Posts by Stephen Judd
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Want do you want, a medal?
Chill, Tom. This is what moving does. I felt it leaving Hamilton, I felt it leaving Auckland, I felt it leaving Wellington, and god help me I felt it leaving Kidderminster. I think you're reading some very different things into what Daniel is saying than he likely meant.
Also, I've just been on a Brazilian music jag, and I commend the word saudade to you.
Now go cross the same river twice.
-
"fish 'n' chips in Waitangi Park."
Sounds appropriate for the new austerity.
-
I see tax cuts don't stimulate the economy after all and apparently it's not ok to run a deficit. Oh well, at least we don't have an arrogant out of touch govt any more.
I learned about the evils of retrenchment and Gordon Coates in 5th form history -- did English skip class during that module?
-
Bart - I said "much government work", not "all government work." I only wanted to provide a counter to Rich's picture of government work as administration, not to increase your legitimate bitterness about the resources available to you.
Personally I think research is the last area where you would want to impose a standardised set-up -- researchers often need to build their own tools, never mind use recherche applications from elsewhere. Forcing your scientists to use Windows and Office for everything because it's cheaper is like giving your ballerinas standard issue steelcap boots.
-
the few blocks of the CBD that actually had an established big city feel about them, and I really enjoyed wandering around the streets just exploring.
I'd like an ad that voomed downstairs from the Cross St crew's pad, past the Wine Cellar, through Myers Park and then stopped off at every seedy noodle joint before staggering into the sea.
-
Can't speak to the first part of that criticism ;), but as to the second, I don't think that's necessary, as long as there are other benefits. Back in the day when Microsoft didn't have monopoly power behind it, people didn't go to Microsoft products because they were indistinguishable; they went because they met their needs better. No doubt people were saying "1-2-3 will never catch on, it needs to be so wouldn't know you weren't using Visicalc."
And there are other needs than "must be exactly like Microsoft."
For that matter, even Microsoft isn't indistinguishable from Microsoft, in the sense that changes between OS versions and Office versions can be quite marked. If you're faced with an XP to Windows 7 transition, the usability gap between XP and Ubuntu Jaunty hardly seems any bigger.
-
I have OpenOffice at home. It's by no means equivalent to MS Office - maybe 75%, with no features that MS don't provide that I've been able to find. As soon as I can find some sort of semi-hookey MS license, I'm going back to MS.
I reckon if you started giving your average government admin person a Ubuntu/OpenOffice desktop, then you'd be in for a good 2 weeks retraining time at a cost of several grand. That doesn't seem a very good deal to me.
There are several issues at hand there:
1. are there features of Microsoft products that are essential?
2. are the one-time costs of transitioning too high to justify the savings in license fees?
3. are we just talking about "your average government admin"? And are there employees who don't need a full desktop environment?On the first point, most people only use a subset of MS Office's capabilities, so OO may be enough. Many people only need to use applications that can be served through a browser, in which case Linux + Firefox is enough. Custom applications (which are the ones that often cost the most) can be built as well on a free stack (Linux + Postgres + whatever) as on a Microsoft stack. There may well be cases where some particular thing makes migration costly or difficult -- some crucial VBA macro or a SQL Server based data warehouse -- but there will be others where that is not so.
On the second point, again there are many scenarios where no particular transition costs apply: browser-based intranet apps or users who only need email + browser + simple word processing. You also have to bear in mind that in the enterprise world typically license fees are either annual, or the software is obsoleted semi-annually requiring new fees to be paid. So there will probably be plenty of cases where the payback period for retraining is quite small.
On the third point, much government work is not "knowledge work" per se but more data entry or process orientated. A more restricted computing environment is safer, cheaper, easier to administer and eminently feasible to build (and clone and rebuild and update) on a free software stack.
In any event, it would do no harm for government agencies to explicitly cost and evaluate free and open software solutions, rather than assuming that there is no alternative to Microsoft (or Oracle, or BEA, or whatever). Sometimes there will be.
Disclosure: I work for Catalyst IT, which specialises in FOSS solutions. But I thought this long before I started there.
-
Having watched the ad at last, it's a cute little sequence, and I agree with Gareth. Can't say I recognise any faces though. And that chap needs a helmet.
-
Auckland heeft meer meh.
-
Are services, such as Google Analytics and Nielsen, that use tracking code on third-party websites any different?
Fundamentally, they can't be. All their tracking code does is set cookies and request resources from the third party site that result in log files (or database entries) at the third party site. They have no way of capturing any more information than what you or I can capture with a log file and a willingness to slice and dice it. The value that Google Analytics or Nielsen provide is:
- they can benchmark against other sites
- Google Analytics can tie in with Adsense
- they do the statistical slicing and dicing for you
- they have pretty visualisations that encourage you to digAs to what can be captured:
- IP number (network and geographical location can be inferred from this with some inaccuracy)
- cookies (allows inference of returning visit, visit length, bounce rate -- it's based on inspecting cookies, and so depends on the browser accepting cookies and the user not deleting them)
- referring page (with more or less precision -- based on information provided by browser, and some browsers lie; this is inspected to extract google search terms too)
- browser type (some browsers lie)
- operating system (some browsers lie; it's possible to sniff OS type with some accuracy by low-level inspection of packets, but I doubt they do this)
- screen dimensions (relies on Javascript in the browser)
- time to serve a resource from first to last byte (connection speed can be inferred from this)Anyone who operates their own web site on their own server can do this. There is no special magic.
One of the naughty things these services do is imply a degree of precision that they don't really have. For example, if I tell you that you had 32,516 visits last week that sounds as though I have an efficient and accurate system, but actually, that's a best guess based on shagging with cookie data and the actual number could be higher or lower (and depends on how "visit" is defined too). It would be more honest to say "around 30,000, probably somewhat more."