Guest post from Dave Bath – can we please have RSS feeds from Auditors General (and other agencies methinks [NG])

A very reasonable request – so it seems to me – from Dave Bath who has asked me to post the guest post below. I guess there’s a message there – not just for Auditors General but for all right thinking government agencies.

It’s bleg time… for people who’d like to get all Auditor General offices to provide an RSS feed of announcements and reports.  I’m after ideas (a list of questions near the bottom of the post) and even cosignatories for a request that each audit office provide such a feed.

These days, with Gov 2.0 a buzz-word, and given the excellent work by Troppo’s own Nicholas Gruen, you would expect almost all agencies to provide RSS feeds, and preferably, like the parliament, a range of news feeds for different purposes.

You might expect that an Auditor General would provide RSS feeds, as such an agency is not responsible to the executive, but to the parliament.

Yet only the audit offices of Western Australia ( and the federal ANAO ( provide RSS feeds.

I’d searched both with Google ( and visually through the Victorian Auditor General’s site ( for news feeds, with no joy, so sent a nice polite email to the office asking for a link to their feed for new reports and/or media announcements.

I received a reply, but no joy.

Dear Dave,
We do not have an RSS feed for our reports as we have a subscription service.
We use this purely to send alerts when we publish our reports.
You can subscribe vis this page:
Lesya Bryndzia
Communications Assistant
Victorian Auditor-General’s Office
Level 24, 35 Collins St. Melbourne Vic. 3000
t: (03) 8601 1666 f: (03) 8601 7010 e:

Well, I knew all about the subscription service – I get those emails.

But I want an RSS feed, and because setting up an RSS feed is pretty trivial compared to an email subscription service (with all the costs associated with keeping an address list of private citizens private), I’d have thought the agency responsible for reviewing value-for-money issues would jump at the chance for a cheap way of making it’s work well known.

The privacy issue of an email subscription list is non-trivial – especially as subscribers are likely to be a grumpy bunch, wanting the official dirt on government – a list of people of interest to political minders.

It is hard to think of a valid reason why an audit office cannot or should not provide RSS feeds – after all, the audit offices federally and in WA do or are in the process of doing so.

So… the bleggy bit…

  • Do you think all audit offices should provide RSS feeds?
  • Do you think RSS feeds are trivial to implement?
  • Should a common letter be sent to all audit offices, (excluding the “good guys” in WA and the ANAO)?
  • Should the national Auditors General Club ( get a copy?
  • Should government and opposition spokespersons get a copy of the letter, (and if so, in your state, who are they)?
  • Would you consider being a cosignatory to the email I plan to prepare?
  • Do you want to write individually, to the office in your own state?
  • Do you have an draft fragments you think should be included?

I’ll probably work on a draft, putting it in a follow-up post for further comments.

This entry was posted in Web and Government 2.0. Bookmark the permalink.
Notify of

Newest Most Voted
Inline Feedbacks
View all comments

[…] will probably be more comments on this over at Club Troppo which has a slightly different version of this post.  A big thankyou to Nicholas for the […]

Stephen Bounds
13 years ago


You’d be amazed how many government departments are using antiquated systems that don’t support RSS. My guess would be that the subscription service is completely manual, ie some schmo is responsible for manually triggering the subscription emails each time after updating the pages. To create an RSS feed would be a similarly hand-managed job.

So it’s not a question of complexity but technological capability.

Note that in some cases you can simulate RSS feeds through page scraping of the relevant pages. I’m pretty sure you wouldn’t get a take down request :) This might be a quicker approach??

Dave Bath
13 years ago


I *wouldn’t* be amazed about antiquated systems. (Hell, there’s one I put into VicRoads that has been running near-unattended since 1988).

I’d expect the bulk emails are fired off manually – after manual preparation of the email.

There are a zillion different ways to skin this cat technically and cheaply – and Troppo’s BitWrangler Jacques could make quite a few. (Is is simply a matter of dropping in a WordPress appliance and hooking it up to a feedburner?)

The biggy though is that managing privacy-related information costs money.

13 years ago

The answer to all the questions in a word: Yes

However, until such a time we can make use of the Firefox Add-on: Update Scanner 3.1.4

Dave Bath
13 years ago

(1) Screen scraping for me is not the problem – I get the emails. Any individual broadcasting is unreliable and non-authoritative, and therefore distribution will always be limited – contrary to the purpose underlying existing announcements.

(2) I put my focus on audit offices because, like, they are maintained (and funded) outside the day-to-day whims of the executive. While they are responsible to the parliament, there is the duty to the public (explicitly in Qld with the Right To Information Officer). Nicholas is correct saying RSS feeds should be available from all “right-thinking agencies”, and while a central aggregator and set of particular feeds could be managed within a central government “feed warehouse”, feeds from agencies outside ministerial directions, the parliaments, ombuds and audit offices, are special cases.

(3) RSS feeds are convenient, easily aggregated, almost universally familiar to wonkish information consumers – which is why RSS feeds are universally used by political parties themselves, and most agencies. The services offered by are a good example for all agencies. ANAO has already created a twitter stream, which is easily grabbed as RSS, but this is not specific to reports.

(4) Cuurently, media announcements and the like from agencies that do not have RSS feeds are through what seem to be static pages, with poor tagging (you’ll all be familiar with tags from your own blogs).

BTW: There is a slightly longer version of the above guest post, with links to the various audit offices and (most of the) contact points, back at Balneus, which I’ll expand/reformat as I get more information.

13 years ago

But I want an RSS feed, and because setting up an RSS feed is pretty trivial compared to an email subscription service (with all the costs associated with keeping an address list of private citizens private), I’d have thought the agency responsible for reviewing value-for-money issues would jump at the chance for a cheap way of making it’s work well known.

There should be a relatively simple algorithm for converting a subscription into an RSS: feeding the email into a robot and then republishing. I guess it depends on the format of the emails. Not sure about the legality of republishing, I doubt they would freak out about it, but potential penalties might theoretically be large *shrug*.

May I just say, that problems like this are really signs that the WWW was badly designed, because if links were content-addressable rather than origin-addressable things like RSS would be intrinsic to any website rather than an extra tack-on (see for example how git works, that’s the way of the future). Of course it’s not too late to fix things, these things tend to get redesigned every 15 years or so.

Dave Bath
13 years ago

Tel@6 : The emails contain multiple reports announcements, and are basically freeform, so even republishing by those in control of the emails would be painful, unless assembled from individual parts.

the WWW was badly designed

Actually, the early web included semantic markup, HTML1 had not just “this is a definition”, “authoritative source over there, I scraped it at such-and-such a date”, “subject”, “cite”, keywords, author… but also semantic navigation – “next”, “prev”, “up” … and those semantic tags were respected and actioned by inbuilt buttons of the first GUI browser, NCSA Mosaic. Then the web was opened up to the public, markup was used for style rather than semantics (I still use “strong” and “em” rather than the default rendering of those styles “b” and “i” – but I typically write static html by hand), the semantic elements because unused, abused and finally abandoned.

Similarly, too few still use the full range of meta tags (Dublin Core) to describe the content, so semantic searching is very difficult. The Australian Government Locator Service relies on this, using meta in the head, not body, but it is usually poorly filled out – despite strong regulatory advice.

Hell… the meta tags are SUPPOSED to contain audience and sensitivity tags, and this would make authorizing views of pages trivial, and allow a default of “unless prohibited, display”, and then let google robots present all info to the public – the server inside the organization looking at the page to see if it was authorized to go “outside” … so FOI would be simple – you’d be googling within a site. Looking up laws would use the “Jurisdiction” and valid date ranges and keywords. See SOME of the semantic elements here, and you’ll see documents are actually supposed to be content addressable. But… as these are rarely filled in properly by organizations (indeed, if they exist, they can be misleading because they’ve come from MS-Word documents that have been cloned from documents with a completely different purpose, and the keywords and descriptions haven’t been adjusted).

Unfortunately, project managers don’t know the regulations, and thus don’t put them into requirements, while “architects” don’t make boilerplate chunks to be dropped into RFIs and RFQs. *sigh*

You can imagine the scars I got as EA Policies and Standards Analyst in a large GBE: to even raise compliance issues down the line would mean all the tier 1 types would have to admit negligence or incompetence in NOT having those things in place already.

Stephen Bounds
13 years ago


The reality is that no-one’s prepared to pay for the overhead. What you are talking about if pretty analogous to the work that used to be done by a battalion of recordkeepers, but is now (theoretically) devolved to individual staff.

But since it’s not their “real job”, there’s no incentive to make sure it happens.

And not to put too fine a point on it, I don’t think any of the non-compliant audit offices see that it’s a problem. After all, you can still access all the content you should be, it’s just a bit less convenient for you. So it ultimately boils down to a “principle” of how information should be accessed rather than a yes/no end-user capability.

Stephen Bounds
13 years ago

To be a little less pessimistic, I would personally try to take the path of least resistance to improve notifications from the various auditors’ offices.

Setting up a posterous account would be a good start, since then the agency would have an RSS-enabled blog that they could post to simply by adding the posterous address to their subscription email list.

13 years ago

I have audited several government websites, all of which are stuck on restrictive cms. RSS feeds were recognised as an issue and recommended. Luckily, more and more departments are upgrading, however remember we are dealing with large beasts.

Dave Bath
13 years ago

Windfery: Your point about beasts is appropriate, but raises more questions that are even more damning.

The point of any content management system is that content is managed, and therefore easily reworked.

The “fool with a tool is still a fool” adage applies – either the agencies didn’t buy a CMS, or they are not using it properly. The older the installation, the greater is the sin of continued misuse and lack of training.

Now, assuming some form of database underlying a CMS, assuming the structure is either documented or discoverable (if not, it was an inappropriate purchase), something more common with open source (even if not free) tools, then it should take no more than a week’s work for someone in the agency’s IT area to write a script (I’d probably use perl) that searches and constructs the feed, a week of testing and cutover (calendar)… So let’s say $10K.

This is made harder, but not intractable, when the “CMS” was built by a recent grad, unsupervised, without a proper requirements document, for about $1500 … A shocking tool, then misused… such as Alston’s National Office of the Information Economy (thankfully destroyed and replaced by AGIMO).

If, on the other hand, a CMS was COTS (the big bucks paid on the pretext of supportability), then assuming proper staff training and usage, migration or extension should be trivial.

The metadata requirements have been in place for over a decade … If obeyed, remunging or migration would not have been onerous.

Gordon Grace
Gordon Grace
13 years ago

In the last 5 years, I’ve been able to observe a steady increase in the publication of RSS feeds by government agencies, lagging slightly behind demand. Observe:

The web standards movement has started to take root in government – publishing platforms and procedures are maturing, and the benefits of separating content and structure from presentation and behavior on the web are made obvious when an agency is asked to support several online distribution mechanisms.

I agree that RSS feeds should be part of most (not necessarily all) web content RFTs. It’s not currently a mandatory requirement for government websites, but AGIMO does provide advice at:

Like many of the previous comments, I still dream of the day when agencies across the board are using metadata correctly. In the short term, I’d settle for good date metadata – at least then, any agency using the Agency Search service would be able to provide RSS feeds of custom searches across their web content on demand, without making any changes to the underlying CMS.

It’d be worth asking AGIMO what sort of take up they’ve seen of their whole-of-government media release RSS feeds – aggregated from across government at a portfolio level.

[Disclaimer: Previous AGIMO employee, current Funnelback employee, RSS fan]

James Dellow
13 years ago

I wrote this back in 2009, an introduction to RSS for local government and how to go about introducing RSS even if their current WCMS doesn’t support it: Don’t forget the RSS