[Open-Legislation] French civil code on git

stef s at ctrlc.hu
Wed Apr 8 14:57:05 UTC 2015

On Wed, Apr 08, 2015 at 12:20:12PM +0100, Francis Davey wrote:
> UK legislation is very very much more difficult to work with than the patch
> amendments mentioned above.

that is interesting, why? i see that all parliaments have different ways to
dodge accountability, would be interesting to create a blacklist ;)

> But we do not yet have a complete accurate machine-readable corpus of
> up-to-date legislation,

i can imagine. what's the oldest law in force in the uk? i assume it's much
older than the earliest EU "laws".

> Many jurisdictions have the same, or worse, in terms of formatting of
> amendments, but most that do also spend money on maintaining a codified
> version, we haven't done so.

apparently there is a huge difference in MS handling of this stuff. would be
interesting to have a comparison.

> > what exactly is your goal?
> Read legislation in a simple and clean way and see how it has changed over
> time (i.e. amendments). If there are other ways to do it, then great.

i think the regardscitoyen tool is very cool domain-specific tool for
reading/browsing. have a look:

> But also requiring a bit more effort to read.
> I know a lot more people who can read github (or any other web based git
> repository such as bitbucket) than can put together SQL queries to read
> material from a database dump. Sure, I can do both quite happily but I am
> untypical as lawyers go.

i think this is a misunderstanding. of course i do not mean endusers to
consume a db dump. and i see how people see synergies in git and the blame
feature, but - even if the source data is good enough to actually attribute
changes - this feature can be had also with other tools, that are in general
better suited for this specific domain.

the thing is the goal defines the tool, and i have yet not encountered a goal
where git is the right tool (except for publicity stunts). for example
consider that the history only shows the winning amendments, all the
non-winning will not show up in the history, or we have tons of branches (like
as much as representatives / proposal). i guess MS parliaments might go a bit
smaller in size, but consider the Common Agricultural Policy in 2013 had more
than 8000 submitted amendments. in the committee phase. navigating such size i
doubt is reasonable with git. and it is actually this kind of proposal that is
interesting to most people, i guess attention is proportional to
count(amendments). furthermore consider the very different user needs of such
a tool, an activist needs very different view, than an academic, or a
journalist, or a staff member of a representative. 

i think the git topic died out, but turned up some quite interesting points we
should ponder a bit more:

 1/ some catalog of ways to dilute accountability, or obstructing citizen
 2/ use cases and tools, like naming-shaming, campaigning, amending, browsing,
 diffing, reverse-attribution, etc.

i guess this "githubber" succeeded at least in one goal, getting this topic
discussed again. as a publicity stunt, i have to give credit for that ;)

otr fp: https://www.ctrlc.hu/~stef/otr.txt

More information about the open-legislation mailing list