? · english · espaņol  



Senator Conde's answer to CESSI






Source: References: Translation courtesy of Matt Dorn




La Plata, Buenos Aires, Argentina, 28 November 2002



Chamber of Software Companies and Information Technology Services (CESSI)


Ladies and Gentlemen,


First of all, I thank you for your letter of October 4th, 2002, in which you establish CESSI's official position with regard to Bill E-135/02-03, which refers to the use of free software for the public sector in the Province of Buenos Aires. This letter, in which I was pleased to find explicit recognition of the benefits of free software, constitutes a valuable contribution to the discussion of this Bill. In the same spirit, I am taking the opportunity to respond to you to clarify some misunderstandings, encourage the continuation of the respectful exchange of ideas about the use of technology in the public sector, and analyze some aspects of this initiative that your message didn't consider.

To facilitate this exchange of ideas, I will take it as a given that the references made in your letter to the "National State" are the result of an involuntary error. The scope of law E-135/02-03, when it is approved, will be limited to the public sector of the Province of Buenos Aires.

Allow me to begin by easing your discomfort regarding the two specific items that the aforementioned missive emphasizes as the origin of your opposition to the Bill. Both fears are unjustified, given that the project:

  1. does not propose to eliminate anyone from the list of providers to the state. On the contrary, the language has been carefully prepared in such a way that nobody be excluded from the possibility of offering services to the state, establishing at the same time clear and egalitarian conditions for all those interested in doing so.

  2. does not create any regime of exception, but on the contrary sets the foundation for future information technology bids to be freed from the current situation, in which the contracted software and hardware are generally specified by brand name and model, or, lacking that, demanding features that only one provider is in a position to provide. This distortion can in fact be considered a true regime of exception with respect to the rest of the State bids, and its elimination is one of the secondary benefits of the Bill.

Having clarified both points, I must mention that your letter, which focuses mainly on commercial and technological aspects, seems to have been composed without keeping in mind the fundaments, the spirit, and the text of the Bill. Your objections would be valid if the project demanded the use of free software based on cost estimates or technological functionality--characteristics often difficult to quantify and, moreover, subject to changes over time. However, the fundaments of the Bill mention those aspects merely as ancillary benefits, recognizing that they are not present in all cases, and arguing that, even in those cases where the use of free software implies greater cost and/or more restricted functionality, the State must avoid using proprietary software because doing so endangers its ability to fulfill adequately its mission.

To better understand this, we must recall that the mission of data processing on the part of the State is not to produce pretty documents, nor to create animated Web sites (tasks which, of course, are as easily performed with free software as with proprietary), but rather to be the guardian of the public record: to maintain updated and accurate information about the identity and patrimony of its citizens, of its own interactions with these citizens, of its own acts, etc. In order to adequately fulfill this mission through the use of information and communication technologies, the State must preserve three essential aspects:

  1. Security. In particular, the State must go to great lengths to ensure that only authorized persons have access to data, and at the same time guarantee that access on the part of those persons cannot be denied by third parties.

  2. Persistence. The state administers data whose useful life is often measured in hundreds of years, so there must exist a guarantee that the data will be available and accessible for very long periods of time.

  3. Transparency. The state has the obligation to publish all of its acts, except those whose revelation might negatively affect national security or the well-being of its citizens.

If the State, in its eagerness to use cheaper or more modern technology, were to neglect one of these principles, it would be failing in the fulfillment of its mission.

The incompatibility between the proprietary licensing model and the principles of security, persistence and transparency, comes from express prohibitions or insurmountable practical restrictions that proprietary licensing presents for carrying out the tasks that would allow for ensuring the fulfillment of these principles. Some of the required tasks are:

  1. Inspection of the program's function. Without this ability, which, in the vast majority of cases, is made difficult when not expressly prohibited by the proprietary licensing model, it is impossible to guarantee any of the three aspects mentioned above.

    1. Security: There is a debate in academic circles about whether free software is inherently more secure than proprietary software, but it is not expected that this question will be resolved in the short term. There is a consensus, however, about the fact that it is far more simple to hide malicious code (back doors, time bombs, etc.) in software whose source code is not publicly available than in software that can be inspected by any interested person, and there is abundant evidence of hidden functionality in proprietary programs, even in those produced by the most prestigious companies.

    2. Persistence: Without the possibility of analyzing the source code, it is also impossible to determine if the the author of the system included temporary or remote disabling mechanisms that can compromise the possibility of access to data in the future. Similarly, when using proprietary software, the user stores his or her data through the use of essentially unknown software, in an essentially unknown format. Without the possibility of inspection, it is impossible to know if the format uses technologies subject to patent or copyright restrictions that could result in future access to that information being dependent on the possession of corresponding licenses.

    3. Transparency: When the State uses software as operations support for its processes, the software becomes an inextricable part of these processes, and therefore is subject to the requirement that all government acts be made public. The State cannot, in these cases, use software whose source code is not publicly available, without violating basic constitutional principles defined in the National Constitution and in the Constitution of the Province of Buenos Aires.

  2. Improvement of the program's functionality. With a program licensed under the proprietary model, only its original author has the ability to correct errors or add or remove functionality. Beyond the obvious and unacceptable dependency on a single provider that this restriction implies, it implies that the State remains without an acceptable course of action in many situations. Some examples:

    1. Absence of, or delay in the correction of problems: The priorities of the software provider are not necessarily the same as those of their customers. If an identified security problem in a program is not on the list of priorities of the provider, or if it refuses to correct it or demands excessive compensation to do so (for example, demanding payment for an upgrade), the State doesn't even have recourse to its own means to obtain a correction by a third party.

    2. Incompatibility with previous versions: The cases of "improved" program versions that have problems reading the data of previous versions are well known. The very short cycle of software obsolescence, motivated much more by marketing reasons than by real demand for new functionality, obliges users to keep their software updated, and the price of the update includes (often without the users' knowledge) surrendering access to valuable data stored in files.

    3. Incongruence between software and the law: The State uses software to implement the mandate of law. The problem is that when there is a conflict (a result, perhaps, of a different interpretation) between the letter of the law and the functionality of the program, the citizen finds that the software is more powerful than the law, since it manages the process, and the solution of the problem depends on whether the provider of the program is prepared to correct it, do it in a reasonable way and within a reasonable time and budget.

    4. Disappearance of the provider or the product: Examples abound of proprietary programs whose users find themselves forced to undertake very costly migrations due to the provider's going out of business, to its acquisition by another larger company, or to the simple discontinuation of the product by the author's unilateral decision.

  3. Preservation of Technological Neutrality. When software is licensed under the proprietary model, the user renounces the the ability to make certain decisions, since they are dictated by the author of the program. These decisions range from the hardware platform (since the author decides on which platforms to offer the program) to the programs to use for related tasks (since the author makes sure that his or her products function better, if not solely, when they interact with others of his or her own making). This implies difficulties for guaranteeing the persistence of the data, since the obligation to accompany the provider in its decisions can bring unsustainable situations (for example, an unstoppable run of hardware upgrades). In the same way, the selection of a given proprietary software product on the part of the State limits the citizen's freedom of choice regarding the products that he or she can use to interact electronically with the public administration, which constitutes a violation of the equality of citizens before the law.

  4. Offering services independent of licensing details. Offering services of the State is not optional, nor does it admit delays or obstacles. The different models of proprietary licensing are a constant source of confusion about the legality of using a certain program for a given purpose, on a given computer, by a given group of users, a situation that is significantly more complicated when we speak of using a combination of several programs. This constitutes a serious risk to the continuity of service offerings, since one misunderstanding about licensing terms, a change in them, their expiration, or a raise in price could oblige the state to suspend the offer of a service for lack of necessary licenses.

I would like to point out three aspects of this analysis, however incomplete it is:

  1. None of the cases cited have merely academic interest, but rather all of them are concrete examples of problems that present themselves in practice each time the public administration uses software licensed under the proprietary model, and before which the State, deprived of a legitimate alternative course of action, ends up acting in a way unfavorable to its own interests, and those of its citizens.

  2. Given that the absence of these capabilities don't have their origin in the technical characteristics of the programs in question, but that they emerge directly from the proprietary licensing model, it can be immediately concluded that the licensing model is much more important than price or technical characteristics, for which reason your suggestion of leaving it aside as a primary criteria of evaluation would be irresponsible on the part of the public administration.

  3. The only way the State has of using software to conscientiously carry out its role as guardian of the public record is through the application of programs whose license does not affect the necessary capabilities, that is, that it permits, without limits on time nor on number or type of computers, the execution, study, correction, improvement, expansion, and adaption of the program in accordance with the necessities of the State, not those of the provider. It will not have escaped your attention that these are precisely the abilities that free licenses grant.

In fact, even if I really am convinced that respecting the rights of the citizen is a principle superior to commercial considerations, these have also been carefully considered at the time of composing the Bill. It is clear upon reading it that nobody is obliged to deliver products and services to the State free of charge. The amounts of the contracts are deliberately left out of the framework of the Bill, leaving this aspect free to the action of the market. I also want to take exception to your letter's statement that the State has the same freedom of action in economic and technological terms as any other user. Much to the contrary, the contracts of the State must respect much more restrictive norms than those with which a private entity must comply. That owes itself, again, to the fact that the mission of the State is very different from the rest of the entities that make up society, and therefore all of its actions must reflect that mission and those differences.

Additionally, as you point out very clearly, the Provincial State has the ability and the right to choose free software solutions in however many instances it considers convenient. This ability, and this right, are not fundamentally different from those of any person, or any company, to set policy regarding information technology. However, as a function of the essential principles that I have summarized above, these are all instances in which it is possible to employ free software. And it is appropriate that the legislative branch, while representative of the combination of the wills of the electorate, fix the fundamental public policy lines through laws.

The bright outlook that your letter forecasts for software makes me enormously happy, of course: a projected growth of 60% in jobs at an economic moment which expects, very optimistically, a growth of 6% is certainly a very encouraging objective. Even more, I am convinced that Bill E-135/02-03 in itself, far from a threat, is a great opportunity to feed this growth: an analysis of the list of the 149 members that make up your organization shows that many of them are dedicated, in a prominent or exclusive way, to providing services. As a consequence, a Law like the one proposed would end up particularly beneficial for your members, since it opens to them the possibility of competing freely in providing services associated with free software, with the only requirement being their qualification.

The principle evident in your letter is that "the State must choose its providers based on the merits of their products [...] and not discriminate on the basis of their model of licensing of products." Perhaps it is necessary to repeat here that the Bill does not discriminate between providers. The fact of commercializing programs through a proprietary model does not disqualify any company as a potential provider of the State, as long as the terms of their contracts with the State abide by the foreseen conditions of the Bill.

On the other hand, we share broadly the conviction that the State must choose the software that employs based on its merit. But software merit is, without a doubt, a perception relative to the user. Thus, for example, for a user with vision difficulties or impaired motor ability, software with a vocal interface will be practically indispensable; and at the same time this software will lack merit (and probably be useless) for a user with hearing difficulties. Accordingly, the most meritorious software for the State is that which permits it to make effective the guarantees of security, persistence and transparency previously mentioned, thus making it possible for public institutions to fulfill their mission. However, as I indicated earlier, the proprietary licensing model is incompatible with the preservation of those guarantees, and therefore the differential treatment between proprietary and free programs and public administration becomes inevitable.

I send you my regards, expressing once again my appreciation for your contribution to the discussion,and inviting you to continue the debate for the benefit of all of society.

Attentively,

Senator Alberto Conde (UCR)
H. Chamber of Senators
Province of Buenos Aires



Hosted by GrULiC Valid HTML 4.01! Last modified: 11/10/03