Changing the Collabora Online default template language

The latest version of the Collabora Online Development Edition Docker image at the time of writing defaults to German language. This manifests as unexpected behaviour for English language input in new Collabora Online documents, such as quotation marks appearing inverted:


Fixing this involves changing the default language of the underlying ODF template files. In this example we’re using Collabora Online in a Nextcloud 12 host, running in Ubuntu 16.04:

1. Locate the ODF templates at /var/www/html/nextcloud/core/templates/filetemplates

2. Copy the templates to another system with LibreOffice loaded. Open them in LibreOffice (there are three files in ODF, ODS, and ODP format) and update the language to the English variant of your choice (for example English (New Zealand).

3. Copy them back to the Nextcloud host. Restart Apache and the Collabora Online Docker container, and confirm that new documents have the expected behaviour.



Run Docker in an LXD container

I’m a fan of Canonical’s LXD containers—which essentially copy the same approach to lightweight virtualisation enjoyed by Solaris Zones users (and by extension, any illumos-based distros such as SmartOS) for over ten years. One area however where Canoncial is playing catch-up compared to commercial UNIX is in incomplete documentation spread out absolutely everywhere—blog posts, articles, wikis, and so on. Trying to find consistent information on the level of support for Docker running in an LXD container is a perfect example of this. It’s a real mess.

At the time of writing, running Docker as installed from the official Docker repository will fail in an LXD container. This is noted in the following two bug reports:

The advice provided in both reports is to use Ubuntu’s Docker packages:

“Only Docker coming from Ubuntu ( package) works inside LXD containers.

“The Docker coming from upstream is missing a number of patches to make it work, leading to the problem you describe above. We’ve been pushing for those changes to be merged upstream and some were, but we’re not yet at a point where the upstream packages work.”

Otherwise, the prerequisite for running Docker in LXD is that the container is launched with the docker profile applied, and is configured as a privileged container (by default, LXC containers are unprivileged). In the following example, the nextcloud-dev-1 container is created with the default and docker profiles applied, and its configuration is set to be privileged:

$ sudo lxc launch ubuntu:16.04 nextcloud-dev-1 -p default -p docker -c security.privileged=true

Post installation, log into the container and install the Ubuntu Docker package:

$ sudo apt install

From there, Docker should work as expected.

More on privileged containers is here:


Flip NZ broadband’s constantly changing IP addresses

Here’s a sample of the IP addresses assigned to a Flip NZ domestic broadband connection in 30 minutes of use:


This leads to the expected behaviour with certain web-based services:


I’m also informed of at least one bank advising customers that online banking via Flip broadband is as a result borked. Performing tasks such as temporarily firewalling cloud services while performing remote setup and configuration is also certainly a waste of time.

Flip’s response:


I’ve used Orcon domestic broadband in the recent past, and while they of course reserve the right to reallocate addresses at any time, I’ve never seen addresses rotate literally every few minutes.

Long story short, this makes Flip useless for home office work – one hopes these limitations are made clear in their marketing collateral.


Watermarks template for LibreOffice Writer

For folks wanting a LibreOffice Writer template with a default set of high-quality, vector-based watermarks good to go, head on over to and download/share away. Any feedback or suggestions, let me know.

2015 in review

The stats helper monkeys prepared a 2015 annual report for this blog.

Here’s an excerpt:

The concert hall at the Sydney Opera House holds 2,700 people. This blog was viewed about 19,000 times in 2015. If it were a concert at Sydney Opera House, it would take about 7 sold-out performances for that many people to see it.

Click here to see the complete report.

Slack (the “email killer”), summarised

This post is for the benefit of those observing with curiosity what seems to be every second tech news outlet proclaiming Slack as the technological Second Coming, or otherwise some miracle tool that will have us ditching email en masse the day after tomorrow.

The worst exemplar of advertising masquerading as tech journalism in this regard would appear to be The Verge – who clearly cannot go a full week without some breathless, sponsored-story-level article on Slack. For instance, “More than a million people are using Slack every day now (Get your Slack facts)”, or “Why Slack could be the future of conferences”, or “That week I tried to unplug from Slack”, or (my absolute fave), “Slack is killing email (yes, really)”. More on the latter below.

Anyway, for lay folks wondering what Slack is and isn’t, read on.

Slack is a chat app, designed for internal, corporate teamwork.

Read the above title a couple of times, just to put the extreme hyperbole oozing from every journalistic quarter in perspective. Yes, it has “integrations” for a number of third-party services (Google Hangouts, Dropbox, basically any proprietary cloud-only service that comes to mind). Yes, it has a very nice user interface. Yes, it has lot of other bells and whistles, and so on.

But, at core, it’s a chat application, designed for internal, corporate teamwork.

Slack is yet another proprietary, walled garden.

All the media oohing and ahhing, mutibillion-dollar market valuations and slick features in the world cannot escape the plain fact that Slack is yet another SaaS-only/cloud-only/subscription-only/off-premises-only/rental-only product. Businesses can’t self-host it (should circumstances require), and remain entirely exposed to Slack’s whims regarding terms of use, pricing… basically every other well-known, well-documented risk associated with a SaaS-only product. Works great, until it doesn’t.

Another risk: Slack would make a very attractive target for acquisition by a larger player, perhaps only to be mismanaged into irrelevance (paging Microsoft-owned Yammer…). Heck, Slack’s CEO also founded Flickr, subsequently sold to Yahoo – and who exactly uses Flickr to any relevant degree nowadays?

Slack is expensive.

Assuming you want business-level uptime and support:

“With 25 users in the account, including external teams and contractors, that would mean we’re paying $200/month ($400 for business SLAs). Per month. For chat. Crazy.” (via Hacker News)

Slack is not killing email any time soon. Or likely ever.

With a nod here to The Verge’s ridiculous clickbait headline, Slack is not killing email. Slack’s own CEO has said Slack is not killing email (while also somewhat disparagingly referring to email’s standards-based, guaranteed-to-work-anywhere benefit as being “lowest common denominator”).

Smarter businesses figured out a long time ago that using email specifically as an internal communication and collaboration tool does suck, and augmented it with chat, conferencing, blogging, wikis… in all manner of closed and open source combinations. For these businesses Slack simply further validates these tools, and offers – if anything – an arguably incremental improvement. Slack will only come as a revelation to organisations that for whatever reason historically refused to look at email alternatives specifically for internal team-based collaboration.

As soon as you step outside of Slack’s walled garden and need to communicate with folks external to your business, it’s almost certainly going to be over email – one of the few electronic communication mediums guaranteed to work across every conceivable platform.

To paraphrase Larry Ellison: watching Slack “kill email” is going to be like watching a glacier melt.

Slack does not have unlimited accounts for free.

Yes, Slack’s marketing department claims this is the case (“Slack is free to use for as long as you want and with an unlimited number of people”), but as is the marketer’s wont, the reality is of course different:

cf. So Yeah We Tried Slack… and We Deeply Regretted It.

Yes, one could argue the folks in the article were naive about the free plan, but the point is a promise from any technology vendor should generally be considered worthless.

How not to stream on

EDIT: Due to some admittedly rather ambiguous wording on my part, I’d like to clarify (particularly for folks reading from wherever the surge in views is coming from) the few representative channels linked at the bottom of the post are folks who are doing streaming right. Hence why I am linking them for recommendation, and hence why they are favourited and followed.

After having watched for a few months now, my points of view on how to ruin your game stream:

  • Endless, non-stop jabber. Extra marks off for non-game-related jabber.
  • Thinking you’re the next David Letterman (who was incredibly unfunny anyway), or a budding talk-show host or comedian. Streaming is about the game first and foremost. Not how marvellously entertaining you think you are (but actually aren’t).
  • Having your voiceover over-cranked such that we can’t hear the game audio.
  • Having your voiceover over-cranked such that it distorts horribly every time your voice is even slightly raised. Combined with the above, this will have me switching off in no time.
  • Skipping over cutscenes, dialogue, critical pieces of in-game information, speed runs – instead of allowing the folks who have bothered to tune in to your channel to enjoy the experience vicariously. Studios spend millions on this stuff. The likelihood you’re improving the experience by skipping it? Nil.
  • Oversized webcam, chat, followers, and other assorted overlays that occupy anything more than a small fraction of the screen.
  • An excess of window decoration, animated or otherwise.
  • Having goddawful screamo music blaring over the in-game soundtrack and effects. Do you know how much money studios spend on producing quality audio to begin with?

Here’s a few streamers that I always tune into, who have largely successfully observed the above:

(Edit: updated for screamo music annoyance)
(Edit: updated for wording ambiguity)

Let’s *not* let them (NZ government workers) use Chromebooks

Technology journalist Bill Bennett muses in a post from a year ago titled “Let them (NZ government workers) use Chromebooks” about the potential cost savings and productivity gains to theoretically be had from deploying the Google office productivity stack for NZ government workers. Some excerpts:

“Put aside for a moment the security risks and the NZ$2 million paid to Microsoft for extra [Windows XP] support… One solution would be to write off all the existing computers and replace them with Chromebooks… There would be immediate savings. Chromebooks can’t run Microsoft Office. Government departments can shift to Google Apps… Getting all government employees and applications into the cloud means there will never again be a situation like 40,000 computers using out of date software.”

The security risks and $2 million dollar figure can be viewed as part of the exit cost of adopting the Microsoft platform to begin with. 13 years ago there may have been a reasonable case for Windows XP being an adequate desktop solution, which is clearly no longer the case in 2015. However, suggesting that NZ government trades one proprietary ecosystem (Microsoft) for one even more closed is not an advance at all. From a technology perspective, Google Apps for instance is completely welded shut – with a non-standard and entirely Google-secret document format at its core.

Yes, there’s a strong case to be made for the combined benefits of moving computing services off the Microsoft desktop platform and onto a Linux-based OS (Chrome OS is but one example, Ubuntu is another), replacing full desktop computers with thin clients, and adopting cloud-hosted applications. However, I would seriously question the wisdom of government migrating to a platform so tightly controlled by a single vendor – not to mention one that derives ~90 percent of its total global revenue from advertising, or that has a long track record of startlingly short product lifespans. It would drastically curtail competitive supplier choice, and in the case of Google Apps (and to an extent Chrome OS), eliminate the freedom to self-host the technology in-house or via a third party as and when the situation or requirement arises.

The $2 million dollar figure for Microsoft’s extended Windows XP support might well pale in comparison to migrating off a proprietary, cloud-only solution 13 years from now. The point being, any serious discussion must take into account exit costs, and not simply the superficial low cost of entry, be it Office 365, Chrome OS, Chromebooks, and so forth. This is an aspect Bill Bennett’s article omits, much like most other opinion pieces on the matter.

Apropos of this, in the time since the source article was written, the reasons for the UK government mandating ODF for document interchange (to the detriment of both Microsoft and Google) makes for worthwhile reading.

ASUS & McAfee’s cringeworthy antivirus campaign

Exhibit A:

ASUS and McAfee antivirus campaign screengrab.

Oh, if only it was this much fun. Which – if you’re the poor sod who’s ever spent hours having to wrestle with antivirus software such as McAfee that’s proved to be completely useless in removing any number of infections from Microsoft’s godforsaken products – it’s really not.

Enough with the cutesy ads ASUS, do the right thing by your customers, and give us some laptops with Ubuntu preinstalled with first-class support, please.

(By the way, ASUS products from a hardware standpoint are generally incredibly well designed.)

On Microsoft killing the Internet Explorer brand

The problem is, Internet Explorer – while a simply awful piece of technology – will forever be associated with being tied to an equally mediocre OS (Microsoft Windows). Open source it already, and kill off its dependency on Windows. Not rocket science.

Microsoft is killing off the Internet Explorer brand (The Verge)