I recently upgraded from Hardy Heron (Ubuntu 8.04) to Intrepid Ibex (8.10) but I was certain that the old pppd configuration I had for just this situation must have survived the upgrade. It was painstakingly put together from searches conducted over GPRS with Opera Mini on the self-same phone I wanted to connect with in a situation not unlike the one I found myself in – removed from all semblances of civilization (i.e. cabled broadband).

Just for fun I decided to check if there were any significant changes to nm-applet (or NetworkManagement Applet). There hadn’t really been any changes to it worth writing home about in the last 3 versions of Ubuntu (at least) but for some reason a force, perhaps even the Force (the one without midi-chlorients), reached out from within the Ether(net) and compelled me to right click on the applet icon and click “Edit Connections…”

And lo the clouds parted, choirs chanted, and a beam of white light shone upon the lowly notebook (specs to come later). Or the light shone upon the roof that protected the notebook from the wonderful summer rain falling outside at least.

For the Network Connections dialog had indeed evolved, nay mutated. Here we gazed upon the new “Mobile Broadband” tab (though this connection is anything but broad) and all was at once well with the ‘Tubes. For from within this tab one could click ‘Add’ and a pretty clickety-click Wizard adorned the desktop. And with the simple selection of country and service provider, alteration of the arbitrary contents of a descriptive text box, and the click of an ‘Apply’ button a connection capable of making use of a GSM modem was created.

But surely it couldn’t be that simple? Where do you tell it which modem device to use? Upon connecting the phone to the notebook all was revealed… nm-applet now automatically discovered modem devices. Connecting was as simple as clicking the correct radio button in the left-click drop-down menu of nm-applet. I could cry.

A part of me is sad that another aspect of Linux’s divine complexity has gone the same route as the ill-supported wireless networking drivers. Whatever will the geeks of tomorrow do to get their kicks from computing? Imagine your children never having to ever compile their own kernel? My blood runs cold. And yet I’m joyed that the illustrious Mark Shuttleworth (or his team at Canonical and the brave me and women of the Ubuntu community to be exact) has brought Linux yet another step closer to the mass market.

Screenshots and more to follow upon my return to the arms of DSL, my mistress. Apologies for the current lack of bling but this post is being written from an old laptop running Ubuntu, connected to the Internet via a Sony Ericsson W850i which is paying premium rates for bandwidth.

While trying to upgrade from WordPress 2.1 to 2.6 I was presented with the database upgrade page when trying to get to the backend (/wp-admin). Upon clicking the Upgrade button/link a half-ream long list of “Table doesn’t exist” errors were displayed.

WordPress failed database upgrade - Table doesnt exist error messages.

The reason WordPress can’t find the tables is because it fails to create them during the database upgrade. For the full reason behind why this happens you’ll have to read the whole post. If you’re only interested in what to do to get it working the next three paragraphs and two code samples should explain all.

In WordPress 2.2 two new settings were introduced in the wp-config.php file namely DB_CHARSET and DB_COLLATE. By default, that is in the wp-config-sample.php file, it sets these variables to ‘utf8′ and ‘utf8-general’ respectively.

If you added the DB_COLLATE option to your old wp-config.php file and retained the default option as given in the sample file then it is highly likely you experienced this error since utf8-general-ci is not a valid MySQL collation. The silly little error that causes so much grief lies in the dashes — they should be underscores.

Most MySQL installations use utf8_unicode_ci as the default collation. You’re likely to get joy by setting the DB_COLLATE option to

I had an old WordPress 2.1 installation that needed to be upgraded. Besides the obvious security issues that have been fixed and the features that have been added, I wanted to use a theme that required at least WP2.5 to work.

I don’t do anything on any of my live websites without first ensuring that I won’t break anything by trying the changes on the copies I keep locally on my development machine. Upon attempting the famous 3-step upgrade process I found that not only does WordPress struggle to create a bunch of new tables for some reason, it proceeds to delete the old WP2.1 categories, link2cat, and post2cat tables even though the CREATE TABLE queries for the replacement tables, terms, term_taxonomies, and term_relationships fail.

The only relevant thread I could find at WordPress’ support forums suggested incrementally upgrading and from the responses it suggested that it worked for the original poster. But why were the tables not being created? What do the other upgrade scripts do differently from the 2.6.2 one?

After some hunting I found the CREATE TABLE queries that are supposed to create the new tables during an upgrade to the file schema.php in wp-admin/includes/.

Copying and pasting the query verbatim into the MySQL Query Browser and replacing the variables with the values I suspected them to be (determined mainly from wp-config.php) indicated that the query was indeed failing. After ensuring that no amount of quoting or backticking would fix the query, the DEFAULT CHARACTER SET and COLLATE directives at the end of the CREATE TABLE query caught my interest. Could the use of dashes rather than underscores really be breaking the query?

A quick test by removing the directives entirely, forcing MySQL to use its defaults, verified it. The defaults WordPress supplied in its sample configuration file (wp-config-sample.php) were in fact not valid.

The WordPress documentation does mention that people who are upgrading should not include the DB_CHARSET and DB_COLLATE variables in their wp-config.php files unless they know what they’re doing… But that’s only on the ‘Editing wp-config.php’ page! This should be made very clear in the Upgrading WordPress documentation.  For someone that doesn’t know the format MySQL expects arguments to COLLATE this can translate to days of struggle… Bad WordPress! No treat for you!

By now is old news that Morgan Tsvangirai, leader of the Movement for Democratic Change, has withdrawn from the final run-off elections in Zimbabwe which would determine the next president of the country.

This comes after at least 70 MDC supporters have been killed. Robert Mugabe‘s answer to the “imperialist West” to the allegations that opposition supporters are being killed by mandate of the government is that the media is once again misrepresenting what is happening in Zimbabwe to justify their interference in the country.

Steve Pierson at The Standard, a political blog, says:

Military action seems to be the only option left. A smart campaign could be conducted relatively bloodlessly but it would have to be led by South Africa, the only country in the region with the standing and the military might to carry out such an action. And South Africa is showing no will to act. [http://www.thestandard.org.nz/?p=2298]

In some respects I agree with Steve. It is obvious that non-violence seems to be a tactic only the opposition are interested in. But I disagree that any military campaign conducted by South Africa against Zimbabwe will be “relatively bloodless.” I also agree that New Zealand should offer its support – but that South Africa should decline. At least initially.

I have it on relatively good authority that should a military incursion from South Africa to assist in a coup d’état degenerate into an all-out bush war (which is likely) then our army will be given a thorough shelacking. We might have the technological upper-hand and superior numbers, but our military has deteriorated significantly over the last 10 years. I would venture a guess that this is due to more pressing concerns at home than keeping a well-oiled fighting machine in reserve.

While I don’t profess to be a military genius, I do not logically see the technological advantage we have counting for much if we’re

1. Fighting in the bush; and
2. Trying to preserve innocent life.

The Zimbabwean army should also not be underestimated. I don’t know about the state of repair of the Zimbabwean air force or armoured divisions but I know that ours leaves much to be desired, mainly due to funding reasons.

Our special forces known as the Recces could be deployed to covertly assist rather than risking open warfare, but our Special Forces Regiments also suffered due to the shift in priorities that came with the changes in 1994. For instance, as part of the “military transformation process,” 1 Special Forces Regiment was disbanded in 1996.

Today I’m not coming up with solutions. The prospects in this situation look bleak. Either people continue to die or live in squalor or we send in our troops to and they die to try and prevent the Zanu-PF from killing more innocents. Today I am saying that whatever the ANC decides to do about this they have to weigh the cost in human life carefully – hardly an easy task and one that will most assuredly be met with criticism no matter which path they take:

• Try to convince Mugabe to buy into a “government of national unity” (maintain status quo). Should this fail it will appear as if the ANC have done nothing but it will likely cost fewer lives. Mugabe is unlikely to give up his seat of power, however.
• Get the AU or SADC to approve the deployment of a peace force. Should we then lose to the Zanu-PF in a bush war at least it would look like the ANC tried but failed. No matter how you slice it, a military campaign will cost more lives. The chances  of the SANDF winning a bush war against the Zanu-PF (ZNA) is unknown.

Regarding the support from New Zealand – it would be great to know that we’re not in this alone, but in the end I don’t think an African nation could accept help to depose another African government from the ‘imperial West.’ This includes funding, arms, and personnel, I think. It’s sick that politics doesn’t seem to have anything to do with making mature, well calculated decisions but with street cred.

So for the initial campaign South Africa should decline help, but once the rebuilding of Zimbabwe is underway I reckon all aid will be welcome again.

Thus ends my reply to Steve Pierson’s post. Below is just a little bit more added for flavour.

In an article on News24 it is reported that Henri Boshoff, of the Institute for Security Studies said that a UN peace mission requires that a resolution to the effect being adopted and this could only happen if the host nation requests it or there is evidence of gross human rights violations. In the latter case (which most everyone is willing to agree on) the UN will still use a regional organisation like SADC or the AU for peace enforcement. According to the article it is very unlikely SADC or the AU will condone the deployment of troops in the region because of their individual policies on interference.

Karin Alexander of the Institute for Democracy in South Africa who is currently in Zimbabwe went so far as to say that deployment a peace force was an unlikely solution because Mugabe’s possible response created risks.

This basically means that a military campaign into Zim is out of the question for the ANC for the time being.

The Minister of Provinial and Local Government, Sydney Mufamadi and Thabo Mbeki’s legal advisor, Mojanku Gumbi are in Harare at the moment to discuss the “goverment of national unity” with Mugabe. I guess we’ll have to see what results our diplomatic efforts and the elections on Friday yield before grabbing our guns.

History has a funny way of repeating itself, though, and Robert Mugabe has shown very little inclination to relinquish his presidency. I guess the question is not if diplomacy fails and Mugabe is re-elected but when? And what will the ‘imperialist West’ do if South Africa, SADC, and the AU disappoints again? Will they risk the repercussions of interfering in African affairs?