Open Source Limits

Tony Healy, a research software engineer and policy researcher in Sydney, Australia has recently authored a 4 page report (hereafter termed as the Report) titled ”Has Open Source reached its limits?” for The Institute for Policy Innovation.

IPI,as the Report states is ‘a non-profit,non-partisan educational organization founded in 1987. IPI’s purposes are to conduct research, aid development, and widely promote innovative and non-partisan solutions to today’s public policy problems’.

In his Report, the author bases his conclusions on the following points:

[1] The numerous examples of Open Source deployments do not actually support the claims of success

[2] Agendas of advocacy groups mask the weaknesses

[3] Existing rationales for Open Source adoption are false

Thus he concludes that Open Source is not necessarily the best way to develop software. While it performs an useful function in a specialised computing environment, the demands of the mass market is not well addressed to by the model.

As a participant in a software development model which relies solely and completely on the content and intent of the principles of Free/Open Source Software, my views are necessarily prejudiced towards the inherent advantages of the paradigm. Thus, in this text, I’ll attempt to take a look at the basis for the Report and provide an insight into how these do provide an unique opportunity to the developing nations.

Agendas mask weaknesses - is this the fact ?

Free/Open Source Software development model is based on the concept of community participation. Including members from the software development community (the ‘hackers’), the industry as well as various other stakeholders forms one of the important cornerstones of the initiative. In his regular feature for the Linux For You, Frederick Noronha (fred at bytesforall dot org) asks this important question - ‘Why is the industry consortia in India so keenly interested in Open Source ?’. Indeed this has been the most pointed questions being discussed in the various forums and User Groups. However, stating that those pushing for the implementation of Open Source are those with vested interests who profit at the expense of software developers is simply missing the point.

The age-old precept of the goose-that-lays-the-golden-egg does apply to F/OSS development also. The Report finds that the communities involved in OSS advocacy are of 4 types:

[i] IBM
[ii] Hardware makers
[iii] Commodity firms and
[iv] Lawyers

Curiously enough the worldwide predominance of User Groups (UGs) as one of the vibrant community matrix seems to have escaped the study while IBM has turned into an industry tokenism. If IBM can be posited as a member of the community, why not Microsoft ? With the extent and reach of its Get the Facts campaign, this company has single-handedly provided innumerable talking points for presentations and pre-sales talk for the industry members of the F/OSS community ! More often than not, the ‘Fear, Uncertainity and Doubt’ I receive when talking about F/OSS is based on the humorous case studies provided by Microsoft.

In a bottomline driven business model that dominates the service industry, the very fact that big business powerhouses like IBM, Oracle, SAP among others see OSS as a viable model provides an insight into the potential nature of doing business the OSS way.

Businesses can only be too willing to put their might behind a software development paradigm when they see that it makes for synergy with their strategic objectives. Hardware makers are no exception and thus stating that their advocacy masks a selfish desire to make money - doesn’t really make sense. And even if that premise is agreed upon, then the vice-versa case does also make sense - hardware makers promote proprietary software to ensure longevity ! Wonder how NVidia would react to this conclusion ?

Examples are facile - are they ?

The Report takes on the example of the games industry to prove a point that OSS is not suited to mass marketing. This is more of an unintended mismatch or perhaps an intended disconsonance.

The premise of F/OSS is not yet focussed on the gaming industry where a lot of other external variables determine the success of games. Concluding that the concept of ‘copying code’ (and this something that F/OSS developers should strongly object to) does not lend itself well to this industry is only half the story.

The conclusion is also based on the studies of the academic demands of computing. Suffice to state that using F/OSS technologies, a complete computational biotechnology laboratory could be set up within INR 150,000. The cost using proprietary software is left to the reader to calculate. A software development industry dedicated towards providing such customised solutions is actively present (at least in India) and thus the comparison seems more like apples and pears.

Re-usable component based models are generated through the use-reuse and improvement of existing components. This, in a strict sense of the term does not lend itself well to being ‘derivative’. It is more ‘innovative’ in nature.

One of the first and foremost as well as the lesser visible aspect of the success of F/OSS has been based on the Internet. Thus, web developers and web development based software firms presented as the ’success stories’ do augment the claims that F/OSS is industrial strength and market ready.

Rationales are false - to what extent ?

F/OSS activists primarily state that the rationale for F/OSS adoption by governments and industry for a developing nation is 4 fold. It provides for:

[1] Economic freedom
[2] Political freedom
[3] Cultural freedom
[4] Social freedom

In his essay on ‘Why GNU/Linux is God’s gift to India ?’ (available here at the Wiki), Prof Venkatesh ‘Venky’ Hariharan (venky at indlinux dot org) makes a case for the adoption for F/OSS by the governments.

F/OSS creates a local talent pool of developers by making available world-class code available for study and improvement. In a peer-reviewed meritocracy based development model, this leads to a generation of developers better equipped to understand the nuances of mass markets. F/OSS also provides the societies the cultural and social freedom to develop computing paradigm in their own languages. Two leading examples of this phenomenon are Ankur Bangla Project and IndLinux

A major point of attacka against F/OSS adoption has been the Total Cost of Ownersip (TCO). But as Rishab Aiyer Ghosh conclusively proves, for developing nations, the license cost centre(s) in TCO make this a compelling reason for adoption of F/OSS.

This is not a complete and exhaustive rebuttal of the Report, and one hopes that someone more competent would take the onus on to provide a more compelling bouquet of evidence.

Leave a Reply

You must be logged in to post a comment.