Metal (Technical)

Weeeee’re back!

You may not have noticed, but my server crashed a while back (I did get one unprompted comment, so I’m not completely writing into a vacuum). Since it no longer runs anything business-related, and I had been looking for a cheaper server, I wasn’t in any rush to restore it.

I had already got an account with GreenGeeks. My motivation to move wasn’t that I was unhappy with Asergo (previously Easyspeedy), where I’d been hosting happily since 2006. It was simply that having a sports car to putter up and down the driveway isn’t the best use of resources, and a much cheaper Golf Cart would do just fine.

GreenGeeks state that they have been running since 2008, and contribute back 3 times the power they consume into the grid in the form of renewable energy. I’m already hooked. And starting at $2.95 per month, rather cheaper than what Asergo are currently asking for a dedicated server, it seemed a no brainer.

Except that, by the time of the crash, I’d been running both in parallel for a while already, so was in effect paying $2.95 more than before.

When the server crashed, I thought it’d be a good time to move.

Except that the experience wasn’t particularly fun. Used to running my own server entirely, where I know what to do, and can freely access and customise what I need, I found the limitations of working through cPanel immensely frustrating. I needed to contact support multiple times to allow me to ssh into the server (no fault of theirs – it automatically deactivated when unused for a certain period of time), but when, after months of delay, I finally set aside the time, to be unable to even ssh in wasn’t a great start. Through cPanel, I couldn’t find out how to access the server error logs for the secondary virtual host. The primary logs were accessible through the interface, but nothing else. Accessing an error log on my server would take seconds, and I could find and fix the actual error. Now, after fumbling around for way too long, support would have to be contacted, and an evening wasted.

I have no complaints with GreenGeeks, but time is a little more important to me these days, and the thought of wasting endless hours on stupid obstacles like this became too much, and I returned to my trusty sports car.

Besides, any day now, I could be releasing the next Twitter or Zoom, and unless it uses Silicon Valley levels of compression, running it from a $2.95 server is not likely to happen.

A clean install was a chance to run some up-to-date software, and it now runs the latest version MariaDB 10.4 and PHP 7.4, up from the dated MySQL 5.5 and PHP 5.6 releases I was running previously. I had to do a bit of rewriting to get everything to work with the latest PHP, making me realise a) how rusty my coding is b) how ancient some of the scripts living on the server are, but it was soon running smoothly again.

Still, it’s up, and like a new toy, you may see more posts than usual for a while. And for the one person that noticed my absence, thanks for watching!

Related posts:

Image from Wikimedia Commons

Metal (Technical)

Fourteen years of silence

But it’s been 14 years of silence
It’s been 14 years of pain
It’s been 14 years that are gone forever and I’ll never have again

Yes, the lyrics of a Guns N Roses classic, and yes, also how long it feels to have been locked up for two weeks alone in my house due to South Africa’s coronavirus restrictions (we are not allowed out to go jogging, or even windsurfing).

But 14 years is also how long its been since I wrote about Free Software attempts to create an alternative to the then-dominant Skype.

At that time, the two candidates I looked at were Wengophone and Ekiga. Both didn’t cut it back then, and sadly neither ever did. Ekiga’s last release was in 2013, and Wengophone was rebranded as Qutecom in 2008 after the original sponsors, Wengo, handed over ownership to another party, and development stalled.

With both projects effectively dead, has another champion stepped forth?

Enter Jitsi, which began as a student project in 2003. The real game-changer was the introduction of WebRTC, which allowed Jitsi to morph from a desktop app into the modern Jitsi Meet, which essentially allows communication with the use of a browser only.

Initial WebRTC support was added in 2013, bringing us to today, where it takes nothing more than visiting to start video calling with multiple people.

Jitsi is actually a number of projects. Jitsi Desktop is no longer supported by Jitsi, but there is a community effort to keep it going. However, the one getting most attention is Jitsi Meet. It’s a completely Free and Open Source video conferencing solution, fully encrypted, offering all the usual features such as chat and screen sharing. It’s possible to run your own dedicated Jitsi Meet instance, or to use one of the publicly available ones. The official Jitsi instance,, is available for anyone to use at no cost.

Jitsi is integrated into Zulip, a distributed Team Chat I use both at work with the MariaDB Foundation, and with Wikimedia South Africa.

So, how well does it work? The results have been mixed to far. We tried Jitsi a number of times in 2019, and while it was usable, voice quality was not as good as alternatives. Note that there’s a difference between 2-person calls, in which participants communicate via peer-to-peer (P2P) or calls with more than 2 people, which are made via the Jitsi Videobridge.

The instance on uses soft moderation, where it assumes everyone is a responsible participant, and can mute others if they, for example, are talking loudly on another call and have left their mic on. This probably wouldn’t work in all contexts, with a lot of anonymous participants and the risk of someone kicking off the coordinator.

Moving forward to lockdown 2020. Everyone is rushing to move everything from tai chi classes to date nights online, and there’s been a scramble for solutions. Zoom, the proprietary video conference software, has probably been the biggest beneficiary. However, Zoom is not open source, and has been challenged on its privacy protection, as well as on its claim to to support end-to-end encryption.

Jitsi security is transparent, and Jitsi have released a statement on Jitsi Meet Security and Privacy if you want to know more.

This past weekend proved serendipitous for my Jitsi use. First, a meeting I was attending ended up using Jitsi Meet by accident. The video conferencing meeting wasn’t set up in advance, so, since we were already in Zulip, someone created a link to a instance. I was late to the meeting, and a little reluctant to rely on it given previous negative experiences, but call quality was great with about eleven people participating.

Literally seconds after the conference ended, I got a message from someone else, asking what I knew about Jitsi, as they were keen to try it out. I gave them feedback, and we set up a small test, with multiple different devices, including laptop browsers, Android and Apple devices, and pushed it by activating video on all. It passed with flying colours.

Intrigued, I pinged and got a better response time than with many other high profile, responsive sites, including Google.

So Jitsi look like they’ve ramped up their capacity. I’m keen to watch their progress, try them again, and quite possibly install my own dedicated instance.

Related posts:

Earth (Arts and Literature) Fire (Social)

South Africa in the 2019 World Press Freedom Index

It’s been a while since I looked at the World Press Freedom Index. South Africa was ranked 21st in the world when I first started looking back in 2003, and had slumped to 52nd back in 2013.

Looking at the rankings since, there has been a notable improvement from 2013 until 2017. While the Nordic countries are as usual on top, South Africa has been ahead of the United States each year since 2013, is currently ahead of the UK as well, and is fourth in Africa, behind Namibia, Cape Verde and Ghana.

2018 press freedom rankings2018 Press Freedom Map

Looking at 2019 index specifically, of the BRICS countries, South Africa is far ahead, with Brazil 105th, India 140th, Russia 149th and China a dismal 177th.

While the trend has been positive, the last two years have seen setbacks, in particular 2018-19. The report attributes this to harassment by state security agencies spying on some journalists, as well as intimidation campaigns by politicians. Much of this has been of ruling party politicians accused of corruption, but particularly noteworthy was the EFF’s attack on journalists, with their supporters making death threats and rape threats, fueled by their leadership.

As far as media diversity goes, there’s been some shift, with online-only media starting to make an impact. Daily Maverick, which launched in 2009, gaining prominence as an alternative media voice, and Ground Up, which launched in 2012, a small media house focusing on high-quality, ethical journalism. The once-dominant Independent News and Media is a shadow of its former self, with accusations of interference in the editorial process by its owners, and many credible journalists leaving. 2015 also saw the launch of the first Xhosa-language newspaper, I’solezwe lesiXhosa.

Below is the table since 2003. Note that a lower score is better, and also that the scoring system has changed over the years, particularly in 2012 and 2013.

Year Score Ranking
2019 22.19 (-1.80) 31 (-3)
2018 20.39 (-0.27) 28 (+3)
2017 20.12 (+1.80) 31 (+8)
2016 21.92 (+0.14) 39 (=)
2015 22.06 (+1.13) 39 (+3)
2014 23.19 (+1.37) 42 (+10)
2013 24.56 (-12.56) 52 (-10)
2012 12 (=) 42 (-4)
2010 12 (-3.5) 38 (-5)
2009 8.5 (-0.5) 33 (+3)
2008 8 (+5) 36 (+7)
2007 13 (-1.75) 43 (+1)
2006 11.25 (-4.75) 44 (-13)
2005 6.5 (-1.5) 31 (-5)
2004 5 (-1.66) 26 (-5)
2003 3.33 21 (+5)

Related posts

Fire (Social) Water (Personal)

Farewell Matthew Buckland

It came as quite a shock to hear this afternoon that Mathew Buckland had died.

I first met Matthew at the Digital Citizens Indaba (DCI) in Grahamstown in September 2006, where I was speaking on the Web 2.0 panel. Matthew was editor of the Mail & Guardian Online at the time. Mail & Guardian online was then the pick of the local online media, punching way above its weight, and much of this was due to Matthew’s leadership.

I remember him passionate as we explored the possibilities of new media, and friendly and generous with his time.

Matthew Buckland at the DCI in September 2006. Picture by Gregor Rohrig.

Thinking back to when the DCI took place, it was a dynamic time, full of possibility and new creations. Muti had come out in January 2006. Written by Neville Newey, apparently I inspired him to write it and offered some ideas, but in reality I had little to do with it. I think such was the vacuum in those early days that this was why I was invited to DCI! The Rat and Parrot, and many great conversations. My first encounter with the famous Mushy Peas on Toast. Blogging was revolutionary and everything was possible. In the weeks, months and years following the Digital Citizens Indaba, there was a burst of innovation. And Matthew was at the heart of much of it.

Amatomu, which Mathew co-founded, and Afrigator launched within days of each other. Thought Leader launched in 2007. Memeburn a few years later, and he was MC at the launch of Silicon Cape.

The word inspire can be defined as to fill others with animation, a quickening or exalting influence. Reading the outpouring of support since the announcement, it’s clear he touched many people’s lives, and left a legacy far more influential than he probably realised.

Go well Matthew.

Related posts

Fire (Social)

Election arithmetic

It’s almost election time, and with 48 parties standing nationally (and a further 32 standing only at provincial level), there are lot of hopeful party leaders out there.

It’s safe to say the majority of them are going to be very disappointed when the results are announced. Let’s look at some election arithmetic:

First, nationally:

Year Parties contesting Parties that won seats Parties that didn’t win seats
1994 19 7 12
1999 16 13 3
2004 21 12 9
2009 27 13 14
2014 29 13 16
2019 48 (14?) (34?)

Even if the vote is split more than ever before, and, optimistically for the small parties, a record 14 parties are returned to parliament in 2019, that still leaves 34 parties that will fail to win a seat. 34 parties, more than have even contested before, that have spent R200 000 just to register, and who knows how much more to campaign (judging by the activities of some of the parties, that figure will be close to zero) for the promise of making a difference, or at least a cushy job in parliament, and will be sadly heading back to reality the day after.

What exactly did it take to win at least one seat in previous elections? The following table shows the smallest party (all earning one seat, except for the ACDP in 1994), and the number of votes they won to earn this.

Year Votes Smallest Party
1994 88,104 ACDP (2 seats)
1999 27,257 Azapo
2004 39,116 Azapo
2009 35,867 APC
2014 30,676 APC

Election queue 2004
Even if everyone in this queue votes for the same party, it’ll take almost 1000 of these queues for that party to earn a seat

What about provincially? In the Western Cape province:

Year Parties contesting Parties that won seats Parties that didn’t win seats
1994 14 5 9
1999 15 5 10
2004 20 6 14
2009 22 5 17
2014 26 4 22
2019 34 (5) (29?)

Do parties have a better chance of getting elected in the smaller pond of a province?

Here’s the Western Cape equivalent of the earlier table.

Year Votes Smallest Party
1994 25,731 ACDP
1999 38,071 UDM
2004 27,489 UDM
2009 28,995 ACDP
2014 21,696 ACDP

Looking at just the Western Cape, the answer’s a resounding no, parties don’t have an easier time provincially. You can even say it’s harder, as parties have to win a much higher percentage of the available votes to win a seat. Every party that managed to win seats also won seats nationally, leaving no room for the smaller, provincial only parties, while the reverse was very much not true. The majority of parties that won sets nationally failed to win a seat in the Western Cape. The most parties that have ever been represented was 6, and currently there are only four represented. Bad news for the 10 hopeful parties that have registered for the Western Cape only.

Again, even if we see an increase to 5 parties winning representation, a record 29 parties, more than have ever even contested before, will be sorely disappointed.

Notice that the number of votes won by the smallest parties is not much less than nationally. And in 1999, it was a lot more.

New parties have a hard time breaking into parliament without something major happening. A major split from an existing party, or a high-profile personality entering politics. Looking through the years, there are very few exceptions


1994, as the first non-racial election, can’t be compared to a previous election, but of the seven parties winning representation, only one, the African Christian Democratic Party, didn’t have a substantial political presence before.


In 1999, the number of parties elected nationally lept from seven to thirteen. All seven represented parties were returned to parliament, while the six new parties were:

  • The United Democratic Movement, made up of prominent former ANC and NP politicians.
  • Azapo, a prominent liberation movement which had boycotted the 1994 elections.
  • The Minority Front (MF), a party led by prominent politician Amichand Rajbansi, ex-leader the of the National People’s Party, the dominant party in the House of Delegates election (available to South Africans classified Indian/Asian only) during apartheid.
  • The United Christian Democratic Party (UCDP), led by Lucas Mangope, president of the Bophuthatswana bantustan during apartheid, who had also boycotted the first election (putting it mildly).
  • Federal Alliance, led by South African rugby supremo Louis Luyt
  • Afrikaner Eenheidsbeweging (AEB). Probably the lowest profile party to win a seat, leader Cassie Aucamp was still a relatively prominent Afrikaner nationalist and I believe church leader.


In 2004, the number of represented parties was reduced to 12. One new party, the Independent Democrats, was led by the well-known PAC politician Patricia de Lille, and who had increased her national profile with her work in exposing the arms deal.

Two parties were not returned to parliament, both newcomers in 1999.

  • Federal Alliance – the party ran jointly with the Democratic Alliance, and later merged with the Freedom Front Plus
  • AEB


2009 saw an increase to 13 again. The New National Party folded, while two new parties were elected:

  • COPE were formed from a major split in the ANC after the ousting of President Mbeki
  • The African People’s Convention were formed from a split in the PAC, with the deputy president leaving to form his own party


2014 saw the number of return parties remain static, but some churn amongst the makeup. New additions were:

  • National Freedom Party, formed from a split in the Inkatha Freedom Part, and led by the former IFP chairperson.
  • Economic Freedom Fighters, formed after prominent ex-ANC youth league leader and President Jacob Zuma supporter Julius Malema was expelled from the ANC.
  • Agang, led by high-profile Mamphela Ramphele, a prominent anti-apartheid activist, managing director at the World Bank and former partner of murdered activist Steve Biko. Her dalliance with the Democratic Alliance before the election also increased her visibility.

Out went:

  • UCDP – Mangope was expelled in 2012
  • MF – Rajbansi died in 2011
  • Azapo – probably a victim of the EFF’s success.

The point is, it’s extremely difficult to get elected. Once you’re in, with all the resources and free media coverage, it’s a little harder to leave again, but the death or removal of a prominent leader puts the personality parties in particular at risk.


So, what does this mean for 2019? Putting on my soothsayer hat, the following parties are in danger:

  • Agang is at risk after Mamphela Ramphele left shortly after the 2014 elections. Although new leader Andries Tlouamma has been making the most of his media coverage, and perhaps his twerking will be enough to keep the party in parliament.
  • PAC – again experiencing bitter infighting, as they have before every recent election, will this be one election too far for them?
  • AIC – some thought their surprising result in the 2014 election was due to being placed right next to the ANC on the election ballot. With similar colours and a similar logo, perhaps people were confused while voting? But their results have been steady since, and with their new national profile, they may be on track to at least retain a seat.
  • APC – the party seems to be quite active on the ground. but their aim of a million votes (up from thirty thousand in 2014) is pie-in-the-sky stuff. They rather need to look behind them to hold on to their seat, with Azapo running again, the EFF gaining prominence, and other new entrants such as Black First Land First, the Land Party, all playing in the same zone.

So, who are the possible 2019 newcomers to parliament?

  • Good, the party formed by Patricia de Lille on her expulsion from the Democratic Alliance, are very likely to earn a seat, and will be hoping to do as well as the Independent Democrats in 2004 and 2009.
  • The Socialist Revolutionary Workers Party have prominent trade union backers, and if they can tap into their organisational capacity, they should win representation.
  • The ATM is backed by the South African Council of Messianic Churches in Christ, who have a hefty following. If they can tap into this, they also have a good chance of representation.

Beyond that, I struggle to see any of the newcomers gaining a seat. There’s still some time left, and things could change, but most seem to be relying on wishful thinking that the support of a few people close to them, or in their community, will get them over the line.

And most are going to be horribly disappointed.

Image from Wikimedia Commons

Related posts

Fire (Social)

Wondering who to vote for in May?

It’s very rare that you come across a political platform that jumps out at you in the way this one did. From the moment I first set eyes on it, I knew this was the one.

Meet the Alliance for Transformation for All.

ATA logo

They have two main policies:

1) Subsidies for taxis (we’re talking minibus taxis here of course, not sedan taxis)
2) Less government regulation for taxis. Taxi drivers should regulate themselves.

What more needs to be said. The surefire way to cure all of South Africa’s ills is to give more money to the taxi industry, and at the same time back off on those annoying regulations, and let them decide what to do with it.

I’m sold, I don’t need to look at what the other 47 parties are up to.

OK, so maybe ATA are not for you. I present another surefire winner…

“Prisoners go to the Badlands where they grow their own food, build their own accommodation. Prisoners are chipped and cannot remove the chip or leave the Badlands without it exploding.”

Wait, which party is this? Perhaps the Death Penalty Party is getting soft and now thinks prisoners should rather just have bombs attached to them?

Well, to my dismay, it’s the Green Party of South Africa. If I had to pick a political home, it would be a Green Party, but a European-style Green Party, such as the German Greens, which were born out of the environmental, peace, new left, and new social movements.

Not one that thinks strapping bombs to prisoners is a good idea.

Did I misinterpret things? No, it’s a direct quote from their website. A combination of Mad Max and Survivor (just don’t get voted out, or kaboom), I can only hope this will be televised. It’ll be a great foreign exchange earner.

Except that “Banks and use of money suspended”.

Hmm, there may be some chaos, and the Badlands will be getting a little full, but don’t worry:

“A state of emergency is declared…” To deal with the collapse of the financial system, starving hordes trying to break into the Badlands where at least some food is being grown?

No. “A state of emergency is declared till we have set South Africa onto a path to reduced population”

Except that the country will immediately be on a path to rapid depopulation, since women will be permitted a maximum of two children, below the replacement rate, because “Women giving birth to second child have tubes tied”. Forcible sterlisation of women, not men. Maybe a kind of reverse Handmaid’s Tale?

GPSA logo

Continuing the weird sexist theme: “Women volunteer to be trained for running local clinics in skin, nutrition, common ailments to adjust behaviours that are destructive to health and pick up medical problems before they become too bad and direct them to professional care.”

In case you’re worried that suspending money overnight will lead to some disruptions in the food supply, and the country may need to rely on food aid from the rest of the world, fear not. I can assure you, no food aid will be arriving, because, “within a month, no fossil fuel powered ships may dock in South African ports”

So there may be a shortage of things like computers arriving too, which won’t be a problem, because we’ll be experiencing a cultural revolution II, with “De-urbanisation for anyone prepared to work with renewing the biomass of the soil, or plant and grow organic food, reforest or care for animals will hopefully motivate many to disperse over the land.”

It’s a little unclear what that means – there no mention of coercion, but what exactly will attract everyone to a life of toil on the fields isn’t specified.

But wait, maybe we need those computers after all, because “Schooling will be over the air so the necessity to be near the city for education of kids will be reduced/ This will mean that families prepared to work on the land or on any of the above, can move out into the country.”

If working on the fields doesn’t appeal, there are other options. You can register as a plumber:

“People with aptitude register to be trained as plumbers and plumbing parts manufacturers. The whole country is to be replumbed for circular systems of use naturally treated in wetlands. This means that the same water will be used over and over again for toilets, paper and solids will be separated locally and composted. No more pit latrines, chemical toilets sewerage in aquifers, rivers or sea and plenty of food for microbes.”

And also no composting toilets and biomass incinerators?

Plumber doesn’t appeal? How about drug manufacturer?

“Because drug abuse shortens life span, people who wish to take drugs and are prepared to watch a film showing what eventually happens to drug takers get a licence for free drugs.”

Good to know that the motivation behind free drugs is not primarily reduction of harm to society, or care for the addicts, but because it’ll speed them on their way off this plane leaving more space on the fields for the rest of us to toil on.

I think I’ll choose the free drugs.

(On the off chance that the party doesn’t sweep to power and their website doesn’t stand the test of time, here’s a screenshot of the page as of 19 April 2019)

Related posts

Images from Wikipedia:

Earth (Arts and Literature)

March 2019 African language Wikipedia and Wiktionary update

It’s been quite a long time (four and a half years in fact) since I looked at the state of the African language Wiktionaries. For those new to Wiktionary, the idea is that it will describe all words of all languages using definitions and descriptions in the particular language edition. An ambitious task!

So, how are the projects progressing?

A portion of the Octateuch in Ethiopian

African Language Wiktionaries

Language 30/5/2010 15/5/2011 29/10/2014 22/3/2019 % +
Malagasy 4,253 3,599,084 5,482,632 52.33%
Afrikaans 14,669 14,731 15,794 20,831 31.89%
Swahili 13,000 13,027 13,903 14,029 0.91%
Wolof 2,689 2,693 2,310 2,312 0.09%
Somali 1,635
Sotho 1,389 1,398 1,343 1,343 0.00%
Lingala 673
Zulu 131 510 586 599 2.22%
Igbo (incubator) 375
Kinyarwanda 306 306 367 366
Tsonga 359 363 92 359 290.22%
Oromo 218 264 322 335 4.04%
Swati 371 377 290 292 0.69%
Amharic 319 377 206 217 5.34%
Egyptian Arabic (incubator) 195

In short, although it’s been so long since the last update, there’s not much to show. The only project to more than double its articles in four and a half years is Tsonga, off a minute base. Malagasy has always had a huge amount of bot activity, and is still growing from a large base, and Afrikaans shows some signs of life. But overall, the state of the African language Wiktionaries can be described as dormant.

Perhaps the African language Wikipedias will fare better?

African Language Wikipedias > 1000 articles

Language 26/6/2015 5/9/2017 30/6/2018 2/4/2019 % +
Malagasy 79,329 84,634 84,996 91,528 7.68%
Afrikaans 35,856 46,824 50,275 76,965 53.11%
Swahili 29,127 37,443 42,773 49,555 15.86%
Yoruba 31,068 31,577 31,672 31,867 0.62%
Egyptian Arabic 14,192 17,138 18,605 20,405 9.67%
Amharic 12,950 13,789 14,286 14,558 1.90%
Northern Sotho 1,000 7,823 8,050 8,018 -0.40%
Somali 3,446 4,727 4,898 5,456 11.39%
Shona 2,321 2,851 3,630 4,278 17.85%
Hausa 1,345 1,525 1,856 3,494 88.25%
Lingala 2,062 2,915 3,023 3,113 2.98%
Kabyle 2,296 2,887 2,844 2,986 4.99%
Kinyarwanda 1,780 1,810 1,823 1,821 -0.11%
Kikuyu 1,349 1,357 1,358 0.07%
Igbo 1,019 1,384 1,320 1,392 5.45%
Kongo 1,176 1,179 1,193 1.19%
Wolof 1,023 1,157 1,166 1,184 1.54%
Luganda 1,153 1,162 1,169 0.60%
Zulu 683 942 959 1,067 11.26%
Language 26/6/2015 5/9/2017 30/6/2018 2/4/2019 % +

The Zulu Wikipedia is the latest addition to the 1000 club, having reached this milestone just before Wikimania last year, and progress has been steady since then.

At first glance, Hausa looks like it’s in great shape, with an 88% increase in the number of articles. But this is misleading, as many of these are one line articles on football players, the entirety of which translates as, for example, “Kenny Allen (footballer) is an English football player.” No disrespect to Kenny Allen, but I’m not sure he and the 100s of other footballers listed there are critical components of Hausa knowledge. There’s a move to delete these articles (you can see the impressive list here while it’s up), but even if they survive, it’s not a sign of a healthy project.

Leaving aside Hausa, it’s once again Afrikaans, growing at an impressive 53% over the period, that provides an example for the rest. At current rates, it’s on track to pass Malagasy and reclaim its position on top in about a year or so.

Besides Afrikaans, only Shona, Swahili, Somali and Zulu show a growth rate above 10%, while quite a few sit idle.

Moving on to the South African language editions specifically:

South African Language Wikipedias

Language 26/6/2015 5/9/2017 30/6/2018 2/4/2019 % +
Afrikaans 35,856 46,824 50,275 76,965 53.11%
Northern Sotho 1,000 7,823 8,050 8,018 -0.40%
Zulu 683 942 959 1,067 11.26%
Xhosa 356 708 738 789 6.91%
Tswana 503 639 641 641 0.00%
Tsonga 266 526 562 585 4.09%
Sotho 223 523 539 546 1.30%
Swati 410 432 439 467 6.38%
Venda 151 256 256 265 3.52%
Ndebele (incubator) 12 12 11 -8.33%
Language 26/6/2015 5/9/2017 30/6/2018 2/4/2019 % +

Afrikaans remains the only project that could be described as a usable Wikipedia – the other languages are still very much in the formative stages. Zulu is also showing signs of life. Besides these two, only Xhosa and Swati see growth rates above 5%. It’s sad to see the stalling of Northern Sotho, while Ndebele shows no signs of getting out of the incubator anytime soon.

2019 has been proclaimed the Year of Indigenous Languages by the UN, but so far there’s not much sign of a change in the status of the African language projects. Later today sees the South African Centre for Digital Language Resources, in collaboration with the Academy of African Languages and Science from the University of South Africa, present an interactive day workshop on contributing to Wikipedia in South African languages. It’s great to see this initiative, which arose with no help that I’m aware of from Wikimedia South Africa. I’m always hopeful with events like these. Generally very few people to stay around to edit Wikipedia, but as projects like Northern Sotho and Swahili show, one person can make a huge difference in the early stages, and it justs needs a committed editor to stick around. It’s a lonely job editing in the early stages, wondering if it’s worthwhile, no community, no idea if their work is being read. Hopefully someone will take on the challenge!

If you are looking to contribute, but don’t know where to start, please reach out to Wikimedia South Africa and we’d be happy to assist.

Related posts

Image from Wikimedia Commons

Water (Personal)

A Seasoned Travellers Guide to Jet Lag

As a seasoned traveller I’ve mastered the art of avoiding jet lag, and will share it in this blog post. The solution is simple – no need for sleeping pills. I will try and catch some sleep on the plane if it’s night time at my destination, but usually this is sporadic, and I arrive tired, so this isn’t necessary.

The real secret is just to push on until night time in the new time zone, no matter how tired I’m feeling, and, exhausted after the long day, get a good night’s sleep, waking up refreshed in the morning of the new time zone.

Travelling east is usually more difficult than travelling west, and so it was I began the long journey home from New York to Cape Town, via Brussels and Doha. Leaving the hotel in New York in the afternoon, an hour on the subway, the JFK Air Train, a meal at the airport, before catching the 7.30pm flight (0.30am Brussels time, 1.30am Cape Town time). I had an aisle seat, but usually don’t manage to sleep unless I have a window seat and free space next to me. I tried to nap on and off during the flight, rather unsuccessfully.

The plane arrived in Brussels 9am (10am Cape Town time), and I was already pretty tired. I had a long wait at the airport before the 3.40pm flight (4.40pm Cape Town) to Doha, barely functional, nodding off a few times in the airport.

A late flight should be the perfect time to sleep. Except it arrived at 11.55pm Doha time (10.55pm Cape Town), and the connecting flight was 1.45am (0.45am Cape Town), boarding at 0.45am (11.45pm Cape Town). With only 50 minutes between arrival and boarding, and the need at Doha airport to scan one’s luggage in-between, the changeover gets the adrenaline pumping, not ideal for sleeping the rest of the way.

Finally I’m boarded – exhausted, ready to sleep for most of the final leg – a 10 hour 20 minute flight to Cape Town.

Except, besides the Doha-airport induced adrenaline, I’m in the middle seat, squashed between two men, legs cramped up against the seat, and little prospect of sleep. A movie it will have to be, but since it’s the 8th flight in a month, I’m all airplane movied out.

I land in Cape Town 11h05, and get back to my house, about 37 hours after leaving New York. Right, all I need to do is push on till 20h00 or so and get a good 12 hours sleep…

No chance, I fall asleep at 2pm, waking up 9pm. Waking up refreshed at 9pm is not the recommended way to avoid jet lag. From then on, my body didn’t know what was happening.

Day 2: Sleep 2.30pm to 5.30pm. I needed to be somewhere by 6, so used the dreaded alarm.

Day 3: Sleep 3.15am to 2.30pm. A long, much-needed sleep, and it looks like the times are improving, right?

Day 4: Sleep 7.45am to sometime late afternoon. Considering drugs right now.

Day 5: Sleep 3am to 10am. Woohoo! Back to normal. 5 days of hell, but all worth it.

Day 6: Sleep 11pm to 2.30am. Awake till 7am, sleep till 2.30pm. Hmm, maybe not…

Day 7: Sleep 8.40am till 3.40pm. Aaargh!! I have a party in the evening, but don’t feel up to it, so head to bed early.

Day 8: And fall asleep 1.30am, waking up 6.30am. Have I cracked it yet?

Day 9: Sleep 11.30pm to 6am.

Six and a half hours is not usually enough, and I was nodding off in the day. As I write this it’s 11.35pm, and I feel alarmingly awake. Time to finish this off and find a boring book…

Related posts:

Images from Wikimedia Commons 1 2

Fire (Social) Wood (Spiritual)

The Emerald Isle?

At its peak, following the last ice age, Ireland was covered in forest. The first humans, initially hunter gatherers, arrived around 9000 years ago. About 6000 years ago the forests began to disappear as farming began to take hold. By 1600, forest cover was still substantial, but as population pressure grew, and in particular clearance for sheep grazing, by 1800 less than 1% remained. Even by European standards of destruction, this was an impressive feat.

Sheep farming in Ireland

Today, Ireland has around 10.5% forest cover, still placing it at the bottom in the list of European countries. The average in Europe is about 30% – Finland leads with 73%, and even heavily developed Germany has 32%. Ireland trails even the Netherlands, with a population density of 416 people per square kilometre, compared to 69 for Ireland.

Ireland is nearly unique among developed countries in that its population is far below its peak. In 1841, the current area of the Republic of Ireland had a population of 8.51 million. The Great Famine initiated nearly a century of population decline, as both starvation and emigration took their toll, and by 1926 the population had been reduced to less than 3 million.

The causes of the Great Famine are complicated. The best land was reserved to meat production for primarily British consumers, and when a potato blight struck, destroying around half of the potato crop, exports of potatoes to Britain continued apace as impoverished primarily Catholic Irish could not pay the required prices. Ireland always produced enough food to feed its people, but most of it was exported to Britain, similar to what happened with other famines under British rule, such as in India. Some Irish saw and still see this as a form of genocide.

But back to the forests.

Ireland is now afforesting quite rapidly, and aims to get to 18% cover by 2046. The government has policies in place to support private landowners planting forest, and now almost 45% of Irish forest is privately owned. Unfortunately, the vast majority of this is non-native, mostly Sitka Spruce, planted mainly for its potential use for timber and paper.

The non-native Sitka Spruce

Wandering around Ireland, the Sitka Spruce ‘forests’ feel dead. Almost no birdlife, and an acidic soil that prevents anything else growing under them. The largest patches of natural forest grow in Killarney National Park, which by contrast is lush and alive, supporting a wide range of flora and fauna. However, the entire reserve is a tiny 102 square kilometres.

Wild forests in Killarney National Park

It was Ireland’s first national park, created in 1932, and until 1984, the only national park in the country. There are now 6 in total, protecting just under 1% of the country. Ireland is certainly lagging in this regard,

But it’s intriguing to me too see how much sway meat production has in the country. I have visited three national parks while here. Killarney, which is heavily wooded, as well as the Burren and Connemara. The latter two have almost no tree cover. What’s more, they don’t look like getting any soon. In the Burren, the reserve seems to find it important to preserve the concept of winter grazing, and is concerned about animal farmers moving to less taxing means of production, and no longer grazing the Burren.

The Burren is unique – beautifully stark and rocky, with numerous rare species. The famous quote “There isn’t tree to hang a man, water to drown a man nor soil to bury a man” purportedly comes from one of Oliver Cromwell’s general during his campaigns in Ireland in the 1600s. And yet once it too was covered in forest.

The Burren

Connemara National Park is similar. Sheep and goats graze right to the top of the highest peak. I can understand that it’s been thousands of years since trees grew in these areas,and so conservation has a different meaning that in an area much more recently touched by humans, but I still find it strange that, with such a demand for afforestation, and government subsidies, there seems to be so little attempt to regrow the ancient wild forests that would once have covered the Emerald Isle.

Sheep grazing just below the highest point in Connemara National Park

Images from Wikimedia Commons (1 2 3 4 5 6)

Related posts:

Metal (Technical)

Simple hosting?


All I want is a simple website!

OK, I have a website, and you’re probably reading this on it.

But I want a simple website. Since 2006, I’ve had a server hosted with Asergo (previously EasySpeedy). It ran a lot at one stage. Huge mounds of organic food found new homes through its circuits. Today it’s not so busy, but stands ready, waiting to handle any imminent Slashdot Effect when my latest masterpiece goes viral. Except it never does, and I’m basically paying for a private jet when I could make do with a rusty bicycle.

In other words, a bog standard WordPress hosting offering will do just fine these days.

Moving from total control of my own server to a tiny hosted offering seems limiting, and its kept me from moving for a while. But gradually I’ve come to make peace with it, and will make do with a tiny instance offering limited functionality.

So what am I looking for?

* a reasonably new version of PHP
* a reasonably new version of MariaDB, or similar
* perhaps even hosted locally? Other sites I help with that host locally have made me wary, but there must be some good ones out there?

Space (my current server feels like it’s got backups going back to the dial-up days, so this will be a good chance to clean up), bandwidth etc. are all much of a muchness, as most offerings give some flexibility here

Let’s look at some of the options.

As a starting point, seeing as another site I manage is hosted there, 1-grid. They’ve gone through some wobbles. At first part of Web Africa, the hosting division got spun off into Gridhost, which then got renamed 1-grid.

* R89 a month
* Unlimited MySQL databases
* PHP 5+. Really? What does this mean? PHP 5.6 and PHP 7.0 reach EOL in about a month… I happen to know PHP 7.1 is available, but sites really need to publicise this stuff!
* No mention of MySQL version. The instance I manage has the venerable but still supported MySQL 5.5, with no mention of possible upgrade.
* But, the dealbreaker. A measly 5GB storage. The next tier, offering 50GB, is R219 a month.
* Another big dealbreaker – they charge an extra R519 for SSL, for one domain only!

* Hetzner were one of the first locally to move to MariaDB, a major plus in my books (yes, I know I work for the MariaDB Foundation so perhaps I’m slightly biased)
* Another plus is they offer Lets Encrypt SSL with all hosting packages
* R99 for 5GB, 10GB costs for R149 a month
* However, they only mention MariaDB 10.0, which, while a version above MySQL/MariaDB 5.5, is still quite old. In discussion with them, they do offer MariaDB 10.1, but again their website lags with no mention of this.
* PHP 5.6 and PHP 7.0 only.
* Only offer Debian 8. Debian stable is itself already pretty venerable when it comes out, and Debian 8 came out way back 2015. I don’t need cutting edge, but Debian 9 came out in Jun 2017, so this makes Hetzner’s offering a little on the old side.

Ok, moving internationally. I moved my domains to Gandi a few years ago. They offer a Simple Hosting service as well.

* Unusually, they offer Percona Server as a MySQL equivalent. Based on MySQL 5.7, which came out at about the same time as MariaDB 10.1.
* Support PHP 7.2
* 20GB for ~R45 a month (a discount as I have domains with them)
* In price, space and up-to-date software they absolutely blow the local offerings out the water
* They also offer a free 10-day trial.

Their Simple Hosting seems to be just that, really simple, not offering as much as the full CPanel or KonsoleH services offered by 1-grid and Hetzner but hey, I did start by saying I wanted a Simple website.

I spend way too long investigating the governments of Luxembourg and France to decide which of the two server locations to choose, but in the end I’m sold (I chose Luxembourg for those interested).

Except I’m unable to sign up, as their interface isn’t working (I tried on three browsers) and I cannot actually sign up. I contacted them a few months ago, pointing out inconsistencies in their documentation (the site mentioned MySQL/Percona 5.5, 5.6 and 5.7 being the latest available version in various different places). The site still lists MySQL 5.6 as the latest release in at least one location.

So not exactly getting a good feeling from them either.

Am I destined to keep paying for my jumbo jet hosting, as it least I know it works?

Any personally recommended hosting suggestions?

Related posts:

Image from Wikimedia Commons