but intentionally misleading advertising combined with the complexity of the schemes is enough to prevent the average consumer from making an informed decision, and unless they are accurately informed, the capitalist free market cannot correct for the problem (assuming it was in operation, which is a bit of a stretch given the collusion of a company with a convicted antitrust abusing monopoly on desktop OS's colluding with several cartels, likewise convicted of antirust abuses).
the trouble with your griping is, the package manager on a gnu/linux distribution is vastly superior to any way osx or windows has to install software.
No, installing software with a a package manager is vastly superior for installing a subset of software. It is inferior for installing other software in other situations. This isn't a contest anyway, it is about pointing out the real deficiencies and addressing them instead of trying to reassure you about your insecurities and that you're not an idiot for using Linux
just because you like things to be needlessly complex doesn't mean that the many million ubuntu users also like that.
Use case: I want to find and install a program to make stop action movies with my Webcam.
Ubuntu procedure:
Open Firefox
search for "linux stop motion programs"
read reviews of several programs until I find the one I want to try "stopmotion"
Go to "Applications:Add/Remove"
search for stopmotion
change the default search from "supported applications" to "all applications"
search again
select the package
click "enable"
click "apply changes"
click "apply"
enter my password
click "close"
Select "Applications: Graphics: Stopmotion" from the menu
Mac OS X procedure:
Open Firefox
search for "OS X stop motion programs"
read reviews of several programs until I find the one I want to try "istopmotion"
Click the "download" link on the page.
double click on the.dmg icon that is in the downloads list that pops up.
double click on the iStopMotion icon (or drag it to the Applications and then do so if I plan on keeping it for sure).
click "ok" to run a new program from the internet
So the above provide two procedures for a very, very common workflow for finding and installing a program. For said use case, Ubuntu really doesn't win on the usability, but hey it isn't too bad and neither are as good as Windows for usability for this workflow. When you look at other workflows, however, like if your friend has a copy of a program you want installed on his laptop, but which is not being distributed anymore by the manufacturer and your friend is on IM, well Linux really falls down compared to OS X, but is way ahead of Windows. If you look at the use case of software distributed on a DVD or software you want to run off of a portable flash drive, or if you want to run software that is not in a repository and is commercial and needs to b registered, Linux is likewise behind.
For most of these cases Ubuntu is more complex and less usable. Linux has a big win with the functionality offered by package managers, but for the most part that seems to have blinded developers to the big losses they have in other areas of application installation. Linux is not inferior, it just has different weaknesses and referencing its strengths do nothing to mitigate those weaknesses nor ameliorate the problems of novice users such as described in this article.
The problem is that this approach often doesn't work. For one thing, it annoys the piss out of experience users.
Experienced users can find the preference to turn them off.
For another thing, new users tend to ignore most of that information... mainly because they are being overwhelmed by new information and can't possibly assimilate it all.
A lot of this depends upon the presentation of the information. It can be done well and has been in other products.
Take, for instance, the problem that was encountered when changing screen resolution. The tester changed the resolution easily, but then she clicked the "Keep settings" immediately, which locked her into graphic settings that were hard to change back. Part of the problem, I suppose is that the system allowed the user to make a ridiculous change.
The problem was not that the user was allowed to change the resolution to very low, but that the UI did not scale the window or provide scroll bars so the parts off the screen were accessible.
But part of the problem is also, perhaps, that the user is very used to clicking "OK" on any dialog that gets in the way: there are too many new things to read and learn, and the easiest way to get things done (in the mind of a new user) is to dismiss those annoying boxes as quickly as possible.
The OK/Cancel flaw is a well known and documented usability snafu. Users should never be presented with the same two choices over and over again or presented with dialogue boxes unnecessarily. Dialogue boxes should be rare and should always include actual verbs as button names. If you present a dialogue that says "would you like to keep this resolution (OK)(Cancel)" then you've already failed from a usability perspective. A proper dialogue box would read, "Would you like to keep this resolution (Revert to old screen resolution)(Keep new screen resolution)."
Most users would just click "OK" without reading it.
User should never be given such an option. Make sure the buttons themselves are actions and the user has to read at least one and thus understand what they are about to do. Even if they click randomly, at least they are not being operant conditioned to click "OK" in future.
All this to say that I'm by no means convinced that adding more balloons, wizards, and dialog boxes will magically make it easier for users to figure out what's going on.
Agreed. A good start would be following some of the best practices for UI design, which have been ignored in both Ubuntu and major applications. The next step is proper usability testing where you find the pain points and then find solutions for them and test them to see what does work.
Anyway - most of the 'problems' were with proprietary software. That is the fault of the proprietary vendors - how about laying blame where it belongs.
People truly interested in making Linux better don't care about blame. They care about solving the problems for normal users. Who cares if Adobe is at fault if users still have problems and go back to Windows? The correct question is, how can things be made easier for Adobe or how can Linux compensate for what Adobe does, such that users don't have problems. This might be standardizing on one package format so Adobe doesn't have to make special provisions for each Linux distro or it might be seeing what Adobe is giving the user and firing up the package manager to grab the proper version for Ubuntu instead.
A bare bones windows install doesn't even have half of those programs - she'd be pretty proper fucked on windows doing all those things if someone hadn't loaded her box with 'pirated' shit to start with.
People don't have bare bones Windows installs and there is no evidence to suggest her copy of photoshop is pirated. Some people don't mind paying for quality software. The trick is making sure they have that option on Linux instead of being limited to just OSS in the repositories.
Some of the comments of the author indicate he doesn't really know what's going on either...
Why should he be an expert in every software application. He just presented some common tasks, tasks his GF was likely to want to accomplish, and saw how easy they were and where they could be made easier for new users.
How is it a mess? If someone or something tells you need flash. Synaptics -> Search Flash -> Install -> Restart Firefox, and you are set.
That is fine, except it is a longer process that just clicking a link on a Web page, which, realistically, is where a lot of us are when we want to install software. People generally don't decide they need to install a Flash plug-in or anything else. They need to solve a problem so they look for a solution, usually with their Web browser. What if she wanted to install an application to make stop action movies with her Web cam. should she go to the package manager and look or should she go to her Web browser and search for different ones and see which is most suited to her needs?
The issue is further complicated by the fact that many Web sites do provide packages for download, so unlike Windows you're asking users to learn two different methods for installing software. Further, there are usability issues with installing software from a CD, running it from your flash drive, getting packages that are for a different Linux distro or which are tar.gz files, installing software that is on a friend's computer, and installing commercial software packages.
Now I'm the first to tell others about the benefits of package managers and I think Apple and MS should get off their butts and create proper package managers for their OS's. That said, package managers do not solve all the problems and current package managers are very poor at many common use cases (anything other than find and install software that is in the repository when you already know what that software is). Ignoring these usability problems and use cases, will not benefit Linux's usability or desktop market share.
I'd say usability testing like what was done in this article is the ONLY way to succeed, if you're going to proceed with a true goal of "ease of use". Anyone savvy enough to help build an OS is FAR too removed from comprehending all the points of confusion the "average user" might run across.
I've performed usability tests in the past and regardless of if you're a coder or a usability expert, they almost always reveal usability issues that seem unintuitive or which people want to "explain away" because they can't wrap their heads around what is happening. Usability testing is, however, a very underused scientific process in computing today.
That said, over the years a lot of general guidelines have emerged that can help designers avoid some of the really common mistakes that happen over and over and over again. From my own use of Ubuntu, it is clear there has not even been consistent use of those guidelines to get the low hanging fruit before going to the expense of a real usability test. There is much work to be done.
Personally, I think "file extensions" are a pretty evil concept, the way most operating systems today handle them.
They are, indeed, a common usability problem, especially when one goes into hidden extensions and multiple extensions. Expecting users to know what three letter combination mean "program that does stuff" and which ones mean "data you can read or see or hear" is absurd, especially when they can only sometimes see them and what they see is not always the "real" extension. It is certainly an area where OS's could be significantly improved.
applications; add/remove; select an app from the list; apply changes.
Except that only works for some software. Not all software is in the repositories.
try telling a true new user (not someone who has used windows for a while) to install a new program on windows. where do they begin? hint: probably not with 'internet explorer'.
Actually most people I know start all unfamiliar computing tasks from the browser and usually from Google. This includes finding and installing software. Heck, as a Linux user when I need a new application I don't go to the package manager to look through it. I go to Firefox and search for applications that do what I want, preferably with reviews.
...deb packages and rpm files are easily as 'intuitive' as window's 'download from a site...
Except not all Linux software is distributed as.deb and rpms, much of it is yum or tar.gz or something else, a usability issue Windows users don't have to deal with and which most Linux distros have ignored.
...they aren't yet as common, but neither are osx packages, so does that make osx packages harder to install than windows installers?
No, how common it is doesn't effect how usable it is. OS X packages are harder to install for some use cases though. They're easier to drag and drop from media, but there is an extra step when installing from the Web because they are distributed within.dmg files for the most part.
I think you're missing the point of this usability study. It exposed real problems with using Ubuntu today. It doesn't matter if other OS's also have problems. The only consideration of other OS's should really be to see if they have solved that usability problem and if so, can that solution be used in Ubuntu.
Have you seen the Web Acess client? There's NOTHING out there that compares.
Okay, I've used the Exchange Web access client and, well I was less than impressed. One of the people we did business with included a PGP public key in his signature. He's a security professional and and academic, so that's pretty acceptable and he sure wasn't going to stop doing it just for his e-mails with us. The problem, whenever anyone tried to view that message with the Exchange Webmail client, it would lock up their account entirely and they would be unable to send or receive e-mail using the Webmail or normal client until an admin would log in and delete that message from their account manually, without looking at it. We were all less than impressed, let me tell you.
That was by no means the only recurring problem we had, but it was one of the most common. And don't even get me started on Sharepoint. Both these led to strict "No MS server products" rules without a formal evaluation from IT, no more departmental IT introductions of this crap.
Slashdot Mirror
You're saying when running a charity, Microsoft should be "allowed to compete"?!? This is the same Microsoft that has been repeatedly convicted of undermining fair competition through criminal antitrust abuses? This is the same Microsoft that is still in the process of being prosecuted for ongoing antitrust abuses? This is the same Microsoft that is being investigated for bribing government officials and standards bodies?
I say let them, they are going to have to up the hardware costs of the machine to get an XP port running, and it will inevitably be significantly less functional than what is already available for less money.Great, then the OLPC brand is poisoned and there is confusion about what the capabilities of the different versions are. Then MS can undercut others on price in order to lock in a new market early and then bleed them for the next twenty years like they have been other markets. I'm less than impressed with this idea.
I sincerely believe that a foray into M$ for the OLPC will bring to light the inherent advantages of free software.You're assuming people will try both and objectively compare them on a level playing field, then choose what is best for the kids. Given MS's history, their piles of cash, and their incentive here, why do you think that?
We've paid more per person in tax subsidies than many other nations. Take Sweden, for example. Their population density and median population density are both about the same as the US. Their subsidies, however, had legal teeth that required the telcos to actually provide something in exchange. They also had a huge embezzling scandal where much of the money was stolen. They still have significantly faster internet at significantly lower prices than the US, in exchange for a smaller per person tax.
The high speed internet problem comes down to pretty much the same thing as many other problems in the US. Politicians are willing to give private companies billions is subsidies, in exchange for hundreds of thousands being returned as campaign contributions. So long as this legalized bribery is allowed, companies will simply pay off politicos in exchange for subsidies or for not having to fulfill the agreements they made when the subsidies were given.
Yes. Dell needs to have multiple vendors in their pipeline so they can get them to compete where they really care, on price, as Dell usually buys the cheapest part X available on a given day from whomever. At least for consumer desktops, that is how they operate. I seriously doubt open source drivers are going to be a consideration on the low end and they sure haven't been in the past (speaking as someone who ordered hundreds of the same model from them only to find a wide variety of parts actually inside, some of which had OSS drivers like the test model we bought and some of which did not). More likely, Dell will push for this on midrange and higher end software where they actually offer Linux as a supported option and probably only then for machines targeted at business.
All it takes is for one company to start releasing open drivers and the rest will have to fall in line or risk loosing[sic] a lot of business practically overnight. In the end, everyone will benefit.Except it actually does cost money to release drivers as OSS if you have to clean up the code or get rid of the hacks you use to work around your hardware flaws. In which case, it potentially amounts to costs raising for the first movers and people who don't plan to use OSS drivers still have to cover that cost (although it might be negligible on such a scale).
Reasons include: they don't like providing anything they do for free because a competitor might use it, they don't want to expose their embarrassingly poorly written code, they're afraid their poorly written code will expose their security flaws, they don't want consumers to know about the hacks they use to work around hardware flaws or which compromise quality for speed.
It won't necessarily be, however it does present a situation that falsifies the assertion I was addressing. Physical access does not necessarily mean access to the data.
It's a bigger assumption that your password or encryption key hasn't been paged out to virtual memory and then never cleared...Those were issues with early versions of hard disk encryption built into OS's, but has been addressed by Filevault (OS X), Bitlocker (Vista), and TrueCrypt (Linux). Basically, unless the machine is in hibernate mode, or was shut down less than 5 minutes ago, no that isn't really a concern.
...and that you haven't reused your password for a less-secure purpose.No. That isn't an assumption for me.
Anyways, you seem to be missing the point. The OP said if you have physical access, then you have access to the data anyway. That is not necessarily true, even when not including any third-party security features.
Sure they will, but assuming you picked a strong password, they may not be able to because it would take too much computing power to break the encryption. My point was physical access != ownage, even with stock systems using only the built in tools.
Side note: I recently built a new rig for gaming and have no issues with Vista. In fact I kinda like it.Vista has some interesting new features as well as some anti-features. For the most part, it seems okay, but as of my last test still had too many issues with old software (including 1 show stopper for my work) and it seemed prudent to wait for the ecosystem to stabilize. I, also, don't see its lack of functionality preventing it from taking over from XP in the next 5 years, although the bad mainstream press and lack of benefits for enterprise customers may slow it down.
This is interesting. An FBI guy told me they shipped all the Macs they seized to the RCMP, who had staff experienced in analyzing them, whereas the FBI did not.
Dude. Even Windows (Vista) supports encrypting your disk these days. Assuming it was turned off when seized, that does not mean your data has been compromised or is realistically recoverable, especially by your average cop shop.
From what I've seen, no. According to an FBI guy I know, as of a few years ago when the FBI found a Mac during an investigation, they shipped it to the RCMP (canadian mounties) for analysis. There is also a fairly well known computer forensics program at the university nearby (one of the largest of such programs in the country). They do cover Linux and NTFS but very sparsely. Most of the Linux stuff is about setting up a and using a Linux box as an investigative tool, not investigating other Linux machines. They don't even mention OS X, *BSD, Solaris, etc. in any of their class descriptions. They do mention DOS.
Okay, I'm happy to go over it.
you say it takes longer because i have to use a different application to install the software than to find itThis is correct, it adds another series of steps to the installation process.
...quite apart from the fact that many webpages offer ubuntu compatible debs for download.Some do, but many don't. Many offer package formats used by other package managers or tar.gz files. Ubuntu's package manager does not handle either. Because of this, most users won't know which link to click if there are multiple ones and won't know if it will work if there is just one. On top of this, many pages offer stand alone binary installers. The usability problem is that there are many different mechanisms, most of which are not as usable as just clicking the one and only binary installer for Windows.
This problem can be mostly solved by making sure the Web browser hands off all such packages to the package manager and that the package manager handles all the formats. Alternately, all the Linux distros could standardize on one format that is easily linkable from a Web page and make it easy for developers offering software to just offer one version for Linux instead of having to try to make a bunch of different packages every time.
...the advantages of apt are legion...There are many advantages to using a package manager in general, but those advantages don't remove the usability problems noted here. The goal should be to remove all the usability problems, those that are solved by apt today as well as those that are not solved. Ignoring the latter by claiming you have the former, misses the point and is just a way of ignoring the problem.
...while as the mac "software installer" is a huge security risk.The security risk of running a binary installer is a nother problem to be solved, not only on the Mac, but also on Linux and Windows. Almost all commercial software on Linux and Windows ships as a binary installer which is a significant security risk. Linux mitigates this problem for non-commercial software using repositories and package managers, but have not made it easy to solve the problem for commercial developers. Commercial developers need more than just a bare bones installer. They need to do licensing and registration of the software, updates, and process payments. None of the package managers I've used on Linux have taken on the challenge of providing that functionality for commercial developers, so said developers tend to use existing tools to build binary installers that also happen to ship with solutions to those problems. The end result is Linux has to deal with the same security problem as OS's that don't have package managers at all.
I'd note that OS X has partially solved this problem as well with drag and drop software bundles that don't require installation and thus remove that security risk for software that uses that (and pretty much anything not adding a kernel module can use it). OS X fails to provide an official service for licensing, registration, updates and payments as well, so many developers don't use their drag and drop bundles for installation either, and use binary installers like everyone else.
In short, there is a lot of room for improvement from all vendors. MS and Apple should be introducing package managers to solve some of the usability problems that are more or less solved on Linux, but at the same time Linux distros need to beef up their package managers to solve problems introduced by the current generation of package managers, as well as problems solved by other vendors (like Apple) with their OpenStep style packages.
I see little or no value in looking to see where your favorite OS is better than competing offerings. I see a lot of value in seeing where it is worse and borrowing solutions from other vendors and I see a lot of value in seeing where no one gets it right, recognizing that there is room to improve, and proposing solutions.
Her task was to view YouTube, not install Flash. Given that, she was already on the Adobe Web page with one click. Install via a link seems like a reasonable expectation of functionality. Ubuntu failed at that functionality.
The problem isn't that Ubuntu is hard to use, it's just different from Windows.The problem is that Ubuntu is hard to use for certain workflows... and those workflows are different from the ones Windows is good at.
I think that complete computer newbies are better off with Ubuntu than Windows "power users" for that very reason.She wasn't a computer newbie though, she was an average user, you know the main group Linux on the desktop advocates need to target to grow market share. It seems like a valid failing, and one that should be fixed if people can get over being all defensive of their favorite OS and what it implies about them.
I think you mean, "Why MS and textbook publishers must control OLPC." OLPC is not media. MS uses DRM to lock people into their OS. Content publishers use DRM to lock people into their control channels and for planned obsolescence. MS needs to control OLPC for the same reason they implement DRM to keep people locked into their OS, partly through having DRM'd content that will only work on Windows.
MS is an exception to the rule. They don't compete based upon features the way other software has to. Most major software developers certainly do make use of usability testing to some degree. That said, some Linux developers do usability testing as well, just not for the entire distro that I've seen.
in context, ubuntu is doing a *fantastic* job of making ubuntu usable for ordinary people.The Ubuntu distro is facing a serious uphill battle. They have made strides, but it is clear they don't do significant usability testing or even apply usability guidelines to software before it is included.
I'm not writing this to belittle Ubuntu. I use it daily and it is fairly functional. It does have a lot of flaws, however, and it is a very good thing to expose them and discuss them and hopefully fix them.
Good idea.
Why is it always, "to test the usability of X, we will use the lowest common denometer of computer knowledge- the woman"?This was an anecdote. Would you prefer they remove references to the gender of their significant other and they try to awkwardly work around the lack of a gender neutral pronoun for the article? I can understand feeling threatened by the lack of women in technology as represented by Slashdot, but that doesn't make it sexist. No one was claiming she was technically non-proficient because of her gender.
I'm sure there are other women lined up to complain about the periodical "Look, I tested usability on my girlfriend" Slashdot story.Why? I see plenty such anecdotes referring to their father or grandparent as well. I've seen anecdotes in this very discussion about testing usability on non-proficient business executives and lawyers.
When my bf or family/friends have computer trouble they call me. All these stories have a thick sexist undercurrent.While their certainly are comments here with more than a sexist undercurrent, I don't see how the article itself is sexist. You're assuming they chose their girlfriend because they prejudged that girlfriend as not technically savvy, but that is your assumption, probably based upon your being overly sensitive to the idea. Not many women get into computing these days. That is simply the objective facts of the matter. Implying this is because they are predisposed directly by their gender or that any given woman will not be into computing is what would be sexist.
I actually performed this exercise so I would not miss any steps. The download link was right on the main page for the software, and certainly looked at the main page for both the Linux and OS X programs.
...download a file 3 times as large as it should be because it needs to contain versions of every library used by the application...The download of the application for OS X was slightly slower than the Linux one, but only by a second or so and that step was by no means the most time consuming. Given that disk space and download speeds are a pretty small concern for this workflow, I'd also argue the benefits of the OS X package format are a big win. For example, it is portable and I can use it from a network drive on both my Intel based laptop and my old PPC based media server.
these libraries and the application will not be automatically updated so turning the software into a security riskThis is not exactly true due to the dynamic binding OS X uses, but keeping software up to date is a real win of package managers (when an application is in the repository). But that isn't the issue. The issue is that Ubuntu is inferior for the workflow described. Sure it is better at other workflows, including keeping many of the userspace applications up to date. That in no way mitigates the original point. As I said, this isn't a "which OS is best" contest. It is addressing real usability flaws in all OS's. Ignoring the usability flaws present while talking about other usability wins is not at all productive.
should i carry on?If you can stick to the point, please do. Tell me, why you feel this usability flaw that was a problem in the rather informal test, is not really a flaw and should be ignored... Your rationalizations should be quite amusing.
In some cases Microsoft is doing the selling. In other cases MS has profited from adding DRM to content themselves, such as adding DRM by default to songs users ripped themselves from their CD collection or providing adding DRM to office documents.
I'm not saying that they are the good guys, I'm saying that the interpretation that they want to control all media isn't supported by their behavior so far.Perhaps so, perhaps not, but that was not the original topic of discussion here. MS pushed DRM for profit and works hand in glove with those content producers even when it hurts their customers. Whether their long term goals are the same as the current media producers or not, is simply speculation.
Agnostic normally means you neither support nor hinder any given DRM. MS certainly does support and advocate particular DRM.
Name a Microsoft DRM technology that is not allowed to be used on a device or platform that supports playback of unencumbered files.How about playing PlaysForSure files on Linux. MS is about making suer that if you're using any device, they make sure it gives you motivation to use Windows on your computer... this profits them.
I understand that people see the potential for bait and switch, but guess what, that's why I don't buy media with onerous DRM(I buy a DVD once in a while).It isn't even a matter of bait and switch. When you sell music from a store you partly own, using DRM you created called "playsforsure" and then you intentionally make sure it doesn't work if the person tries to move away from Windows that is misleading people from the start. When you then cancel support for that DRM service in favor of your new, incompatible one and don't let people move their music to it, that is bait and switch and there is more than potential of that.
I couldn't care less if the devices and software I own support DRM or not, but I won't pay very much for DRM media.Nonetheless you have to understand that MS is directly profiting from DRM and advocating it. It's not like they're being forced into it by content producers. They have the leverage to stop HDCP, and they have intentionally gone ahead with it.
I don't think this is a supportable assertion. They created several of the most popular DRM schemes in use.
I'm pretty sure that they see a formidable business case for cow-towing[sic] to big content producersI disagree. I think they see a business case for kowtowing to the big content producers, but also see an opportunity to protect their monopolies by further making it more difficult for consumers to migrate away, since their data is locked into MS proprietary DRM and formats.
(and given that HDCP is an industry wide attack on the consumer, it's hard to argue for singling out Microsoft for supporting itTwo points here, so far Apple has resisted implementing HDCP, although they may have to comply eventually for their OS, if not for their iPod. Microsoft, however, is the one player with the ability to stand up and refuse HDCP as being bad for their customers, and be able to pull it off. It is not like content producers can afford to lose Windows users, let alone Windows and xBox users combined. MS has chosen, however, to go along with it because it benefits MS as well (at the expense of MS's customers).
In a world where most consumers don't seem to care about their rights, I'm not surprised that they are failing to use their market position as a lever to support consumer rights.Consumers care about their abilities. They care about their music not working when they switch players or computers. Usually, however, they find that it is too late by the time they discover these limitations. There have been numerous lawsuits after the fact but intentionally misleading advertising combined with the complexity of the schemes is enough to prevent the average consumer from making an informed decision, and unless they are accurately informed, the capitalist free market cannot correct for the problem (assuming it was in operation, which is a bit of a stretch given the collusion of a company with a convicted antitrust abusing monopoly on desktop OS's colluding with several cartels, likewise convicted of antirust abuses).
No, installing software with a a package manager is vastly superior for installing a subset of software. It is inferior for installing other software in other situations. This isn't a contest anyway, it is about pointing out the real deficiencies and addressing them instead of trying to reassure you about your insecurities and that you're not an idiot for using Linux
just because you like things to be needlessly complex doesn't mean that the many million ubuntu users also like that.Use case: I want to find and install a program to make stop action movies with my Webcam.
Ubuntu procedure:
Mac OS X procedure:
So the above provide two procedures for a very, very common workflow for finding and installing a program. For said use case, Ubuntu really doesn't win on the usability, but hey it isn't too bad and neither are as good as Windows for usability for this workflow. When you look at other workflows, however, like if your friend has a copy of a program you want installed on his laptop, but which is not being distributed anymore by the manufacturer and your friend is on IM, well Linux really falls down compared to OS X, but is way ahead of Windows. If you look at the use case of software distributed on a DVD or software you want to run off of a portable flash drive, or if you want to run software that is not in a repository and is commercial and needs to b registered, Linux is likewise behind.
For most of these cases Ubuntu is more complex and less usable. Linux has a big win with the functionality offered by package managers, but for the most part that seems to have blinded developers to the big losses they have in other areas of application installation. Linux is not inferior, it just has different weaknesses and referencing its strengths do nothing to mitigate those weaknesses nor ameliorate the problems of novice users such as described in this article.
Experienced users can find the preference to turn them off.
For another thing, new users tend to ignore most of that information... mainly because they are being overwhelmed by new information and can't possibly assimilate it all.A lot of this depends upon the presentation of the information. It can be done well and has been in other products.
Take, for instance, the problem that was encountered when changing screen resolution. The tester changed the resolution easily, but then she clicked the "Keep settings" immediately, which locked her into graphic settings that were hard to change back. Part of the problem, I suppose is that the system allowed the user to make a ridiculous change.The problem was not that the user was allowed to change the resolution to very low, but that the UI did not scale the window or provide scroll bars so the parts off the screen were accessible.
But part of the problem is also, perhaps, that the user is very used to clicking "OK" on any dialog that gets in the way: there are too many new things to read and learn, and the easiest way to get things done (in the mind of a new user) is to dismiss those annoying boxes as quickly as possible.The OK/Cancel flaw is a well known and documented usability snafu. Users should never be presented with the same two choices over and over again or presented with dialogue boxes unnecessarily. Dialogue boxes should be rare and should always include actual verbs as button names. If you present a dialogue that says "would you like to keep this resolution (OK)(Cancel)" then you've already failed from a usability perspective. A proper dialogue box would read, "Would you like to keep this resolution (Revert to old screen resolution)(Keep new screen resolution)."
Most users would just click "OK" without reading it.User should never be given such an option. Make sure the buttons themselves are actions and the user has to read at least one and thus understand what they are about to do. Even if they click randomly, at least they are not being operant conditioned to click "OK" in future.
All this to say that I'm by no means convinced that adding more balloons, wizards, and dialog boxes will magically make it easier for users to figure out what's going on.Agreed. A good start would be following some of the best practices for UI design, which have been ignored in both Ubuntu and major applications. The next step is proper usability testing where you find the pain points and then find solutions for them and test them to see what does work.
People truly interested in making Linux better don't care about blame. They care about solving the problems for normal users. Who cares if Adobe is at fault if users still have problems and go back to Windows? The correct question is, how can things be made easier for Adobe or how can Linux compensate for what Adobe does, such that users don't have problems. This might be standardizing on one package format so Adobe doesn't have to make special provisions for each Linux distro or it might be seeing what Adobe is giving the user and firing up the package manager to grab the proper version for Ubuntu instead.
A bare bones windows install doesn't even have half of those programs - she'd be pretty proper fucked on windows doing all those things if someone hadn't loaded her box with 'pirated' shit to start with.People don't have bare bones Windows installs and there is no evidence to suggest her copy of photoshop is pirated. Some people don't mind paying for quality software. The trick is making sure they have that option on Linux instead of being limited to just OSS in the repositories.
Some of the comments of the author indicate he doesn't really know what's going on either...Why should he be an expert in every software application. He just presented some common tasks, tasks his GF was likely to want to accomplish, and saw how easy they were and where they could be made easier for new users.
That is fine, except it is a longer process that just clicking a link on a Web page, which, realistically, is where a lot of us are when we want to install software. People generally don't decide they need to install a Flash plug-in or anything else. They need to solve a problem so they look for a solution, usually with their Web browser. What if she wanted to install an application to make stop action movies with her Web cam. should she go to the package manager and look or should she go to her Web browser and search for different ones and see which is most suited to her needs?
The issue is further complicated by the fact that many Web sites do provide packages for download, so unlike Windows you're asking users to learn two different methods for installing software. Further, there are usability issues with installing software from a CD, running it from your flash drive, getting packages that are for a different Linux distro or which are tar.gz files, installing software that is on a friend's computer, and installing commercial software packages.
Now I'm the first to tell others about the benefits of package managers and I think Apple and MS should get off their butts and create proper package managers for their OS's. That said, package managers do not solve all the problems and current package managers are very poor at many common use cases (anything other than find and install software that is in the repository when you already know what that software is). Ignoring these usability problems and use cases, will not benefit Linux's usability or desktop market share.
I've performed usability tests in the past and regardless of if you're a coder or a usability expert, they almost always reveal usability issues that seem unintuitive or which people want to "explain away" because they can't wrap their heads around what is happening. Usability testing is, however, a very underused scientific process in computing today.
That said, over the years a lot of general guidelines have emerged that can help designers avoid some of the really common mistakes that happen over and over and over again. From my own use of Ubuntu, it is clear there has not even been consistent use of those guidelines to get the low hanging fruit before going to the expense of a real usability test. There is much work to be done.
Personally, I think "file extensions" are a pretty evil concept, the way most operating systems today handle them.They are, indeed, a common usability problem, especially when one goes into hidden extensions and multiple extensions. Expecting users to know what three letter combination mean "program that does stuff" and which ones mean "data you can read or see or hear" is absurd, especially when they can only sometimes see them and what they see is not always the "real" extension. It is certainly an area where OS's could be significantly improved.
Except that only works for some software. Not all software is in the repositories.
try telling a true new user (not someone who has used windows for a while) to install a new program on windows. where do they begin? hint: probably not with 'internet explorer'.Actually most people I know start all unfamiliar computing tasks from the browser and usually from Google. This includes finding and installing software. Heck, as a Linux user when I need a new application I don't go to the package manager to look through it. I go to Firefox and search for applications that do what I want, preferably with reviews.
...deb packages and rpm files are easily as 'intuitive' as window's 'download from a site...Except not all Linux software is distributed as .deb and rpms, much of it is yum or tar.gz or something else, a usability issue Windows users don't have to deal with and which most Linux distros have ignored.
...they aren't yet as common, but neither are osx packages, so does that make osx packages harder to install than windows installers?No, how common it is doesn't effect how usable it is. OS X packages are harder to install for some use cases though. They're easier to drag and drop from media, but there is an extra step when installing from the Web because they are distributed within .dmg files for the most part.
I think you're missing the point of this usability study. It exposed real problems with using Ubuntu today. It doesn't matter if other OS's also have problems. The only consideration of other OS's should really be to see if they have solved that usability problem and if so, can that solution be used in Ubuntu.
Okay, I've used the Exchange Web access client and, well I was less than impressed. One of the people we did business with included a PGP public key in his signature. He's a security professional and and academic, so that's pretty acceptable and he sure wasn't going to stop doing it just for his e-mails with us. The problem, whenever anyone tried to view that message with the Exchange Webmail client, it would lock up their account entirely and they would be unable to send or receive e-mail using the Webmail or normal client until an admin would log in and delete that message from their account manually, without looking at it. We were all less than impressed, let me tell you.
That was by no means the only recurring problem we had, but it was one of the most common. And don't even get me started on Sharepoint. Both these led to strict "No MS server products" rules without a formal evaluation from IT, no more departmental IT introductions of this crap.