Open Source City: Making Public Data a Platform for Participation

[ryan_mcgreal]

http://raisethehammer.org/article/893/

It's time for the City of Hamilton to start publishing its publicly available data in an open, accessible format instead of today's hodgepodge of closed, clunky and idiosyncratic legacy formats that are hard to find and even harder to use.

[FairHamilton]

I so agree.

Once when I was looking for something on a annual event for 2009 event and the searches kept returning infromation on the previous years events. Why are they still around and easy to find, when there was nothing on the current year?

Frustrating. Time for a complete archiving of historical information, and overhaul to a 2009 design and funtionality.

[thistleclub]

That site has always been laughably oblique. The taxonomy is passive-aggressive. It's like someone handing you a 64-channel mixing console when all you want (or in many cases, need) is a volume knob. Part of it is that they did it on the cheap. Part of it is that they aren't good with people. Either way, time for a rethink.

Even something as simple as putting a FAQ link next to the search box (rather than buried at the bottom of the page) would help.

[Millstone]

The design is so broken it almost seems as though it were intended specifically to discourage any citizen who dares to try and look something up.

Kind of like that webpage. Yeesh.

The City of Hamilton website isn't that bad, but needs to be categorized a bit better. Googling most stuff like "hamilton tax" or "hamilton parking ticket" points you in the right direction.

[astroblaster]

i'm all for it.

i especially want open access to data about historical construction dates

[emge]

I agree with the concept.

What are some cities that have truly excellent websites and well-organized, accessible data?

[BrianE]

As far as data services goes the City does have a bang up GIS department. their Map.Hamilton.ca website is miles ahead of other municipalities.

Although, as I look at it now... it hasn't updated their Imapper in quite some time. Disappointing. Still works tho.

[sofasurfer]

Definitely agree with your principle.

But I'd take slight exception to your implicit assertion that HTML is a suitable standard for long-term access and archival - it's way too browser-specific these days. At least with XML or something similar, it's feasible to then transform the underlying structured data into something readable/meaningful for a particular audience/platform.

(disclaimer: I spent a couple of years working on a UK/european project that was trying to scope out the whole "digital preservation" can of worms - XML was kinda forced on to us when it was still a standard in development, so this stuff left a scar )

That said, if we're talking transparency, cost-effectiveness, efficiency... SOME form of open standards are the only way to go.

(IIRC, didn't PDF become an open standard not so long ago? Pretty sure Adobe opened it up - which is a good move, cos at least a de facto publishing standard is now wide open. Of course, there's the side issue of just how many PDF docs are actually well-structured and well-formed in their own right...)

It's a great chance for Hamilton to take a lead in something that I strongly believe will become the norm in the long run. So why not lead early and excel from the get-go?

[ryan_mcgreal]

Originally Posted by sofasurfer

I'd take slight exception to your implicit assertion that HTML is a suitable standard for long-term access and archival

I don't think it is a suitable standard either, but not because it's browser specific - after all, any programming language worth its salt already has a good HTML parser (my personal favourite is the elegant, malform-tolerant BeautifulSoup library for Python).

The real problem with HTML is that it's designed for human-readable documents, not for data. Dumping structured data into tables or definition lists with defined id and class attributes is just too sloppy-joe for a clear, consistent, and maintainable API.

I personally dislike XML mainly because it's so needlessly verbose (and also, to be honest, because it's abused so horribly and so often), which is why I prefer something like YAML or JSON. XML does have nice built-in transformations via XSLT; but again, any decent programming language can easily parse structured data into human-readable presentation formats.

Originally Posted by sofasurfer

IIRC, didn't PDF become an open standard not so long ago?

You're right: it was opened formally in January 2008 and given an ISO standard later that year. [Aside: There's actually a Javascript library for dynamically generating PDFs that looks quite promising.]

However, there are two big problems with PDF as a data standard:

1. PDF is specifically designed for printed documents, so it's even less suited to structured data than HTML. The set of available objects are oriented toward 2-D presentation on a page, not toward data linking.
2. The standard is a bit baroque (mainly to provide the ability to position elements on the page and to include both vector and raster graphics).

In any case, I certainly can't imagine any credible argument in favour of using PDF to facilitate programming access to structured data rather than a format that's actually designed and intended for this purpose.