Almost there…one more day

One more day to go and it’ll soon be the weekend – time to get things sorted out at home. Looking back it is funny how I haven’t had soft drink for the week but what I also thought about is how much money has been saved by not buying soft drinking. When it comes to my much needed caffeine fix I’ve been able to ween myself off coffee with milk and I’ve been drinking coffee black (no milk) – a great way to get that caffeine hit at work without the price tag. Could I give it up? Sure, but it would hell on earth and little real benefit giving up a caffeine based beverage for zero health benefits other than the moral high ground of “I’m an all natural person who doesn’t need stimulants because I’m high on life”.

There are two big events coming up in the next month or so with the first being the ‘World Mobile Congress’ where it is rumoured that Samsung will launch the Galaxy S10 with three models that closely mirror the Apple product range of iPhone XR, iPhone XS, iPhone XS Max. It’ll be interesting to see whether there are other products released – I wouldn’t be surprised if we end up seeing a refresh of their wearable line up but I doubt we’ll see Samsung embrace wearOS especially when you also consider that Samsung uses Tizen in their televisions instead of Android TV. It is interesting to see the growing relationship with Apple with the inclusion of iTunes Store with the range of Tizen televisions which makes me wonder whether this is part of a much more long term goal of moving away from Google in terms of a reliance on on their ecosystem resulting in what Samsung wanting and what Google wanting not always lining up.

Towards the end of March there is a rumoured event regarding a subscription news service along with a greater focus on cloud services so it’ll be interesting to see what Apple deliver. What will also make for an interesting discussion is the work being done on 10.14.4 particularly with the work being done with Webkit and the regular Technology Preview’s being pushed out which has done a lot of work in the area of HTML5 compliance especially in the area of web applications given how there is a general trend in that direction with Spotify being the most obvious example but also Twitter doing the same thing when they retired their Twitter desktop based application in favour of a PWA (Progressive Web Application). By getting Webkit to support those technologies it will hopefully translate to vendors at least allowing users to install PWAs on their computer without having to haul along the Chromium baggage. Hopefully the added support for those technologies will allow end users to benefits whilst taking advantage of the inherit efficiency that comes from Webkit when compared to Chromium.

With that being said, it’ll make for interesting discussion given that 10.14.4 comes with added biometric support from within the browser which makes me wonder whether Apple has taken time to backporting those features back to the current stand version of Safari or whether they just said “bugger it” and branched and stabilised a technology preview then crowned it with a stable version number. There is also iOS 12.2 in the works so I’m looking forward to see what that also delivers.

Demystifying the Samsung Android release schedule

I don’t normally make it a habit to defend Samsung and Android but I thought I might as well put something out there in regards to the release of Android 9 on Samsung phones and the 6 month delay between it being ‘officially’ released by Google and it being made available to end users of Samsung phones. The first thing that is important to understand is that Android is like a ‘distribution’ like how one can see a GNU/Linux distribution being based on multiple projects that are all bought together and then tweaked/customised by the distributor itself to give it a unique look, feel and add features that end users might find useful.

In the case of Samsung though they’ve not only got to bring Android over and test it on their Exynos SoC but also fix up Android bugs that are found as their customisation is integrated back into it resulting an otherwise simple process into something more complex. Yes, yes, I know, in a perfect world we wouldn’t have heavily customised Samsung phones but it is the way in which vendors make their devices stand out from the rest – to give the Android on Samsung a uniquely Samsung experience when compared to what other vendors do.

Here is a good example, GNOME desktop has a particular release schedule but that isn’t necessarily going to line up with the release schedule for the OpenSUSE distribution so sometimes what you have a gap between when GNOME is released and when that updated GNOME rolls out to end users. Then add on top of that the OpenSUSE distribution folks building and testing for bugs – and sometimes the bugs that are specific to that particular distribution meaning that patches have to then be created to address the issues and then pushed upstream to their respective project source tree.

Given the dynamics of what happens in the Linux world one can apply the same sort of logic to Android where the code is released from Google to the AOSP then Samsung gets that code, merge it with their driver stack etc then test it, test it and test it some more then release it once it is ready. In other words, view the ‘release’ of Android as a code drop and follow the schedule set by the OEM (in other words view what you have on your phone as “Samsung Android” which has its own release schedule like how OpenSUSE has its own release schedule) – in the case of Samsung they release upgrades 6-8 months after the Android code drop which places it inline with the summer/spring (northern hemisphere) which is just before the usual announcement of new Samsung phones.

Now, regarding security updates, not every update is going to be relevant to your particular OEM hence it might not be necessary to push out an update for a security issue because your phone isn’t impacted by it. For example, in the latest January security update there are 4 security related fixes for the ext4 filesystem but if your vendor doesn’t use it then does it need updating? There is an update for Dragon BSP support from nVidia so unless your device uses that piece of hardware then your hardware isn’t impacted. That doesn’t even go into the various Qualcomm components where 6 of them are specific to Qualcomm so if you have an Exynos SoC then you’re left unaffected by it.

Then there is the other factor, as mentioned further up, as problems arise the vendor will make patches so there is a good to fair chance that the problem is resolved not to mention that the vendor could also pull down the code up to the latest security patch plus also additional patches that have been made so a ‘January 2019’ update might include more than just that patch thus it isn’t entirely accurate to just look at what patch level it is as it is possible that it could also fix more than just what is listed in the security bulletin for that particular month.

Microsoft Office is now in the App Store

As promised at WWDC with the changes around sandboxing, under the hood changes along with Apple changing how much of a cut they received from subscriptions that were paid through the App Store from within the application downloaded off the store vs. the customer go to said organisations website and doing it manually, Microsoft has finally delivered Microsoft Office to the Mac App Store (link). What that will mean is that each application is downloaded as a separate piece of software and all updates will be distributed through the store rather than the situation of having to have a ‘Microsoft Office Update’ daemon running in the background. I’ve given it a try and everything is working well – no differences as far as I see it when compared to Office 365 that is downloaded from the store.

Apple and privacy

Just before I start, I’m going to put some ideas out there – not in a ‘here is a grand unified theory’ but rather a collection of observations that I’ve come across and I’ll let you decide what are worth considering vs. worth throwing away.

Not too long ago I wrong an article (link) and I was heading back home tonight having an internal debate in my mind regarding the issue of privacy particularly when you consider how businesses like Facebook, Twitter, Facebook and others rely on collecting user telemetric data to then use it to deliver services, target ads etc. Just before I start I have to disclose the fact that I use separate logins for all my services with the only link existing between accounts are between this WordPress account and Twitter so that when I make a post on WordPress that it is automatically syndicated to Twitter so all my followers can also see what I’ve just recently blogged about. Other than that sort of arrangement I keep all my accounts separate which stops the sort of situation in the case of Facebook and third party access hence the reason why Apple doesn’t allow such integration with iCloud and why Microsoft stopped doing it a few years ago back in the days of Microsoft Passport.

Apple have carved out a niche for themselves over many years as the company that cares about your privacy (side note, I’ll be writing a blog post tomorrow about Apple and their sales decline and what their new direction appears to be vs. what direction they should take) as a point of differentiation when compared to an organisation such as Google whose primary source of revenue is the collection and monetisation of data. Apple talking about privacy isn’t new given that Steve Jobs talked about the importance of asking the customer for permission before getting access to sensitive data and for the customer to know what they’re signing up for ‘in plain English’.

From a pragmatic point of view it is easy for Apple to play the privacy card as a point of differentiation given that the primary engine of their business model isn’t the accumulation of data and the monetisation of that data but rather selling physical products along with services such as Apple Music, upgrades to iCloud storage along with sales of software and subscriptions via the App Store where each transaction Apple takes a cut so there is no need to do what Google or Facebook does.

This strict approach however, as some have argued, has resulted in Siri lagging behind the competition because it restricts the ability to use large data sets stored in the cloud where as Apples approach is to do as much of the processing on the device itself and keep the least amount of data in the cloud. In the case of Google it allows the AI system too not only learn from the information for that specific user but to also use learn using data from all the users that use the particular service which allows the AI to be refined by feeding it more information. Then with that more information then you have demographic such as sex, age, location etc. that can can be used to predict not only based on your own patters but also the patters of people of similar backgrounds and behaviours as well.

So at the moment there is a tug of war between sticking firm to the idea that AI can be completed locally without compromising user privacy but will it always lag when compared to an AI system sitting in the cloud that is able to suck up huge amounts of data and learn at a faster pace simply by the share volume of information? I ask that because customers have a strange habit of saying in surveys that they care about something but when push comes to shove, what they say is important isn’t necessarily translating to a change in behaviour. For example, scandals surrounding Facebook where in the United Stats the number of daily active users has flatlined at 185 million users and in Europe there has been a modest decline from 282 million down to 279 million but that doesn’t include its other ‘properties’ that it owns such as WhatsApp and Instagram, so if there is a concern about privacy it certainly isn’t translating into a major drop off in users (disclosure: I’m on WhatsApp – I’d prefer to use something else like iMessage but my brother and mum use Android phones so here we are). I guess only time will tell to see whether it pays off in the long run – will consumers be happy to sacrifice the the strengths of Google’s approach to AI learning for the sake of privacy or is the whole concern over privacy something that is created by the media (clickbait, column inches and headlines on television) rather than genuine concern beyond “I’m shocked but I’m going to go back to what I was doing before”.

With that being, things may end up start tipping in favour of Apple especially if governments start cracking down on the very behaviours that impact their business model but are concerning to privacy activists. Personally I think that in the case of Google, the impact would be disruptive but Google does have the benefit of having something beyond simply collecting data to serve up ads to provide a given service. For example, there is Google Cloud platform (aka ‘utility computing’ as it was referred to back when Sun Microsystems was talking about it being the future of computing) which recorded a 29% growth in the most recent quarter then add on top of that there is G-Suite, YouTube Premium, the newly launched Google One and I’m sure if Google wanted to then I could imagine them turning other services into subscription – $50 for a wall to wall ad free experience from search through to video etc. would make for an interesting conversation as to how far users on the internet are willing to pay for services if it means zero ads and companies collecting less data for the sake of monetisation. There is also the various other divisions such as selling physical products such as the Pixel phone, tablet, laptop etc. with the capacity to push sales harder than they are especially with the availability to run Linux applications on Chromebooks which opens up the possibility to third party applications thus making them more widely useful.

Facebook however is going to have a tougher time – yes, they have a corporate collaboration platform called ‘Workplace by Facebook” that is $3 per month per user but I don’t see it gaining much traction in the enterprise marketplace. That being said, as of November 2017 there was a filing with the SEC that Facebook was generating $5 revenue per user so that makes me wonder whether there is the capacity for Facebook to offer a privacy centric ad free version for maybe NZ$10 per month butMark Zuckerberg, when asked by members on the Senate Judiciary and Commerce committees  about the idea of having a paid for ad free version – he avoided answering the question by making a statement that was irrelevant to the question  then later admitted that it could be possible. The Washington Post claims (link) that the cost would be closer to $18.75 which is even more stupid because it assumes that a paid version would have the same overhead costs as a free version where the platform service is not only delivered but also all the cost of delivering ads along with the analytics that occur behind the scenes. Wouldn’t it be better to generate less revenue but at a higher margin from regular paying customers than higher revenue with lower margins involving the very things that’ll bring your company under scrutiny and make life for you more difficult in the long run?

Tim Cook has talked about delivering more services this year so it’ll be interesting to see what they are and whether they’re going to be services that are open to all platforms or whether it’ll require an entry point to the platform via owning an Apple device of some sort (such as in the case of setting an iCloud account with an email address). It’ll also be interesting to see how Apple monetise them – will they be happy by having a small subscriber count if it means that it is a more profitable venture (see Apple Music and the direction of preferring paid only options rather than the ad supported model that Spotify use as a gateway to pump up the user base numbers) without the privacy worries? Time will I guess.

I wish that the weekend was longer

Sigh, one of those times when you wish the weekend was longer given how nice the weather is – windows and door to the backyard open with a nice refreshing breeze going through the house whilst listening to the album ‘Stop making sense’ from the band ‘Talking Heads’. It is one of those albums that can’t but want to listen from start to finish – no interruptions, enjoying the music as it was intended.

In the mean time I’m giving Chrome a try on my iMac, MacBook Pro and iPhone to see how it performs but so far I am pretty damn happy and they’ve finally addressed the issue where spell checking wouldn’t work when using Twitter, WordPress and other websites – I don’t know what they did to fix it but it is now working.

Why am I giving Chrome a go? I’ve started doing my tweeting via the twitter website because unfortunately Twitter moved to a new API that, unless you pay for the full featured API set, the experience for those using third party applications is less that ideal when one compares to using the website. Chrome has the best PWA integration and I have found that Safari tends to really lag especially if you’ve loaded up a long list of tweets and doubly so if there are animated gifs and videos too.

The trend I am noticing with the move towards progressive web applications (PWA) is either the killing off of API’s that third parties can use to create locally/offline clients or if they are continuing that the API’s are either crippled or they offer a full featured API only if you’re willing to pony up with the money to get exclusive access to it along with the engineering support that comes with it (I assume that the amount charge is in part to cover the engineering costs of maintaining a public API that remains compatible with the various client that may use it).

I wonder to what extent the long term goal for these organisations is to eventually move to PWA’s so that there is a uniformity between their mobile application and what the end user uses on their computers web browser along with not having to create and maintain API’s which free them to make changes behind the scenes without having to worry about the downstream consequence to third party applications.

Apple doubles down on the privacy push..but I’m sceptical

Tim Cook wrote an opinion piece in Time (link) talking about the need for greater regulation as well as a data brokerage service where consumers have the ability to find out who is using and selling their personal information along – greater transparency across the board. I have to admit, I like the idea of greater transparency because although Steve Jobs did talk about ensuring that customers give permission when an application wishes to access personal data or use a piece of hardware that can capture information (GPS, camera, microphone etc) the problem is there is little in the way of answering the question of what happens to the information after one gives permission. It is one thing to give permission but without the context of what the vendor is actually going to do with the information not to mention having some sense of control after the organisation gets the information makes the dialogue a tick box rather than something that is actually empowering.

With all that being said, I wonder to what extent the publicly actually do have concern regarding privacy and the use of private information given the public’s willingness to keep using Facebook (along with WhatsApp and Instagram) even after the avalanche of revelations came out over the last few years. You’d think that after all the revelations and even more revelations which turned out to be in some cases worse than the prior ones that people would start closing their Facebook accounts but alas here we are almost 2-3 years later and nothing has happened – people still have their Facebook accounts and new accounts are being opened all the time which makes me wonder whether people are genuinely concerned about privacy or is it noise and protest all for show but behind closed doors people don’t actually care.

I think the larger question is whether certain services can and should be treated as natural monopolies due to the higher barriers entry such as high initial start up costs, the years of running at a loss before either breaking even or becoming part of a larger organisation not to mention trying to bring ‘stars’ over to said platform which make the platform useful to end users – in much the same way that a mobile platform is only useful based on the applications that are available to end users on said platform. This goes back to the question about de-platforming – if the number of platforms are limited and the possibility of setting up new platforms is next to impossible then do those entrenched natural monopolies have a responsibility in much the same way that within the European Union a monopoly has a greater responsibility to ensure that the the monopoly isn’t entrenched (see EU demanding that Microsoft open up various protocols and file formats for better interoperability) and free speech isn’t stifled due to the lack of available platforms now that the internet has become ‘speakers corner’/’the village square’.

That all being said, I think the inherent nature of capitalism with its capital accumulation (and concentration) which leads to political capture (the capturing of politics by those with money) will make any meaningful change impossible resulting in any legislation being passed being half assed and watered down to the point of being useless in much the same way the Dodd-Frank never addressed ‘too big to fail’ by not only breaking up the banks (investment separated from retail banking) but actually turning these banks into credit unions which would undermine the whole concentration of power (the reality is that there are a small number of people/organisations that own the majority of the shares). Who benefits after all this? Apple because they’re already ahead of the curve – “see, we can be trusted because we go beyond what the law requires”.

Intel’s new CPU and GPU direction: Lessons learned and humble pie consumed

Intel has come back into the CPU/ GPU ring fighting and part of that involves a generous helping of humble pie by Intel admitting that they screwed up as seen by the fiasco with 10nm – Intel has its press release (link) where as Arstechnica goes into more detail in their news story (link). It sounds like a repeat of the mistake surrounding the P4 but on a smaller scale but that being said it has had an impact not only on Intel but also its partners which makes me wonder whether the turn around will impact existing plans at Apple (regarding the rumoured ARM) or whether it is full steam ahead regardless because the decision being made is external of what Intel does. With that being said the architectural improvements to squeeze even more performance out. There is also a big focus on improving GPU performance (link) along with rumours that Intel looking at maybe even offering a discrete Intel dedicated GPU which will make for interesting competition to the current duopoly of AMD and nVidia in the professional and gaming market.