Tuesday, December 22, 2009

Sen. Conroy's Censorship Plan

In the last week, Sen. Conroy, the Minister for Broadband and Communications, has released a report on his proposed mandatory internet filter, and aims to introduce legislation during 2010 to bring it into effect.

Having read the report, it is becoming increasingly clear that the Mandatory internet Filter is technically, financially, and morally a failure, but Conroy seems determined to go ahead with it anyway.

Technically, the report shows '100%' success in blocking the ACMA list of sites, around 1100 at the moment. That was a requirement for the 9 ISPs participating in the trials. Blocking 1100 URLs is trivial though, and the ACMA list generally only grows by people complaining about sites. Some participants in the trial used more dynamic filters to block web sites, which were then checked against two blind lists, one compiled of material that would generally fit the 'RC' classification, and one containing innocuous material.

When measured against those standards, success fell to between 78.8% and 84.6%, while the over-blocking rate (erroneous blocking of materials that should not be), was between 2.44-3.37%, which is quite high by industry standards, and against the 3 trillion and more web sites, a staggering percentage. There is in general a correlation between blocking and over-blocking, so the more effective a filter is, the more over-blocking it will incur.

Add to this the abject failure of anti-circumvention measures, up to only 16% in the case of the ACMA list, compounded by the fact that the ISPs themselves recognise that any 'technically competent user' can byspass filtering of this kind, and the internet filter begins to seem less and less 'feasible'. (The ACMA+others filtering tested circumvention blocking between 37.8-94.5%, though again the increased circumvention blocking generally correlates to over-blocking and increased speed degredation).

In addition, the standards used by Enex to determine what speed loss was 'negligible' are open to query. Anything up to 10% was deemed negligible, and even 10-20% speed losses were considered 'minimal'. Given that the testing was done with (a) opt-in by an unknown customer base, (b) speeds only up to 8Mbps, and (c) on IPv4 protocols only, the impact of large-scale filtering is yet to be tested.

Lastly, the filter does nothing to address the fact that most of the material it claims to be filtering is transferred not over web sites, but by peer-to-peer traffic and other internet means, against which the filter is powerless.

In conclusion, the technical aspects of the Internet Filter are a failure: ineffective, circumventable, detrimental to internet speeds, and likely to erroneously censor legitimate material.

The Real issues in this debate should be moral and political ones. ACMA's own reports reveal that when given the opportunity to obtain free PC-based filtering software, most parents, 2/3 and above, did not do so and felt no need to do so. If parents of Australian children do not think filtering is the right option, why does Conroy think filtering it for all of us is the answer? Is he a more morally just and wise person than the rest of us?

Labor's own "Plan for Cyber-Safety" outlines a number of threats to children online, which include bullying, identity theft, online predators, all of which this plan does nothing to address. In contrast, the threat of accidental exposure to adult material is relatively low. (In case you think the Filter is not designed to address such threats, it is listed in this document, by Sen. Conroy, as the first option in addressing Child Safety online).

Despite Conroy and other's continual rhetoric that it is all about child porn and other explicit illegal material, the Refused Classification category remains both broad and vague. The ACMA list leaked earlier this year (despite Government denials, it was admitted to be fairly close), contained sites including a dentist, as well as political sites relating to euthanasia and anti-abortion.

Since the ACMA's list is maintained by ACMA employees and officials, and not open to public scrutiny or review, and since Conroy proposes to increase the number of filtered URLs in consultation with foreign governments and software filter providers, there is nothing open and transparent about this process. Furthermore, the secret nature of the lists ensures review and appeal is not available. In effect, the term 'censorship' is entirely appropriate to such a scheme. There is no guarantee, beyond Conroy's repeated word, that the prohibited materials will not expand to include political content, anti-censorship information, and legal materials deemed 'inappropriate' (as tragically the Australian Christian Lobby has already said it would campaign for).

While it runs the risk of hysteria to say this, Sen. Conroy's proposal is unique among Western democracies, and uniquely flawed. Comparisons with Iran and China are not misplaced in this instance. It is indeed a censorship regime, one that will ultimately affect the technologically naive rather than the savvy. It will be a huge waste of money, ineffective in its goals, damaging to democratic ideals and debate, and do little to address the victims or perpetrators of child porn and other illegal activity. I call on all Australians to make their voices heard on this issue.

Friday, December 04, 2009

Reasons why Logos is great

(Sorry Mikey).

Reasons why Logos is awesome:

1) Commentary sets.

They're not cheap in print, are incredibly bulky, and you rarely want to read one end to end. (Jobes on 1 Peter is an exception). You do want to have an electronic version, on a laptop, for overseas mission, at a much reduced price.

2) Searchable Lexica:

Having LSJ, granddaddy of all Greek lexicons, has probably cut into 4 the time it takes me to work on patristic translation. There are few real advantages to turning the pages of a hard-copy lexicon to find the entry you want.

3) Importable Greek sermons.

Add 2 to this one, and Chrysostom translation got a lot easier. Suddenly some high-end work got a lot less labour-intensive.

I think that's the answer to Mikey's and others' criticisms. Sure, Logos always want to bundle a bunch of stuff no one wants, and sure there's a bunch of public-domain stuff there, and sure you can cripple your language skills (but that's usually one's own fault), but pick what is gold, and use it well, and bible software is for the win.