Pages:     | 1 |   ...   | 23 | 24 || 26 | 27 |   ...   | 51 |

In the ANLoc localise software programme we awarded small grants to teams across Africa who were tasked to translate a number of pieces of open source software. What was different in this programme is that these teams followed a strict translation programme. This programme was created by experienced localisers, allowing the teams to focus on the work rather than having to choose which software to translate. In addition, a team of technical experts guided the translators and handled all the behind the scenes technical issues that would hamper inexperienced localisers.

Dwayne Bailey In ANLoc we showed that anyone can localise. The quality of these localizations, of course, needs to be evaluated by the respective language communities. Using technical experts able to hide the software complexities, or the hard parts of localization, from the localisers, coupled with an easy to use web-based localization platform, that required no installation and was easy to learn, we showed that small teams working in minority languages could be assisted to make a dramatic impact for their language.

The future of open source and community translation is web-based, distributed translation platforms. Within ANLoc we have continued to develop Pootle, a web-based translation platform created by Translate.org.za.

This is critical piece of software for volunteer translation environments because it dramatically lowers the barrier to entry.

It also ensures a behaviour that can benefit localization of minority languages in the long run. This behaviour is the sharing of resources. For minority languages it is important that terminology lists and translation expertise are shared. For dominant languages this matters less. But in a language like Xhosa if you create an ict terms list you need to share it, so that its used but also so that people are aware of it.

What motivates companies to localise Companies localise based around market share, market advantage and policy. If a company can sell more products by localising then they will localise. If a company can gain market advantage then they will localise. Lastly, if a policy says that something must be localised then a company will comply.

While companies will lose no time in portraying this to be about their concern for communities and language, below the surface it rarely is more than the above motivators. The most intriguing motivator though is market advantage and its twin, market disadvantage. With an eleven languages policy in South Africa and no software translation policy, there is no need to translate. With the market mostly not needing localised software and buying capacity of the other languages being low there is no push to drive market share through localization. There is also no market advantage to having software localised.

When OpenOffice.org was translated in South Africa it created a market disadvantage to others participating in the field of office software. To eliminate that disadvantage Microsoft translated their software. Remove the names of the players and you have a general principle that can be Dwayne Bailey used by language activists to advance local languages. When you see an opportunity to create a change in market conditions you can often use open source to execute that opportunity.

Open source allows local languages to create the language environment that they desire. Commercial translation programs were driven by market value until open source pressurised them to change and include other motivators such as the number of speakers, local language policies, etc.

Open source localizations remove the excuse not to have a dominant commercial product localised.

Countries like Iceland can ensure that Microsoft is translated into Icelandic and Canada ensured that Windows 95 was simultaneously released in English and French through their relevant language policies. While most minority languages do not have that policy power, through open source they can garner that power by changing the rules of the market. Where no local language software exists it is easy to suggest that none should be created as there clearly is no market. Once open source translations are available it easily becomes a checklist requirement for commercial software to eliminate their market disadvantage.

If we examine the Microsoft llp program we can see that in many cases a llp language follows open source translations and initiatives. This is true in South Africa, Nepal, Nigeria, Tanzania and other countries. While llp gave Microsoft the ability to respond to this pressure it was open source translation initiatives that created the real pressure for change.

In South Africa, Afrikaans language bodies made many requests for Microsoft localization, including letter writing and meetings with Bill Gates father. While the constitution of 1994, which introduced eleven official languages, had made the need clear, it would take more than a decade before llp allowed Microsoft to respond to this need. This period included five releases of Microsofts Windows product, and to date not all eleven languages have been covered by llps.

It is interesting to note that three weeks after the release of OpenOffice.

org in three of South Africas official languages in 2004 a press release from Microsoft stated that they would be creating local language version in six months. It would take Microsoft two years to deliver on this promise, they could be accused of a vapourware announcement, but they did deliver in the end. Thanks to open source.

Dwayne Bailey The lesson from this is not that Microsoft doesnt care about language. It is that, Open Source creates a motivator that forces companies to localise software because of a market disadvantage. A language activist translating open source is much more powerful than a letter writer making public appeals to commercial companies. Open source is a powerful tool in the hands of language activists.


Its interesting to note that commercial interests are not always aligned with societal interests. Commercial interests are not concerned about languages of ten thousand speakers. While in open source it simply takes one person to make their language a personal priority to enable that language. Open source allows non-commercial motivations to be prioritised and to take root.

These non-commercial motivations can include the desire to promote and advance the language. The desire to make the language a primary language of education. They could include efforts at language recovery to bring a language back from the brink. They could include political motivations, both good and bad, to advance language. Ultimately these are driven by people who care about language. Open source releases them from the obligation of justifying their efforts in terms of economics.

While we have not dealt with open content in much detail, almost all of these principles can be applied to open content. The difference being that while open source provides the tools to create, open content provides the content that stimulates the creation of more and more content to the point where local language content is large enough to make the creation self-sustaining.

The message is clear. If you want to empower local languages in the digital world then you have to localise open source software. By doing that you create the environment, demand, incentive and skills that will lead to more localised software and ultimately a digital ecosystem in the local language.

Dwayne Bailey MLANIE DULONG DE ROSNAY TRANSLATION AND LOCALIZATION OF CREATIVE COMMONS LICENSES Creative Commons is a set of legal rules and licenses for smoothing the flow of creative work. Ease of legal adoption in local jurisdictions goes far beyond mere translation from one language to another. It also calls for building a community of experts throughout the world.

Original article in English.

MLANIE DULONG DE ROSNAY is a researcher at the ISCC, the Institute for Communication Sciences of the CNRS and Creative Commons France legal lead at CERSA CNRS University Paris2. She co-founded Communia international association on the digital public domain. After receiving a PhD in law, she was a research fellow at the Berkman Center for Internet & Society of Harvard Law School and at the Institute for Information Law of the University of Amsterdam.

ith the extension of copyright law duration and the expansion of its scope, possibilities to access and reuse works are being Wreduced, while digital technologies can (and should) be used to facilitate their usage, instead of locking them even more. Copyright law grants automatically to authors an exclusive right to control the copying, distribution and modification of their works, leaving few rights available to the public without authorisation, such as parody, private or educational use.

However, creators can choose to let others copy and reuse their works for free. By deciding to be more generous, they will get more exposure, maybe even foster citizen participation, creative remix or translation by volunteers. Creative Commons (cc), a non-profit organization, is offering a set of open content licenses to the public, in order to remove barriers to access and creativity by facilitating sharing of works 1. When distributing their work under a cc license, authors authorise the public to copy their work given that some conditions are respected, such as providing appropriate credit, reserving commercial rights or requiring modified versions to be made available under the same freedoms.

cc licenses are applicable to works which are covered by copyright law :

text, blog posts, articles, books, images, websites, audiovisual creations, photographs, music, etc. They are used by individual artists and institutions such as Wikipedia, Al Jazeera for footage, mit for educational material or Hindawi Open Access for scientific journals. Other tools are available for data or public domain works which are not covered by copyright : the White House, the Dutch and the Piedmont governments use a cc0 (0 for zero) instrument in order to indicate they waive their rights on public data to facilitate citizen access to information and innovation based on public sector data reuse.

1 Net.lang is itself distributed under a CC-by-sa license to facilitate translation and publication in every country and language.

Mlanie Dulong De Rosnay cc licenses are made available to the public though an online user interface 2 asking authors to specify which rights they wish to grant to the public and to choose optional elements. Licensers may (or may not) request their work be used for non-commercial purposes only or in a non-derivative way only, or request the derivatives (such as translations) be licensed under the same conditions. Based on the answers to these questions, the user will be delivered one of six licenses to be displayed on their website or on the physical copy of their works in order to indicate to the public which freedoms are granted in advance and which rights are still reserved.

The six licenses are the following :

License Logo License Logo Attribution (BY) Non Commercial (BY NC) Non Commercial - No Derivative Works No Derivative Works (BY ND) (BY NC ND) Share Alike (BY SA) Non Commercial - Share Alike (BY NC SA) Licenses are made available in various formats when clicking from one to the other :

A button with the cc logo, containing a link to the licenses human-readable summary.

Embedded machine-readable code containing metadata to be processed by search engines.

A human-readable summary of the licenses core freedoms and optional restrictions : http://creativecommons.org/licenses/by-sa/3. The legal code, e.g. the full license : http://creativecommons.org/ licenses/by-sa/3.0/legalcode Originally, the legal code was drafted according to us copyright law, as the organization is based in this jurisdiction. It has later been drafted in reference to international conventions. Licenses are being translated in over fifty languages and seventy countries. This process called international porting goes beyond a mere translation. For instance, the definitions are expected to be extracted from copyright legislations in each jurisdiction.

2 http://creativecommons.org/choose Mlanie Dulong De Rosnay The purpose of having local licenses is to provide a linguistic and legal translation, as well as to increase access, acceptability and understanding by users and judges who need to interpret the licenses in their jurisdiction.

The internationalisation process also provides local teams of affiliates who are expert in copyright and open content licensing. Beyond ensuring the translation and porting of the legal code, jurisdictions project leads work with local user communities and governments to explain the licenses and facilitate their adoption 3. Jurisdiction teams also collaborate with cc headquarters staff to perform research, provide suggestions to improve the licensing system, report on users questions, use cases and issues arising in their jurisdiction. They translate and create educational material and constitute a network advising on questions affecting user communities around the world.

However, the legal porting process comes with a caveat due to the lack of harmonisation among copyright legislations. As copyright law varies among countries, licenses do not exactly cover the same scope of rights.

As cc licenses are declared compatible among themselves 4, an author is expected to consent that future adaptations of her work be licensed under unidentified terms, which can be a problem in contract law. The legal porting process has been a useful constitutional event for the development of an international network, and ported versions facilitate understanding and adaptation in diverse legal cultures and systems. But it is a timeconsuming task in a complex international law environment. In any case, linguistic translations improve access, acceptability and understanding by non-native English speakers. The licenses human-readable translations, summarising the legal text in a few sentences written in plain, non-legalese language, are making it clear to all creators that works can be reused.

Pages:     | 1 |   ...   | 23 | 24 || 26 | 27 |   ...   | 51 |

2011 www.dissers.ru -

, .
, , , , 1-2 .