Response on 'What is Public Software' IT for Change The key issues raised in the mails are: the term Public Software is a distortion of FOSS and takes away the key principle of freedom implied in FOSS; and that using these terms in a somewhat overlapping manners causes confusion and introduces a new agenda which is harmful to the FOSS movement and its goals. Narendra called it a 'altogether new agenda'. Yes, it is indeed a relatively new agenda, though proceeding from the same old agenda (while we are it, maybe we need to revisit and have a discussion on what our basic agendas are). Therefore whether it is a 'altogether' new agenda may need to be debated. Further, though we are convinced that the two agendas are mutually reinforcing – it may still be useful to discuss here whether the public software agenda is prejudicial to the FOSS agenda. The underlying rationale of the two concepts are different and need to be understood so that we have clarity on both. That would be a good basis of a continued discussion on this important subject. What is public software The point of departure for articulating, and for understanding, the concept of public software is the concept of 'public goods' or commonly shared goods, as against private and commercially traded goods in a society. What are the implications of 'public goods' thinking and requirements vis-a-vis the digital society? While this question merits close attention, unfortunately it has not received that attention, for a variety of reasons. We can construct a response to this question in two parts. 1.How digital possibilities can be best applied for production/ provisioning of existing (pre-digital) public goods? 2.What new public goods, in the form of entitlements to digital possibilities, now become relevant in the digital age? The first aspect of the emerging 'public goods - digital society' dynamic is about what kind of digital resources should be used by actors involved with providing traditional or pre-digital era public goods – basic health, education, livelihood support, security etc; and in what manner, in order to maximise the basic/ original objective of providing these pre-digital public goods. The second aspect is about universal provision of such digital goods and services which can be seen as the 'new public goods' of the digital age. Participating in the digital society requires that basic applications such as operating systems, editors, web browsers, screen readers be seen as 'public goods ' from which no one 'should' be excluded, and thus whose universal availability is a societal responsibility. (Apologies for a brief digression here. It is important to understand in what implications the term 'public goods' is being used here. The term is originally from economics, whereby it means such goods that are non-excludable and non-rivalrous. Non-excludable means, from which no one can be excluded. And non-rivalrous means that consumption of the good by one does not reduce its supply for the other. However, in its larger social meaning the term 'public goods' is used to mean goods from which no one 'should' be excluded, whether they by their nature are non-excludable and non-rivalrous or not. On the positive side, digital goods are inherently non-rivalrous. However, on the negative side, the digital phenomenon enables new means of exclusion which may not exist earlier – for instance broadcast versus DTH TV. These characteristics complicates the digital 'public goods' discussion, but more on that some other time :-) It is in this overall 'public goods' ecology that the concept of public software takes birth and is situated. Now since, a lot of objections on the list to the concept of 'public software' has been of the logical variety, challenging the very validity of the concept, I hope the above, and the following, discussions answers those objections. If not, I am willing and eager to discuss it further. What is FOSS The logic of FOSS arises differently. It came from the idea that locking down knowledge is essentially wrong in curtailing both freedom and opportunities of people. The knowledge embedded in software therefore should be freely accessed by all, and also be able to be used freely to develop more knowledge/ software. To this idea of freedom, the genius of Stallman added a brilliant new dimension. It is only freedom if it multiplies freedoms of others rather than curtail it (which is in fact adding a 'positive' element to the otherwise 'negative' – as in negative rights – conception of freedom). He very cleverly used the legal framework around proprietisation of knowledge (to which the basic idea of freely shareable knowledge is in fact prima facie antithetical) to posit an enforceable legal condition – anyone will be able to freely use free software knowledge only if any further knowledge produced by using this knowledge is also available freely. In fact this legal condition can be said to curtail the 'freedom' of the person creating some new knowledge using the old free software knowledge (the freedom to to keep this new knowledge created by him as private). But well, that is it, take it or leave it. This provision was expressly made for furthering the cause of common digital knowledge, a public good. Convergence and divergence Here, one can clearly begin to see the convergence between two concepts of FOSS and of software as and for public good (public software). One may even be tempted at this point to jump to the conclusion that public software is FOSS and vice versa. Well, it is 'almost' always so. But since the two concepts have clearly different constituent logics – even if practical convergence - it is 'logically' possible that there may be cases where FOSS is not public software and vice versa. A couple of examples used in the quoted write up on public software were made just to present this 'logical' though rare, if ever, possibility. (The exception was cited as an attempt to prove the rule.) Though an example like the one used, and much criticised on the list, of use of software for some extremely secretive purposes but serving public interest, would always be open to contestation. But as said, the point was only to provide a possible example showing the logical distinction, which comes from the very different logical construction of the two concepts. It is however possible that some other examples may be better than the one used in the quoted text. It is certainly at least 'logically' possible that there could be (some extremely rare) times when a software needed to be used in 'public good space' may best not be FOSS. This can and will be contested, but most people traditionally in the public goods space (not only governments but also outside it), whom we have spoken with, understand it in this way. Lets accept that. On the other hand, a software can have its source code open, but its design may be directed towards ways of stealing personal information or for triggering mines (banned under a global treaty) as a person approaches them, which though obviously FOSS (because FOSS is about open publication of the source code and the underlying licensing condition and strictly nothing else) can not be called a public good software, or public software. Why we need the term 'public software' If FOSS and Public Software are mostly the same, then the question would come, why should we have two different terms then. The reason is somewhat obvious. There is a big sector in society long devoted to the 'public goods space' which understands the idea and concept of public good much better than that of FOSS, which, I may be excused for saying, is often thought by them as a technical obscurity that will never be of much interest to them. (I agree, this may not be completely true, but that is how they feel). Now this 'public goods sector1' is a serious business, a big and necessary part of our social arrangements. They need to understand, and internalise in their work, the role of software in the digital society. And they will best understand it, and do what is necessary to do thereafter, if it is presented in the 'public goods' framework which they not only understand but take it to be their serious responsibility to work on. But it will be wrong to tell these 'public good' actors that FOSS was always meant to be the 'public goods software' or 'public software' and that the two are exactly the same, because that would be unfair to both this group and the FOSS groups. As mentioned earlier, there is a clear logical distinction between the two concepts even if a very large practical overlap. If those involved traditionally with public goods space or sector in the society find it useful to use the concept of 'public software', why should they not be able to do so? Public software is defined in terms of its public good nature, inherent in the outcomes arising from its use. 'Logically' it has nothing to do with publishing the source code or the nature of copyright licence involved, though it is quite clear that publishing the software and using a GPL licensing will almost always serve the best interests of the public. On the other hand, FOSS is 'logically' only about publishing the code and copyrighting under GPL licence and it has nothing to do with the purpose for which the software may be used – which in fact could be quite destructive, and whereby the software cannot be called public software. In fact, FOSS being GPL licensing condition based concept will exclude software released in the public domain. However the concept of 'public software' could include such software it is best qualifies the conditions of 'public good' in the given circumstance. This above was about the logical basis of the term 'public software' and the distinctions as well overlaps involved vis a vis the concept of FOSS. Now we can move to practical matters. Even if logically defensible, an obvious question is, why should or did we expend so much energy in developing and promoting the concept of public software. Public software – the practical imperative It has mostly to do with having encountered great difficulties in promoting FOSS among public sector agencies (which agenda we found very important per se, as well as to promote the overall cause of FOSS in society), and less than satisfactory progress in promoting it with academic institutions, NGOs and community based bodies. Through these experiences we realised that these agencies responded so much better if engaged through notions of publicness and welfarism vis-a-vis different software models. For instance government officials engage so much better if we start with the objectives of the work of the government and of her particular department, and then extend the characteristics of the public goods work she is involved with to the kind of software that should be used by her/ governments. Similarly, in discussions with government school teachers, we find that it is intuitive for them to grasp the idea of software as a basic learning resource that should be free, and a universal entitlement. They are also immediately attracted to the idea that the learning software be produced and supported by public interest groups/ bodies rather than commercial ones, whereby instinctively there is greater trust. It is then easier for the teachers to relate to the fact that since the interests and motivations of the public interest/ goods actors (or public actors) are only to help them, the software has all the qualities that makes their and students work easier and education more fruitful. They then relate to the features of the such software as its openness to modification, sharing etc as the way they see normal public education processes. Using the term 'public software' (accessible to all, involving participation of all) seems to them quite aligned with the underlying philosophy of the public school system (accessible to all, involving participation of all). At this point, they can of course be explained the production and licensing model underlying the software they are using, and why it is called FOSS. Frankly, starting with the license model of the software they are going to be introduced to, makes little sense to them. Principles of universal access, full inter-operability, not getting exclusively dependent on a private vendor for any government (or public education) process, collaborative building of governance processes (including digital ones, and software is nothing but structuration of such social/ governance processes), principle of transparency, of community monitoring, right of information, full and perpetual public ownership etc are clearly understood by public sector actors. It is easy to argue with them that same principles should apply to software used by and in the public sector. We could also easily agree mutually to call such variety of software as 'public software' as opposed to commercial software used for commercial sectors of the society with completely different contexts and objectives. By emphasising that the starting point for public software is the role of the public sector, (including the government) for the purposes of achieving larger societal goals of equity and social justice, we could even get down to write principles for public software, which public officials clearly could own (rather than FOSS principles which looks to them coming from areas largely alien to them). We could speak together of coming out with a public software policy, which would simply list what would be the characteristics of software that governments should produce/procure and use (in terms of public service principles listed above). Within this larger advocacy it was much easier to argue that FOSS is the right kind of software for governments to use, and that this fact should specifically be mentioned in the public software policies. In these discussion we, the government officials and us - seemed to be going forward together, collaboratively, in a manner that the agenda and discussions were co-owned. This unfortunately mostly does not happen when we take the FOSS agenda – direct and simple – to government officials, since, many tend to treat software per-se as a 'technology issue' which is best dealt with by technology experts or IT associations - see for instance the role that NASSCOMM, a industry body with vested interests, plays in many e-governance processes, including at the policy level2. They tend to treat FOSS as just one kind of software model which can be considered beside other proprietary, models. They start talking about 'overall' cost implications and performance factors as the 'obvious' key factors for taking the software procurement decisions. The ideology involved, which motivates the FOSS advocate, is largely lost on non-techie public sector actors. Public Public Partnerships Beyond governments, there are many social actors who involve themselves in production/ provisioning of public goods. FOSS groups are one such set of social actors. All kinds of voluntary, community groups are examples of such social actors. However, we will have to accept that the state or governments are a very big part of this ecology of social actors producing public goods. Unlike other actors involved in this process, governments, especially democratic/ welfarist ones, uniquely also have the 'responsibility' for producing these goods (and for this reason, the other groups in the public goods ecology are often called 'voluntary' groups3). It is a part of what has been called the 'deepening democracy' project to work towards larger partnerships and programs in the public policy as well as public goods space, involving non-state actors working with government's in a mutually supportive and complementary manner, whereby these relationships are characterised by trusts and mutual respect, though the elements of dissent and even antagonism on many socio-political issues need not be completely forgone. It is even more important in the digital space, with its unique collaborative and distributed system management capacities, that we seek to build partnerships among all the public interest or public goods actors. It is perhaps impossible to sustain even FOSS ecologies beyond a point without some kind of institutional public support – whether of a big NGO or a government agency. On the other hand, governments on their own are not upto the task of making and maintaining the best public software needed to maximise public interest opportunities in our society – especially of ensuring that the egalitarian and social justice potential of digital technologies is in fact realised. The 'news' of Oracle stopping support for ORCA development illustrates this case. There were mails on the FOSSCOMM list that we should write to Oracle to continue its support to ORCA development4, but what is Oracle's accountability to us. (That is the basic difference between a private/commercial actor and a public actor.) Since for the visually challenged, a screen reader is basic to their participation in the digital society, it needs to be an entitlement and not contingent on corporate social responsibility or voluntary effort. By definition for anything to be an entitlement or a right, there needs to be a corresponding societal commitment or obligation to ensure that right is fulfilled. And this requires the positive/committed action of the government/public sector as the primary societal agent for development and even democracy. This notion of public software thus puts the onus on governments to ensure universal availability of such basic applications, through funding, distribution, promotion etc. However, use of this concept of 'public software' also draws all other social actors motivated towards public interest to collaborate as well as they can for universal provision of such public goods. What Brazil's Public Software Centre has been doing for past few years, is along these lines - creating collaboration between government entities and FOSS enterprises/ communities to develop public software to promote governance goals. We need to build similar collaborations in India and that is one of the principal goals of our work. It is important to note that our work on public software arose from our own experiences in advocating FOSS in governments and schools and took shape independent of the Brazil project. (We learnt about the Brazil project much later, after the Kochi workshop, where the public software site was launched). Neither for the Brazilian agency involved, nor for us, the idea and concept has taken complete shape and is largely work in progress. We invite feedback and comments in this process. However we are convinced that there are very useful possibilities in using this concept for both the public sector and for the FOSS community.