Oracle Linux ships MariaDB

I can’t remember why I was installing Oracle Enterprise Linux 7 on Oracle VirtualBox a while back, but I did notice something interesting. It ships, just like CentOS 7, MariaDB Server 5.5. Presumably, this means that MariaDB is now supported by Oracle, too ;-) [jokes aside, It’s likely because OEL7 is meant to be 100% compatible to RHEL7]

OEL7__Running_

The only reason I mention this now is Vadim Tkachenko, probably got his most retweeted tweet recently, stating just that. If you want to upgrade to MariaDB 10, don’t forget that the repository download tool provides CentOS 7 binaries, which should “just work”.

If you want to switch to MySQL, there is a Public Yum repository that MySQL provides (and also don’t forget to check the Extras directory of the full installation – from OEL7 docs sub-titled: MySQL Community and MariaDB Packages). Be sure to read the MySQL docs about using the Yum repository. I also just noticed that the docs now have information on replacing a third-party distribution of MySQL using the MySQL yum repository.

Trip report: LinuxCon North America, CentOS Dojo Paris, WebExpo Prague

I had quite a good time at LinuxCon North America/CloudOpen North America 2014, alongside my colleague Max Mether – between us, we gave a total of five talks. I noticed that this year there was a database heavy track — Morgan Tocker from Oracle’s MySQL Team had a few talks as did Martin MC Brown from Continuent. 

The interest in MariaDB stems from the fact that people are starting to just see it appear in CentOS 7, and its just everywhere (you can even get it from the latest Ubuntu LTS). This makes for giving interesting talks, since many are shipping MariaDB 5.5 as the default choice, but that’s something we released over 2 years ago; clearly there are many interesting new bits in MariaDB 10.0 that need attention!

Chicago is a fun place to be — the speaker gift was an architectural tour of Chicago by boat, probably one of the most useful gifts I’ve ever received (yes, I took plenty of photos!). The Linux Foundation team organised the event wonderfully as always, and I reckon the way the keynotes were setup with the booths in the same room was a clear winner — pity we didn’t have a booth there this year. 

Shortly afterwards, I headed to Paris for the CentOS Dojo. The room was full (some 50 attendees?), whom were mainly using CentOS and its clear that CentOS 7 comes with MariaDB so this was a talk to get people up to speed with what’s different with MySQL 5.5, what’s missing from MySQL 5.6, and when to look at MariaDB 10. We want to build CentOS 7 packages for the MariaDB repository (10.0 is already available with MariaDB 10.0.14), so watch MDEV-6433 in the meantime for the latest 5.5 builds.

Then there was WebExpo Prague, with over 1,400 attendees, held in various theatres around Prague. Lots of people here also using MariaDB, some rather interesting conversations on having a redis front-end, how we power many sites, etc. Its clear that there is a need for a meetup group here, there’s plenty of usage.

T-Mobile USA PAYG gets worse…

Back in August, I wrote about how T-Mobile USA changes their pay-as-you-go plans. It is even worse than it seems.

I’m back again this month, and find that if I want to change to the $3/day unlimited calls/text/data plan, I have to wait till the next month (i.e. the next billing cycle). Previously this would just work immediately, as long as I changed it online. 

My options to get data now? Pay $10 for a 1-week High Speed Data Pass with up to 1GB of traffic to utilise. There’s also a $5 day pass. 

T-Mobile reminds me of my experience with DiGi. Both were underdogs. Once they get more market-share (in T-Mobile’s case, maybe a better network/selling the iPhone; in DiGi’s case finally getting a 3G license back in the day), they just can’t handle the load, and graduate from their underdog status to being worse than the incumbents. I’m still a T-Mobile & DiGi customer, but the question is for how long more?

Time wasted on planes when you can’t use devices

I tend to travel on planes quite often, and I’ve always been annoyed by the fact that you have to turn off your electronic devices during take-off and landing. I thought life would be better when you could keep your tablets on in flight mode (thus not requiring me to carry physical newspapers on-board or even magazines). 

F1 airshowIn the USA and in the UK, this is no longer a problem (it hasn’t been at least since the end of last year), I was always hoping that the Asian carriers I tend to fly (Singapore Airlines & Cathay Pacific in particular) would modernise.

Sometimes I fly to Singapore from KL, and that’s about a 45-minute flight. One day I measured how much time I was wasting without using electronic devices:

  • 26 minutes 45 seconds upon takeoff – announcement, taxi, takeoff (KUL-SIN)
  • 14 minutes upon landing – here during taxi, you can use electronic devices 

41 minutes wasted, is almost as long as the flight ;-)

So it is with great pleasure that since the middle of the year (July 2014), Singapore Airlines allows you to use your devices (not laptops, but thats ok – tablets/phones are good enough for now) in flight-safe mode, and Cathay Pacific just allowed this as of mid-September 2014.

Now, when will Malaysia Airlines, and AirAsia wake up? Presumably this has a lot to do with the Malaysian Department of Civil Aviation.

Taxi safe return home service in Seoul

Apparently taxis can be dangerous for ladies at night in Seoul. So there is a program called the Safe Return Home Service now in Seoul.

How does it work? The assumption is that many people have NFC enabled phones. They just need to tap their phone on this NFC-enabled pad, and automatically a text message is sent to one’s guardian giving the taxi’s model, current time & location, and probably more.

IMG_3528.JPG

This ensures that you now travel with a better sense of security. Read the press release, which suggests this is a program from SK Planet (I’ve had the pleasure of working with this group in the past).

I think something like this can be really useful in markets like Malaysia, where the taxi system requires a serious overhaul.

What eventually happens with apps

There has been a lot of talk about apps generally not doing well. There may be truth in the matter.

Let’s take my mother, a classic example of someone who plays Candy Crush (made by King, whom are now public listed in London). She plays this mainly on her iPad, but also on her iPhone.

Life began quite simply by just playing the game. She got up to nearly level 500, and she only made 3 in-app purchases. Countless hours of entertainment, for what amounts to 3×0.99 cents. You don’t get much for $3 these days.

Then suddenly with the help of automatic app updates (and the fact that the Average Revenue Per User (ARPU) needs to increase for earnings), she was required to login with Facebook finally.

She did, and started from scratch!!! Nearly 500 up there, and starting from scratch. You can’t imagine she’s pleased but she enjoys playing the game.

Fast forward to today, and the Facebook app logged her out. Candy Crush became aware & asked for a re-login. Lo and behold, the 70-odd levels she completed were gone and she had to start from scratch.

Her total investment in time? 8 months. Her total unrecoverable investment in in-app purchases? Less than $3. Her frustration levels? Like she wanted to throw the iPad at the wall!

What can we learn from this? Apps provide countless hours of entertainment for very little revenue to the app creator. App updates that break the database, eventually annoy the user. Is the user likely to continue with other games or apps? Possibly. But after a while there is app fatigue.

So it’s not about discovery. Sure the lists help. But being social (ie in-person) aids discovery too.

Being consistent, is key. Who downloads an offline travel guide, that gets updated and needs a resync, when you happen to be offline? I know a few offenders.

Splitting up apps that should be one – Facebook/Messenger, Foursquare/Swarm, etc. Then not providing a consistent interface, removing features or worse crashing when you’ve got to switch to the next app.

App fatigue is caused by putting the company or investor first, and the user last.

And as more contribute to the subpar user experience, the more smartphones will be whittled down to providing their basic functions provided for by Android & iOS with a sparse few extras. Overall, that makes the barrier to success much higher than before.


i