9 Characteristics of Free Software Users

Operating systems come with cultures as much as codebases. I was forcibly reminded of this fact over the holidays when several family members and neighbors press-ganged me into troubleshooting their Windows computers. Although none of us had any formal computer training, and I know almost nothing about Windows, I was able to solve problems that baffled the others -- not because of any technical brilliance, but because the free software culture in which I spend my days made me better able to cope.

The origins of these cultures are more or less obvious. Windows and other proprietary software are the products of a commercial software market. In this culture, information flows mainly in one direction -- from the manufacturer -- and companies' obsession with so-called intellectual property and vendor lock-in encourages them to force users into the role of unquestioning consumers.

By contrast, free software culture has two sources. The first is the Unix culture that Eric Raymond describes in The Art of Unix Programming, with its emphasis on excellence. The second is the Free Software Definition's four freedoms.

True, end-users are unlikely to be interested themselves in the freedoms to study or improve the program. But the availability of these freedoms for developers conditions everybody's expectations. Moreover, the freedoms to run and redistribute programs relieve everyone of some of the more unwelcome aspects of proprietary culture. At any rate, together these sources create a more active, more demanding set of users than is found in proprietary software.

Unsurprisingly, these differences in origins lead to entirely different sets of expectations. Exceptions do occur, of course, and, the more expertise users have, the less pronounced the differences are. Moreover, free software like FireFox and OpenOffice.org are becoming more commonplace on proprietary platforms. And, similarly proprietary culture is seeping into free software as it becomes big business.

Still, for the most part, you can expect free software users to differ from proprietary in a number of fundamental ways. Furthermore, whether you are aware of these differences can have considerable impact in your success when marketing or developing software.

1) Free software users expect open licenses and no activation methods

Proprietary vendors like Adobe and Xara who have experimented with GNU/Linux versions of the software usually conclude that free software users will not buy commercial software. However, as companies such as Mandriva and Red Hat have proved, such conclusions are more of a failure to conceive of alternative business methods than an observation of reality. If nothing else, business users will often buy commercially in order to have the comfort of a traditional relationship with a vendor.

However, given any sort of chance, free software users do reject proprietary licenses or activation methods that restrict their freedom to copy and redistribute software. Some may endure proprietary licenses if comparable functionality is unavailable elsewhere. Others may accept a proprietary license for non-essential software like games. But, at the first sign of an alternative, they will abandon a proprietary product. And many, of course, will not even accept these temporary compromises.

If you want to sell to the free software community, forget about making money on the software and see what services you can develop around the software. Or do you think it's an accident that file-sharing and free culture have roots in the free software community?

2) Free software users expect regular upgrades and patches

Free operating systems are set up for instant gratification. You want a piece of software? Switch to the root account, and in five minutes you have it installed and ready to use without rebooting.

This daily functionality results in the same high expectations for upgrades and patches. In free software, upgrades and patches are not a once yearly event complete with beta versions and release candidates. They are closer to a daily occurrence. Project maintainers take this responsibility so seriously that many have been known to take personal time from work in order to get a bug or security patch out as quickly as possible.
3) Free software users expect to work the way they choose

Switching from Windows to GNU/Linux, the first thing that users are likely to notice is how many customization options are available just for the look and the operation of the desktop. If anything, they are likely to feel that too many options are available. Often, they cannot imagine ever wanting half the options.

These options are a direct result of the sense of control that free software encourages in its users. Not only do they expect to use menus, toolbars or keyboard shortcuts as their preference dictates, but they expect to control the color, widgets and even placement of desktop features easily and efficiently. If they cross the other way, going from GNU/Linux to Windows, they are apt to feel restricted, that they are being forced to do things the way that the developers want them to do, rather than consulting only their own preferences.

4) Free software users want control of their own systems

For a free software user, one of the most irksome aspects of Windows XP or Vista is that you are constantly being nagged by pop-ups. The system itself notifies you about available upgrades, possible security risks, and the current state of your system. And it's not unusual for your manufacturer's software to have its own messages as well as Java and several other programs. Meanwhile, the operating system and one or two other basic pieces of software are phoning home, and lockdown technologies are policing your computing. Sometimes, it seems like your work is being interrupted every 30 seconds or so.

Desktops in free software operating systems are starting to have notifications, but, so far, they are for the entire system. Even more importantly, they can be turned off. Experienced GNU/Linux or FreeBSD users know that routine system events belong in log files, where they can be read at leisure.

As for lockdown or surveillance technologies, forget it. Many free software users are suspicious of comparatively benign automatic survey tools like the Debian Popularity Contest or Smolt, let alone something that takes control from their hands.

5) Free software users explore

I was able to solve two of the Windows problems I faced over the holidays in a matter of moments. One was simply a case of plugging the monitor into the dedicated video card instead of the onboard on the mother board. The other was solved by using a file manager instead of the dedicated tools that came with the new hardware. Asked why they didn't look around for solutions, those I was helping hemmed and hawed, but eventually they more or less admitted that they were afraid to try.

To me, these reactions epitomize the learned helplessness that proprietary software usually encourages. With only a limited number of tools visible from the desktop -- many buried several dialogs down -- and most of those tools giving no indication of how they achieve their results -- the average Windows user has little incentive to learn how to administer their systems.

However, on free software systems, exploration is easy. Most configuration, for instance, is done using plain text files that you can view from your file manager. And since exploration leads to quick and effective results, the users of free operating systems are encouraged to explore and soon grow to expect the ability to do so. Place them on a Windows system, and they'll probably complain that they are isolated from the system as effectively as if they were trying to type wearing mittens.

6) Free software users expect to help themselves

Free software users have no objection to help files. If anything, they love them. To the traditional Unix man pages, they have info pages at the command line, and online help on the desktop. But they are far less likely than proprietary users to expect formal technical support. Instead, what they expect are the means to help themselves -- not only help files, but easily accessible configuration files (preferable in human-readable plain text), and mail forums and IRC channels where they can consult each other. A Do-it-yourself philosophy runs deep in almost every free software user. The longer they have been using it, the deeper it runs.

7) Free software users don't fear the command line

To Windows users, the command line is a fearful place. And no wonder, considering its awkwardness and limitations; a new one was one of the features promised for Vista and dropped. But the command line is much more friendly in free operating systems than in Windows, and many users soon become comfortable with it.

In almost every case, a typed command has more options and power than its graphical equivalent in free software. Users will gladly use the graphical interface, but, when its limitation is reached, many still happily drop down into the command line. Partly, it's a geek macho thing, but a large part of the habit is sheer practicality. Unless interface designers manage to offer the same functionality as the command line, that's not going to change -- and, frankly, not many are trying to do so.

8) Free software users learn software categories, not programs

Blocked from easily learning about their operating system, consumers of proprietary software operate as if casting magic spells -- ritual recipes that, if used exactly right, will give them the desired results. Added to the fact that proprietary software can be expensive, they tend to become familiar with one office suite, and one web browser and mail reader. As a result, switching software can be traumatic to them.

By contrast, free software users come to have both the system knowledge and the software selection to experiment. They may settle on one piece of software in each category, but only after experimenting with all the possibilities. Should they need a feature that their choice lacks, they'll find a temporary or permanent replacement, trusting that other features they need will be in both programs. Far more than proprietary users, their loyalty is provisional, and dependent on quality and selection. They lack the financial investment that keeps proprietary users locked-in to a particular vendor, and see no reason to change that.

9) Free software users expect access to developers and other employees

The free software community prides itself on being a meritocracy, where status is the result of accomplishment and contribution. Since status depends on what you have done recently, it is less fixed than in a traditional office. Even where obvious leaders exist, they are more often first among equals than managers with direct control over others. That, in turn, means that community members cannot isolate themselves behind a wall of authority. Community members generally have direct access to project leaders, generally via email and IRC. Nor do most project leaders object to this arrangement.

Even in companies, traces of this flat structure exists. Instead of resisting it, sensible managers will accept it and claim a special place solely because of their position.

Conclusion

How long these characteristics of free software will continue to exist is uncertain. In the last few years, a new category of free operating system users has begun to emerge: those who remain entirely on the desktop. In the rush to become more user-friendly -- which usually means more like Windows -- the chance exists that the free software user culture will become unrecognizable to long-time users in the next few years.

However, that seems unlikely. For the most part, the purely desktop user's sensibilities are not sapping the free software culture so much as being accommodated and isolated as a special case. Unless they are content to stay in their normal routines, within a year or two, desktop users will face some problem that they cannot solve without becoming either more adventurous or more in contact with the mainstream culture. When that happens, they will have taken the first steps away from being passive consumers and towards becoming the owners of their own machines.

IE still top dog over Firefox in corporate browser kennel

With a new CEO on board and a major update of its Firefox Web browser expected this year, Mozilla Corp. hopes to reinvigorate its campaign to pull users away from Microsoft Corp.'s still-dominant Internet Explorer software.

But Mountain View, Calif.-based Mozilla continues to expend little energy on wooing IT managers to formally adopt Firefox for deployment within their organizations, according to analysts and users of the open-source browser.

In the past, Firefox faced two main obstacles that limited its adoption by corporate users: its immaturity, and its incompatibility with corporate Web applications and intranets that relied on Microsoft technologies such as ActiveX.

Now nearly three-and-a-half years old and nearing the release of Version 3, Firefox no longer can be accused of being callow. And while many IE-only apps remain, plenty of others have been overhauled to support Firefox as well, according to Rafael Ebron, general manager of Browser Garage LLC, a Web consulting firm in Mountain View.

However, other obstacles to broader adoption have emerged. Mozilla thus far has neglected to develop tools to help IT departments deploy and manage Firefox, and it doesn't offer paid technical support services to risk-averse corporate users.

"The enterprise is looking for a neck to choke, and that is absolutely what is missing from Firefox," said Ebron, a former product manager for Firefox and its predecessor, Netscape Navigator. "If you have a problem with IE and you are a big enough customer to Microsoft, [CEO] Steve Ballmer is going to come out and talk to you. That isn't there yet from Mozilla. It isn't their focus."

Mozilla claims that Firefox has more than 125 million users. And according to market researchers, the open-source browser has made some steady, albeit relatively small, inroads against IE on usage.

For example, Net Applications Inc., an Aliso Viejo, Calif.-based company that tracks visitors to about 40,000 Web sites, said Firefox held a 17% share of browser usage in December, compared with 76% for IE.

Similarly, Janco Associates Inc. in Park City, Utah, currently gives Firefox a 16% usage share among visitors to 17 business-to-business Web sites that it monitors. Janco puts IE's share at 67% while giving 9% to Netscape and 3% to Google Desktop. (Netscape is credited with only a minuscule market share by Net Applications, which doesn't include Google Desktop in its rankings.)

Firefox's market share has increased from 14% since last January, while IE's share has eroded by two percentage points, according to Janco. But Firefox's gains have mostly come from workers installing the browser on their own, without IT's blessing, noted Janco CEO Victor Janulaitis.

"Users are frustrated with Microsoft's product, and more people are starting to experiment in enterprises with Firefox," he said.

Firefox may gain more users following AOL LLC's Dec. 28 announcement that it will discontinue Netscape Navigator -- and its recommendation that users switch to Mozilla's browser.

But Mozilla's laissez-faire attitude toward corporate users can lead to awkward situations, such as the one at a leading vendor of Web-based software. The company's CIO, who asked not to be identified, said it is a longtime Microsoft shop that has standardized on IE as its browser of choice.

Even so, all of the apps that the vendor sells or uses internally can run on multiple browsers. And demand for Firefox among its employees is so heavy — "Salesforce.com runs better in Firefox," one worker told the CIO — that the internal ratio of Firefox to IE usage currently is about 60:40.

The big downside is the difficulty of managing Firefox, especially in comparison to administering IE, according to the CIO. For example, he said that the IT department can patch IE via automated central updates. On the other hand, "we have to send an e-mail and have users manually download Firefox updates, which is not ideal," he said.

That won't change in Firefox 3. A Beta 2 release available now lacks features that IT managers typically want, such as the ability to automatically deploy multiple copies of Firefox through Windows Installer package files — better known as .MSI files, for the file extension they use.

The update also can't be patched from a central console, like IE can be through Microsoft's Windows Server Update Services. Nor can it be managed and secured via Active Directory, Microsoft's tool for setting group policies.

As a result, many IT staffs looking for help in rolling out or managing Firefox have resorted to using free third-party tools.

One, FirefoxADM, lets administrators centrally manage locked or default settings in Firefox via group policy settings in Active Directory. Another, called FrontMotion, offers basic .MSI installation packages for Firefox for free, and custom ones for a small fee.

FirefoxADM has been downloaded about 22,000 times, according to Mark Sammons, a senior computing officer at the University of Edinburgh, who created the tool and is using it to manage Firefox on 8,000 PCs at the Scottish school.

FrontMotion's installers has been downloaded nearly 131,000 times, said Eric Kuo, its developer. By day, Kuo works as an IT director at a medical company in Lubbock, Texas, that he asked not be identified.

Both FirefoxADM and FrontMotion are open-source products. But each is largely developed by a moonlighting individual. That has hurt FirefoxADM, in particular; Sammons acknowledges that he hasn't added any new features to the management tool in the past two years.

"I have no illusions as to what FirefoxADM is," he wrote in an e-mail. "I think it works well, but ultimately, it's a work-around for functionality that really needs to be built into Firefox itself."

An even bigger problem is that neither of the two tools has been formally tested and certified by Mozilla.

"It's absolute FUD to say that you can't administer Firefox well within an Active Directory environment with third-party tools," Ebron said, using the acronym for fear, uncertainty and doubt.

Nonetheless, both he and Kuo said that having Mozilla's official seal of approval would be a big plus to corporate users planning major deployments.

"It comes down to a perception of who owns the tools, not the tools themselves," Kuo said. He noted that preliminary talks with Mozilla about selling FrontMotion — a move he would welcome — went nowhere.

Mozilla has no plans to add tighter integration between Firefox and Active Directory, according to Chris Hofmann, the open-source vendor's director of special projects. He dismissed Active Directory as a "proprietary technology" that would hurt rather than help Firefox administrators.

"Multiple levels of permissions applied across different groups add a lot of complexity," he said. "If you look at the track record for that feature, it’s resulted in less security for IE."

But Hofmann, who oversees security and language localization development for Firefox in addition to helping run its enterprise-oriented efforts, acknowledged that users are clamoring for Mozilla to provide more enterprise tools. For example, on a public wiki maintained by Mozilla's Firefox Enterprise Working Group, .MSI installers and better tools for preferences management top a new-features wish list.

Automated installers are relatively easy to create, and such a tool "might come sooner" from Mozilla than one for Active Directory would, Hofmann said. But he added that Mozilla has no plans to acquire or certify third-party tools or to set up a paid support business for corporate users.

Despite the success that open-source vendors such as Red Hat Inc. and MySQL AB have had in getting corporate users to sign paid support contracts, Hofmann contended that companies are starting to favor DIY support approaches on open-source technologies. "There's a growing number of CIOs asking, 'What is the value of a support contract? What are we getting out of it?'" he said.

Mozilla's stance doesn't surprise Kuo, who claimed that the organization is dominated by developers who would be unlikely to find the idea of starting an IT support business sexy.

Kuo added that he doesn't think Mozilla will suddenly change its attitude and develop a browser deployment tool that could render FrontMotion obsolete. Mozilla "could create it themselves," he said. "But it's obviously not their priority."

Michael Kaply, a senior software engineer at IBM who describes himself as a "Firefox advocate," said that the open-source browser is currently being used by about 72,000 of the IT vendor's 360,000 employees.

IBM wrote its own Firefox deployment software, but it doesn't use any group policies to lock down or otherwise control the browser. "Our employees have full control of their machines," Kaply wrote in an e-mail.

The company still runs some Web applications that work only with IE. But it is building in Firefox support, Kaply said. For example, a travel reservation app was recently switched to a cross-browser design. "That was a big hurdle," he wrote.

However, in a posting on his personal blog last September, Kaply lamented that the number of participants on Firefox Enterprise Working Group conference calls had "dwindled." And in an earlier posting, he said that he thought most of the large companies that had adopted Firefox were using it "as a secondary browser" only.

Gregg Keizer contributed to this story.

National Boards to Vote No for OOXML (Office Open Xml) at ISO

"There will be a BRM (Ballot resolution meeting) in Geneva on 25-29 February 2008 to answer the comments by National Standard Boards on OOXML and to resolve issues.

"Microsoft has set up a situation where many of the shortcomings and thousands of comments which need to be resolved at the BRN will not be addressed. They have closed this process as much as possible and have delayed National Boards from having adequate time to review proposed changes..."

Open Standards represent a democratic ideal, which means accountability. When one proposes one’s own property to become an ISO standard, we have a right to know all the answers before we vote you in.


As it currently stands, for the ISO community to adopt OOXML as a standard would be the first step toward our cherished Open Internet and Open Standards becoming an asset on the balance sheet of just one company, Microsoft. Recall that Microsoft was held liable by the US government and the EU as a proven monopolist, which illegally leveraged that monopoly to stifle competition. Here are some of the unresolved questions regarding OOXML that Microsoft cannot or simply will not answer:

* What will be the default file format of Office 14 (Office 2009)? Will it be identical to that described in Ecma-376? We don’t know, because Microsoft is not providing us with a clear roadmap. Further more, there appears to be a very real question as to whether or not MS’s covenant not to sue applies only to the current implementation, MS Office 2007, if you read the fine print. Where is Microsoft taking the world with OOXML? We just don’t know.
* As part of the comment disposition process in JTC1, Microsoft is proposing to make thousands of changes to the Ecma-376 specification. When will we see a version of MS Office that implements all of the revised specification? Currently, not one single version of Microsoft Office is capable of actually implementing the OOXML specification.
* Thousands of changes are being made to OOXML as part of the JTC1 comment resolution process. Is MS going to implement those changes, and if so, and how?
* We have the OOXML “standard” and we have the MS-OOXML “reality”, what Microsoft Office actually writes out. It is known, for example, that MS-OOXML can contain scripts, macros and DRM, features that are not documented at all in the OOXML specification. What other features are output by MS Office but not described in the OOXML specification?
* The Microsoft Open Specification Promise (http://www.microsoft.com/interop/osp/default.mspx) says that it only covers things that are “described in detail and not merely referenced in such Specification.” What are the technologies that are not covered by the OSP, the technologies that Microsoft believes are not described in detail or are merely referenced? What is shown in the OSP simply reinforces that the whole OOXML spec cannot be implemented. Rather than this vague language, why doesn’t Microsoft give an explicit list?
* What comments from the general public did Ecma TC45 receive during its review period in Ecma? Why have these public comments been suppressed?
* If Microsoft Office “Open” XML is truly “open,” why is it that the OOXML specification is tied so heavily to only one vendor’s products, as opposed to 40 applications currently capable of supporting the ODF specification?
* OOXML still has many technical shortcomings, as failed tests demonstrate. A couple of examples:
1) Excel 2007 can produce a fully binary file format that has the same extension as OOXML for spreadsheets, so the application knows what is inside, but the user will never be privy to that information, absent specific accommodation from Microsoft on a case-by-case basis, which is an accommodation that Microsoft is unlikely to grant to the public at large; 2) try setting a password on an OOXML spreadsheet file from Office 2007. The document is no longer in OOXML format and the user is given no indication that they are no longer in a documented file format. Is OOXML ready to be an International Standard?
* From the overall document contents, it is acutely clear that no effort has been made in OOXML to start from the existing ISO standard for the representation of documents in XML. (The current standard is ODF 1.0, formally known as ISO/IEC 26300:2006). OOXML is unneeded and only harmful. Why did MS embark on this deliberate departure from an internationally recognized standard?
* If you read through the draft submitted to ECMA and ISO you will see MS has only disclosed a disabled subset of the markup and functionality of its new file formats. Lots of elements designed into OOXML are left undefined in the specification and require behaviors upon document files that only Microsoft Office applications can provide. Why? Also, when MS released the feature RTF format for interoperability initially in OOXML, it supported both read and write capabilities in Office. But they changed that so Office 2007 can receive OOXML files but will only “Write” now to a different file format. Why did MS make this change?
* In addition to the “open specification promise,” Microsoft should change the licensing scheme for any relevant current patents which can be used to extend OOXML. Those patents should be made available under a license which is clearly compatible with GPL and LGPL. This licensing schema should apply to future relevant Microsoft patents, too. Will Microsoft make its relevant patents legally valid for Open Source use.
* The name “Office Open XML” is often mistakenly called “Open Office XML” implying a non-existent connection to the OpenOffice.org project. This naming confusion has been documented and has occurred numerous times, including by analysts and even in Microsoft press releases and blogs. Since “OpenOffice.org ” is the pre-existing name, by 6 years, Ecma should choose a new name, less apt to continue this confusion. Will Microsoft make this change as a condition of gaining ISO status for OOXML? Don’t hold your breath.
* The Open Specification Promise that covers OOXML explicitly covers only the “Ecma 376″ version of the standard. However, thousands of changes are being made to OOXML as part of the JTC1 comment resolution process. Are these changes covered as well?
* While Microsoft originally made assurances that the ISO would take control of the standard if it were approved, Microsoft has now reversed that position and will keep near-full control over OOXML within ECMA [PDF], an industry group that exists to advocate its members interests. Since the development and standardization of OOXML has been opaque, what guarantee will MS make to assure those activities will be done in an open setting in the future, including changes to MS Office as they are made which have direct effect on the format? Will Microsoft fulfill their promise to transfer stewardship, control and ownership of OOXML over to ISO, a promise made they made publicly and repeatedly?
* There’s a limit to what you can do with metadata. Working with RDF is nice, but custom XML schemas are the complete opposite of interoperability. Custom schemas break interoperability as they are by definition not shared by everybody or every organization. It thus condemns documents containing custom XML schemas to be manipulated and shared only by the users who have access to those custom XML schemas (typically inside one organization, generating maintenance problems for the future). How can the ISO community ever be assured that OOXML will interoperate with the established ISO standard?
* Many countries, including Thailand, expressed that they had no time to review the MS OOXML Spec during the Fast Track process. Other countries combined submitted thousands and thousands of comments. Yet only a small percentage of those comments will be addressed at the BRM, and all of those comments to be addressed will be technical comments only. When will comments on IPR, ISO policies/goals and JTC1 “contradiction” be resolved if not at the BRM? These countries deserve a resolution. Will they be resolved in an open and transparent fashion?
* MS claims that MS Office can support arbitrary user-supplied XML schemas. If that is really true, then the established ISO standard ODF’s schema could be loaded into Office 2007 and future versions natively, with an ODF option as default and the cloaking of OOXML as a standard dropped? Why not?

Information and communication technology (ICT) devices are able to exchange information only if they adhere to common communication protocols, technical interfaces, and information formats. We all desire the freedom and ability to develop, and implement these ICT’s. We all feel the need for ICT within the same field to be able to interchange data efficiently.

OOXML demonstrates a “standard” Microsoft hurried through ECMA and appears to be nothing more than a rubber-stamp of one vendor’s product specification. Microsoft’s OOXML is nothing more than a format Microsoft created by running a program that spits out the guts of MS Office in an XML form. Now MS appears to have realized this and expressed the “Intent” to deprecate features in the future. Microsoft appears to think it is best to get rid of references to legacy formats and other proprietary technology in experimental Version 1.

The trend is that Microsoft is opening up the boring legacy bits of OOXML, in stupefying detail, while neglecting to document the pieces actually needed for interoperability at a competitive level, like macros, scripting, encryption, etc. In essence, Microsoft is opening up and releasing the file format information that competitors like OpenOffice.org have already figured out on their own, while still at the same time restricting access to the information needed to compete. And the more MS realizes it has to open up the specification, deprecate and modernize OOXML, what do you get? You get XML. XML is XML. Strip out the non-XML garbage from OOXML and you will have the OpenDocument Format.

We need for MICROSOFT TO ANSWER THESE QUESTIONS. Rather than hiding all the information we need and trying to cloak OOXML as ODF, we ask Microsoft to please get off the sinking ship, collaborate with the global community (which will welcome Microsoft) and help develop one universal file format for all. Long term, Microsoft can only benefit from cooperating with the market!

To our readers outside Microsoft’s walls, we ask that you please consider contacting the National Board in your country, and request complete resolution for all issues raised by the comments to Microsoft’s OOXML ISO application. Please be sure to insist that both the technical and non-technical issues be completely resolved as a pre-condition before OOXML is granted ISO status.

HELP US PROTECT YOUR FREEDOM