Revealed: ISPs detail their Aussie net filter trials

A report by the Sydney Morning Herald last week indicated that South Australian Independent Senator, Nick Xenaphon could side with the Coalition in an attempt to block content filtering legislation in the Senate.

But ISPs involved in the first round of the trials have told iTnews that while discussion over how the different political parties will vote is "interesting", it won't stop the tests from going ahead.

"Until such time as the Government puts legislation before the parliament, discussion on who goes which way is pure speculation," said Anthony Pillion, managing director of Webshield.

"I've seen a lot of speculation but until there's real live legislation it's just an interesting discussion."

"My understanding is that if legislation is required and if it goes to parliament, then he [Xenophon] may vote against it," said Andrew Robson, general manager of TECH 2U.

"There's so many ‘ifs' we're wasting our time talking about it."

The chosen ISPs have wasted no time in placing orders for content filtering equipment, mostly from the United States.

Highway1 said it expects its equipment will arrive this week; TECH 2U expects to be installing equipment later this month, while Netforce was unable to provide a specific timeline for delivery.

Webshield, which already operates a clean feed service, told iTnews that it already has the equipment running and is "doing some preliminary [tests]" through the Government-appointed Enex Labs.

Enex will conduct all the tests on behalf of the six participating ISPs.

"Most ISPs will have to do the network integration once they receive their [filter] equipment," said Pillion.

"We did that integration five years ago so the same steps they have to go through now we've already done."

iPrimus was unable to respond in time for this article. The other ISP involved in the trials, OMNIconnect, did not return iTnews' calls.

Several of the ISPs also spoke out about criticism they'd received after their selection for the trials. But they said that in all cases, none had experienced it from their own customer bases.

"We've had calls from people - not customers - saying they'll boycott us," said Robson.

"But by boycotting the tests you're skewing the results the wrong way.

"Our customer base is mostly families and elderly users. There hasn't been a lot of anger from them against the tests."

Robson said he also had the support of business customers to conduct tests on their 100Mbps connections.

"The key thing is there is lots of false information out there and it's not helping," said Robson.

"I think people should be open-minded about the tests."

Highway1 general manager Nicholas Power said that customer response to the tests had been "40 per cent positive".

"Given we have mostly business customers it hasn't drawn as hard a response," said Power.

Pillion also defended against criticism that Webshield - or any ISP - would try to alter the test results to support a particular outcome.

"There's been a lot said that ISPs will go out of their way to make sure the results are good," said Pillion.

"Enex Labs are doing the testing. No ISPs have a hand in conducting the tests or the production of the results in any way. All I need to do is provide them [Enex] with access to our network and I'm told I'll receive a copy of the results, but I don't have a chance to fudge them."

Read on to page two for how the ISPs expect their tests to go

Although it is unclear exactly what tests the ISPs will run at this stage, those we spoke with gave mixed indications of how far they would potentially take the testing regime.

Highway1 is doing the basic ACMA blacklist tests.

"We're not concerned about going further," said Power. "The device we've got can do other lists but we haven't confirmed [further test plans]".

Netforce director Scott Atkinson also said his company would cover off the basic tests but may "try something".

Webshield will filter both the ACMA blacklist and also against what it called "a series of user profiles based on the range of filtering we do".

"I'm still in discussions with the Government over some other test scenarios I've proposed," Pillion said.

"I hope to have them included in later rounds [of the trials]."

TECH 2U is equally advanced in its thinking on the test front.

In addition to the basic government tests, TECH 2U has suggested and received approval from the Department of Broadband, Communications and the Digital Economy (DBCDE) to run additional tests on some of its filtering equipment.

"We're contemplating six scenarios," Robson said.

"In one case, the setup and testing of a box was going to involve a certain amount of labour. We suggested an alternative test be run on the same piece of equipment while it was set up.

"It costs the Department nothing - all it would take is an extra four weeks of testing. We're literally taking the same equipment and plugging it in in a different order to see if it alters the test results."

Robson said TECH 2U had also received additional testing suggestions from a range of technicians and members of the public.

These tests could feasibly be given the go-ahead if it doesn't cost the DBCDE anything extra, he said.

Another test will examine speed degradation on 100Mbps connections.

"This was a suggested test and if the claims of the [equipment] manufacturer are true it will be successful," Robson said.

"I'm going to stick my neck out and say that I think there will be zero degradation but we'll soon see if the brochure-ware stacks up. We're also trying to think of a scenario that could potentially trip the box up."

Robson was unable to reveal the nature of some of the other test scenarios but he had expectations for how they will run. He expected at least two would pass technically, one would fail commercially and at least one "will fail abysmally" from a technical standpoint.

"However, that's a technology that's been used on the Internet for years to try and filter content," he said.

Highway1 is also interested in how the tests will affect its "high-volume customers" that have 20-30Mbps fibre connections.

"URL blocking probably won't result in any real issues but if at some point the Government want to filter images as well, that would just kill it from a technical point of view," Power said.

Read on to page three for details on equipment funding, costs and test schedules.

ISPs participating in the first round of trials have until the end of June to complete the tests.

The official test period is six weeks but it appears that the weeks do not need to be consecutive.

"If things come up we may need to suspend the tests for a while," Pillion said.

"The test period could be reasonably fluid in its timeframe. It's not necessarily a six consecutive weeks-type arrangement.

"Every ISP has the opportunity to put things on hold."

The initial six-week block would focus on independently testing Webshield's existing services only. Other suggested tests will not be included in this period, Pillion said.

Highway1 said it expects to start testing from "mid-March". TECH 2U anticipates a start between late March and early April. All dates are dependent on when the filtering equipment is delivered, primarily from United States-based sources.

The DBCDE will, for the most part, cover the costs of ISPs participating in the trials. Equipment costs are covered but it is staffing where most of the real costs lie - at least for the trials.

TECH 2U expects to contribute a team of between 10 and 12 people on a part-time basis over two months.

Highway1 will have a project manager and two other staff "probably full time looking after the filters and requests".

Webshield - which already purchased the filter technology as part of its existing business - did not have any equipment costs to participate.

It did not, however, anticipate staff would need to put a great deal of time towards the trials.

"The Government has put aside money to pay for the staff members dedicated to the trial," Pillion said.

"We just have to submit requests for that money [to the DBCDE]".

Power believes the Government will need to examine more permanent funding arrangements if it pushes forward with a mandatory filtering regime.

"The cost of the equipment [we're testing' is probably beyond the reach of most small ISPs," Power said.

Ongoing assistance from the Government would likely be needed, he said.

The issue of commercial viability will also be covered by the trials. All participating ISPs will do so under what Robson called "detailed costing requirements".

"If filtering is technically possible, the sub-issue not being discussed is whether it's then commercially viable," Robson said.

"If it costs $600 per user, consumers won't be interested but a large company might jump at it."

Indeed, many of the end customers that will participate in this round of trials are businesses, rather than consumers.

"Our business customers want to have filtering available to help them implement their acceptable usage policies internally," Atkinson said.

Although OMNIconnect did not return calls for this story, an FAQ on its website also points to it testing filtering as just another service it can provide to business customers.

"Many businesses already have defined company policies in regards to information access," it said.

"Trialling this technology allows us to offer a filtering system that will give them the ability to enforce their policies."

A survey of 20,000 users by the broadband forum Whirlpool released today found that 90 percent of Internet users would choose to opt-out of any Internet filter.