Tag Archives: Cloud

For play, Four play or Foreplay

That is the game and today it is the setting of the BBC to get their buns burned, or at least that is how I see it. The article by Liv McMahon is nothing short of a joke. We are given “Snapchat, Reddit and Lloyds Bank were among more than 1,000 sites and services reported to have gone down as a result of issues at the heart of the cloud computing giant’s operations in North Virginia, US on 20 October. In a detailed summary of what caused the outage, Amazon said it occurred as a result of errors which meant its internal systems could not connect websites with the IP addresses computers use to find them.” And I particularly liked the ‘application’ of detailed. It is followed by groveling and whatever by Amazon, and an explanation by Zoe Kleinman, so the detailing was left to someone else. We are given “Amazon said it came down to an issue in US-EAST-1 – its largest cluster of data centres which power much of the internet. Critical processes in the region’s database which stores and manages the Domain Name System (DNS) records, allowing website URLs to be understood by computers, effectively fell out of sync.

According to Amazon, this triggered a “latent race condition” – or in other words unearthed a dormant bug that could occur in an unlikely sequence of events.” So, a bug that could in fact happen if an unlikely event would take place. So, a system at the corner of everything could fall over. You know, Elon Musk gave me a simpler setting, He gave me this image through Twitter (still refusing to call it X).

As I see it, this image is clearer than your whatever you called that piece and it shows the setting that this should not have happened and what were these unlikely events? You fail to disclose this, but that is the foundation of the BBC at present, catering to terrorists (Hamas) at every turn and not triple checking your facts. And there is a need to solve this. You see, Dr. Junade Ali (from Institute for Engineering and Technology) gives us (through you) “Dr Ali believes it highlights the need for companies to be more resilient and diversify their cloud service providers “so they can fail over to other data centres and providers when one isn’t available”. “In this instance, those who had a single point of failure in this Amazon region were susceptible to being taken offline,” he said.” He is correct and that also sets the current ‘drive’ to non-existing AI to a halt. If this is set to AWS standards, there is every likelihood that this flaw is replicated through their AI front and at that setting when this curve is hit, error on error will creep into a system that isn’t supposed to have it. I kinda trust Oracle to have is solved, but AWS might fall over. As such what will the damage be at that point? You can doubt and deny this, but I just illustrate a fall over point and if it has to be addressed at this point, what will the damage be to the consumers of Amazon AI? 

Systems built onto systems and managed by systems when a fall back flaw hits is the start of an unstoppable disaster, or at least unstoppable until there is human interaction and it took approximately 15 hours to fix. Now consider that the decisions of an AI are unchecked for over 15 hours, what damages does this setting bring?

In other news, I got “Many major websites and apps became inaccessible due to a Domain Name System (DNS) issue affecting AWS’s DynamoDB database.” The word Dynamo does not enter your story even once. Seems like the BBC left the facts on the floor, is that how you operate at present? As I personally see it, the Image from Elon Musk was more revealing in this instance and he didn’t have to write a word.

In this, the last word was given to Dr. Junade Ali was spot on “In this instance, those who had a single point of failure in this Amazon region were susceptible to being taken offline

He seemingly was right and the damage is seen through a thousand corporations big and small and it seems that this “dormant bug that could occur in an unlikely sequence of events” is exactly what organized crime is looking for, a place to hold over everyone as a hostage to their needy revenue. A point they can attack. I think that it is a massive setting that needed fixing last month to be certain, because what was, can explicitly be again. That is how organized crime works, unless they have Filofaxes, which makes them very organized crime at that point.

So as I see it the players are Consumer, Technology, Amazon and opportunity (by anyone). So there are the four players and I reckon that this setting has plagued DynamoDB in a few ways and at least three months ago we were given “Teams are shifting from AWS DynamoDB to alternatives like ScyllaDB due to cost, latency, and multi-cloud flexibility issues. – DynamoDB’s fixed pricing and limited scalability struggle to meet enterprise demands for hybrid cloud adaptability.” So as I see it, there were more issues plaguing this weakness. Another thing that the BBC never showed us, at least not in this report. So what else was missing?

As I see it, have a great day, don’t forget your intake of Coffee (or tea if you are in the UK) and see where the flaws of others would impact you. Don’t rely on me because I am apparently heavily flawed.

Leave a comment

Filed under IT, Media, Science

Modus operandi on steroids

That is what I see and not everyone does that. There is the setting of oversimplification and I get it, we all want things to work. So when the BBC alerted us all to the outage that AWS experienced there is more to all of this. 

I am not on the side of Amazon here, or on their opposition for that matter. So when I saw the news that thousands of corporations went down I was eager to see the news. And as it was given to me “Platform outage monitor Downdetector says it has seen more than 6.5 million reports globally, affecting more than 1,000 companies” but why? And we get that with “There aren’t many alternatives to AWS – operating on that vast scale is an enormous logistical challenge” and I tend to agree with Zoe Kleinman on this. So as the BBC gives us “Amazon Web Services says it has fixed the underlying problem that has disrupted many of the world’s biggest websites and apps, but a full recovery will take some more time” and I go ‘underlying problem?’ And there Tom Gerken has an answer, he gives us “At 08:00 BST this morning, reports started flooding in of problems accessing a few apps. By 09:00, it was apparent this had turned into quite a big deal.

We know now that the culprit was something called “DNS resolution” not working properly at Amazon Web Services. In simple terms, it all comes down to the bit of tech which lets a computer understand what we mean when we see a url like bbc.co.uk. But the reason it had such a big impact is simply that a massive amount of companies rely on Amazon working properly.

Downdetector told the BBC it had received reports stating more than 1,000 companies were facing problems. The question now is – will some of these companies look to alternatives?

You see, the problem is that ‘everyone’ expects a setting to work outright all the time and the old premise is “You can fool all of the people some of the time and some of the people all of the time, but you cannot fool all of the people all of the time” this can now be ‘tweaked’ into “You can service all of the people some of the time and some of the people all of the time, but you cannot service all of the people all of the time” you might think that this is folly, but it is not. You can introduce larger pools of resolution, but the system was designed to work all of the time, there was apparently no switch over and that might have resolved things. I am also contemplating that an outside source had introduced something to make it fall over. Was that the case? Amazon and its AWS pool of technicians are top notch, as such this hiccup might have been foreseen. 

My thoughts on the third party comes from the news “The latest update comes after AWS said, at around 12:00 BST, it had fixed the underlying issue, but noted there would still be problems as they brought everything up to speed” and this happens around noon? I don’t believe in these coincidences. Like noon British Summer Time? Something seems amiss. We get the usual baby formula stories, because the baby needs feeding. Yet the idea of having something in stock was rejected? And I get it, we all need our sustenance. That’s why I keep 3 days of spare food, so when this happens I am not helpless. 

So that gives us to the ‘latest’ issue. We are given “After today’s Amazon Web Services outage impacted many of the world’s biggest businesses, some customers might be asking whether they can take legal action for any disruption they might have suffered. Henna Elahi, a senior associate at Grosvenor Law in London, explains that whether money can be recovered will depend on “several factors”, including the contracts between the various parties and the severity of the outage. For instance, banking apps are among those that saw thousands of reports of issues.” And I get that, some people will cling to legal settings and that is fine, but that gives me the following questions.

Does these contracts raise glitch issues? Was there an insurance setting to prevent this? Was that insurance paid or did everyone just assume that this is a free service that works 100% of the time?

I reckon that AWS will investigate how this could have been prevented or diminished. You see when this happens on these AI systems and you can disrupt these services, a glitch like that will allow you to short sell what AI data is handled and that implies organized crime intervention on nearly every level (or state players).

We were given:

This implies that the entire setting took less then 5 hours to fix, I say ‘Yay Amazon’ but the underlying setting that what this had such a massive impact, all whilst North Virginia was affected is the cutting question and whilst we can think that it was in North Virginia hence the CIA is to blame is just ludicrous (yet, not out of the realm of possibilities) my issue is that a setting of decentralized cloud computing might be required. Hence as one system goes down one of the other takes over and as we are given that “The AWS Cloud in North America has 31 Availability Zones within 9 Geographic Regions, with 31 Edge Network Locations and 3 Edge Cache Locations.” My question becomes (optionally utterly ridiculous) “Why did it take 4 hours” with the added “When cloud computing is nearly ‘global’” perhaps there are good reasons for this, perhaps this is the first time this went down to this degree and that is fine. Things go broken into the night and the next morning we have a stronger system. This is the track of evolution and it never goes without a glitch. 

But the idea that one centre had this much of a global impact? Consider that when the Stargate contraption goes online and power gets disrupted. See what you optionally lose at that point. Because that is the underlying setting. It isn’t what we have now, it will be what we will have tomorrow that counts as disastrous.

Have a great day and in case it happens again, don’t rely solely on your credit card, make sure you can afford to pay for that coffee (that ancient system using coins). 

Leave a comment

Filed under Finance, IT, Science

The changing of games

There is the thought that games are changing and the first question becomes ‘What games?’ And that would be a correct way of thinking. Whether you decide to kill the bullies and their connections, whether you stand up for yourself or if you become a lot less visible. The latter part is my preferred way, what doesn’t see you, will not hurt you. But in the 80’s I learned the hard way that always on the defense tends to be pointless, as such I would be inclined to scurry over to the kill side. It has the benefit that the the magicians of this world get plenty scared when the bunny bites them. They aren’t used to the sight of their own blood. They tend to cry and wonder why they can’t be bullies anymore. You see, at some point people have had enough. Some like me tend to weigh the consequences of being bullied or to eradicate them and live the fallout. At some point accepting the fact to be bullied no longer weighs high enough and when the Sydney Metropolitan police departments tend to do nothing, even as they have Brodies Law at hand, they prefer not to act. It only works so far. 

This story is important to the real deal, it is a story that ZDNet gives us (at https://www.zdnet.com/article/why-denmark-is-dumping-microsoft-office-and-windows-for-libreoffice-and-linux/) where we see that Denmark is now apparently ‘Why Denmark is dumping Microsoft Office and Windows for LibreOffice and Linux’ and in that setting I wonder if Danish voices might also float towards WPS Office (by Kingsoft), you see, ZDNet gives us “Denmark wants to claim “digital sovereignty.” In the States, you probably haven’t heard that phrase, but in the European Union, digital sovereignty is a big deal and getting bigger.” I see that this is one avenue I never considered. Oh, I’ve heard the term. Yet the larger setting is not what I have heard, but what is behind it. Denmark is likely furious by some bully that wants of annex Greenland (an island West of Denmark a mere 2.166 million km²) and they are decently angry and this was the first setting. After being fed up with the Trump stage, they decided to take Microsoft out of the equation. At that point a lot of settings that ‘drip’ into American data settings and in this the first stone is cast. You see, President Trump might seem to think that business will adjust towards American standards, but that is a little delusional. You see, Microsoft is seen as a 3.56 trillion company, but behind that is a towering amount of debts as well. The totality of debts is according to some A$93.09 Billion. This might not seem as much, but what ‘victories’ have Microsoft made? What spin actually represents revenue? Microsoft is all about revenue and net profit, yet the larger setting becomes “In Q3 2025, Xbox gaming revenue decreased by 7% year-on-year, but content and service sales increased by 8%.” So Microsoft sets a plus to diminish the minus, yet the larger station is that they lose a lot more than they gain, for what is the depending value of the 8% rise? It is not the same as based amount of the 7% drop. Microsoft is losing against Sony 3:1 and now that the Nintendo Switch 2 is out, these losses will merely increase overall. Whatever Microsoft has as a tablet doesn’t even dent the setting Apple has and as some see, their Azure state seemingly doesn’t hold a candle to the system some book dealer has (yes, it is AWS). Then we get the setting that their ‘edge’ yes, their browser only has a 5% market share against Google having 67%, Apple follows with 17%. Now how many failures can such a company hold? And now consider Huawei entering the field with HarmonyOS. Taking market share from both Android and iOS. That was the setting before today and now Denmark is seemingly the first to drop Microsoft for other paths in IT. So how long until Denmark convinces one of the other EU nations to follow suit? What losses will Microsoft endure before they sink some of their badly conceived projects? I don’t know, I am merely asking. 

As such Microsoft is speeding to get a lot of the HarmonyOS population, but as Kingsoft grows Microsoft diminishes and the that population never had much love for Microsoft and America to begin with. And we see part of this with “EU leaders are seeking to reduce Europe’s dependence on foreign technology providers, primarily those from the United States, and to assert greater control over its digital infrastructure, data, and technological future.” And another part is that they’re concerned about who controls European data, who sets the rules, and who can potentially cut off access to essential services in times of geopolitical tension. And the tariff war doesn’t help. That setting instigated by President Trump is likely to ht Microsoft faster than they realise and what happens when these debts will rise as revenue decreases.

The next part is alleged settings and I have seen no evidence of this from other sources “President Donald Trump issued ICC sanctions. This order allegedly prompted Microsoft to lock the ICC’s Chief Prosecutor, Karim Khan, out of his email accounts, according to reports. This came after Microsoft chairman and general counsel, Brad Smith, had promised that the company would stand behind its EU customers against political pressure. Recently, however, Smith stated that Microsoft had not been “in any way [involved in] the cessation of services to the ICC,” according to Politico. When pressed, Microsoft failed to further explain how the email disconnection occurred.” That might (or not) be a complete answer. I have to add that the entire builder.ai fiasco is on the hands of Microsoft. They backed this and they never saw the 700 engineers programming what on existent AI was supposed to do. So where are these 700 systems, their OS and their Azure licenses? Wouldn’t that be firmly on the eyes of Satya Nadella? And as such, how was this worth a billion dollars? If Microsoft was entirely unaware they could be seen as incompetent (or at lease some people on the VP and higher list). If they did know there is a larger failing at Microsoft going on and as Denmark is allegedly dropping Microsoft, it is the start of a lot more bad news. But they can rely on spin to keep the eyes of others somewhere else.

And we see that (allegedly) see that in part with “Whether or not Microsoft cut services to an organization in response to Trump’s order, the fear that it could do so in the future remains. Before the Danish government announced its move, Denmark’s largest cities, Copenhagen and Aarhus, had already announced plans to phase out Microsoft software and cloud services.” So why allegedly? The setting is fear, not data and whilst we see the results we might see the wrong facts leading to this. As I personally see it “plans to phase out Microsoft software and cloud services” might be du to the fact that AWS is as I see it vastly superior (vastly might be overstating it) and fear could weasel in at any point, almost anywhere. Yet the likely accusation that Microsoft is the ‘bitch’ of President Trump or any American administration will be much harder to counter. It could set the tides against Microsoft in Denmark (for starters), Canada and McDonald islands (both users). So there is space to maneuver, yet Microsoft doesn’t do that and we are left with the accusation. And the larger setting that “In particular, the Danes are worried about Trump’s policies and that US political decisions could put public IT services at risk.” Is a decent fear to have in these days, as such Microsoft will be left holding the political bag. And Denmark is not alone here “Bart Groothuis, a Dutch member of the European Parliament, recently said, the EU “should go for a European cloud” since “Europe has a ‘problem’ with American cloud.”” I am not sure how this ‘computes’ in a downside for Microsoft, but the spin masters will have their hands full because that increases the Danish setting by 100% and there is no way telling what else is at risk and who else is to follow suit.

I saw a different variation of the ‘downfall’ of Microsoft, for the most their lack of actual innovation, their dependency on marketing spin (or whatever Microsoft calls it) and their failure to deliver in several fields. And their enemies are at the gates. After Microsoft failed the mobile markets (it is near zero) and as Huawei is gaining massive levels and Microsoft is losing market share after market share and Denmark clearly showed that they see Microsoft not as a partner but as a threat. As such I have to wonder, in what field will Microsoft fail next?

Have a great day.

1 Comment

Filed under Finance, IT, Media, Politics, Science

Delphi in a name

Yup, we are talking about Oracle, not Borland. And whenever I hear Oracle I tend to add the ‘of Delphi’ automatically. It is a Pavlovian thing. This is nothing negative about Oracle, I wanted to join their ranks in the 90’s, and beyond the millennium a few times too. My origin settings was a database programmer (I earned my stripes with Clipper, the Nantucket version). I think it is the very first program where I shelled out $650 (Dfl. 1,200) for a program and I learned a lot through Clipper. I also got the Clipper notes (Norton Notes) and these two kept my in my apartment (on a desk chair) for weeks and weeks at a time. I relish these happy days. Then of course I got into technical support and customer care through a precursor of IBM and my life at that point was pretty complete. I miss those days and I still think fondly of them. Not so much the upper ranks of that company with their political games, but them I was never a political player. 

So when I saw ‘Oracle commits to invest $14bn in Saudi Arabia over next 10 years’ (at https://www.datacenterdynamics.com/en/news/oracle-commits-to-invest-14bn-in-saudi-arabia-over-next-10-years/) my mind starting swirling and twirling (sorry JK Rowling) and my creative logging started to set new parameters. 

You see, we are given “Oracle has committed to investing $14 billion in Saudi Arabia over the next 10 years to expand its cloud and AI offerings in the region. The plans were announced by the company on May 13, and in the wake of President Donald Trump’s visit to the Kingdom” this implies Technical Support, Customer Care and Trainings. Things I can do (all three) and I have had well over a decade of experience in these sections. As such I keep my eyes open for positions needed in either Riyadh, Mississauga or Abu Dhabi. I reckon that the investments are not just for Saudi Arabia, they are all spend in Saudi Arabia, but there will be essentially needed persons in Abu Dhabi because no one walks away from ADNOC and with ARAMCO in Saudi Arabia, a secondary call center would be needed in Abu Dhabi. And they too will have all three settings in that centre, beyond that I reckon that it will a location will be cheaper in the heart of ADNOC than in Dubai, so there.

When we see “Our expanded partnership with the Kingdom will create new opportunities for its economy, deliver better health outcomes for its people, and fortify its alliance with the United States, which will create a ripple effect of peace and prosperity across the Middle East and around the world.” The words “a ripple effect of peace and prosperity across the Middle East” merely implies (not confirms) the setting I see. You see, it makes sense to do this, but it requires knowledge of Oracle policies (and I don’t know those).

So when we see “Oracle has two existing cloud regions in Saudi Arabia – Saudi Arabia West, located in Jeddah, and Saudi Arabia Central in Riyadh. The former was launched in 2020, the latter launched in 2024, and is hosted in a Center3 data center. The company has been planning a third in the upcoming Neom City since October 2021, which remains listed on Oracle’s website as “coming soon.”” Someone would think that another cloud the UAE cloud should be there as well. Merely not mentioned in this stage, but ADNOC is too big to walk away from and Microsoft has dropped the ball too many times. There is a setting that implies that IBM and or AWS are already there, but that gives the larger setting that ADNOC becomes dependent on one supplier and they are as smart as they come. So I am betting that Oracle has that region (as well as Dubai) in mind when we consider DAMAC (valued at US$ 595 million) with the total revenue recorded by DAMAC Properties was AED 7.5 billion (2017), and they are not all. There is also Emaar Properties, which is said to be the biggest of them all and that are the kind of clients Oracle really likes to keep happy, as such I saw the stage evolve, even though they are already there and in January 2025 we were given ‘Oracle to increase Abu Dhabi investment five-fold’, as such I think that there might be a new need to seek employment with Oracle. Now add to that the quote “Earlier this month, the Abu Dhabi government put out a call for the development of a single multi-cloud system that will serve more than 40 government entities” and you’ll see that there might be space for me too, either in Abu Dhabi or in Mississauga and the two cover a little over 20 hours a day coverage in a 24:7 setting. The nice part is that it takes time to get people up to speed, so I might have an advantage (merely a slight one). 

So as I am about to dream the day away on this rainy Sunday. I see the cogs of industry revolve around the settings of the world and I keep having happy thoughts.

So have a great day everyone, preferably less rainy than it is here.

Leave a comment

Filed under Finance, IT, Media, Science

Overdrive, or drive over

That is the setting. We can try to set the premise of DeepSeek (a waste of my time), we can set the premise of Microsoft AI (a waste of everyones time) and yes the 14 billion will have an effect and we can speculate on the 500 billion that StarGate is going to cost and what exactly will be the enabling part. Did anyone consider the ROI of that idea? That prospect will need to make at least 15 billion annual to make it worth. Throwing big printed cash at it will be as useless as the quantitive easing that Mario Draghi promised about a decade ago. Yup, it won’t go anywhere. 

But that led me to a setting many seem to ignore, so lets have the list:

Microsoft 365 Copilot: A monthly subscription that costs $30 per person. Copilot Free is available with the Microsoft 365 Business Basic plan. Copilot Pro is a monthly subscription that offers more advanced features. 
So at present, how many people are on this plan? It seems that Microsoft isn’t to talkative on ‘how successful’ it actually is. We get spread numbers and these numbers doesn’t seem to validate the billions invested.

Azure Machine Learning: A pay-as-you-go service with pricing based on the number of vCPUs. 
Azure AI Search: A service with pricing based on the number of text records or images processed. 

Here I have more issues. You see, we are given “Azure AI. Azure AI provides users with powerful tools that can be used to create innovative solutions using machine learning, natural language processing (NLP), computer vision, and more” How can any machine learning create innovative solutions? If it is machine learning someone else has it already, making it reengineering at best, optionally an innovative patent. I always (perhaps incorrectly) see pay-as-you-go as a dodgy solution. You either commit, or you don’t. 

Computer Vision API: A service with pricing based on the number of transactions processed. 
So, a service based on transaction processing, on that case if the IT department doesn’t throttle its usage there is every chance that an intern could blow up cost as it is happening.

Azure AI Content Safety: A service with pricing based on the number of text records or images processed. 
Azure AI Content Understanding: A service with pricing based on the number of hours of content processed. 

All this is set to a counter (like ConfirmIT) and that is the only company that had a good handle on it, a setting with decades. Now, there is a chance that I forgot a few solutions and that is OK. I am not heading an aspirational setting of academic instance.

You see everyone is on the bandwagon and I am too tired (or too old) to care. The media can’t be bothered unless digital currency is flowing their way. Yet in all this when did you see a clear description of AI solutions in use by Amazon, IBM or Oracle? You see, the DeepSeek issues of the last few days stirred a few minds. They are now also seeking Return on Investment (ROI) and that image is not clear, at least the media seemingly can’t be bothered and the influencers now shouting their wisdom on LinkedIn are also at times tedious and for the most a waste of everyones time. So why Microsoft? I don’t really care about it, but they (and their sickofans) are shouting how good their solutions are, but we see no clear numbers. And at present clear numbers is what the most of the population want. 

Am I wrong?
I doubt it, the signs are there and when we see a small message on the left, the right clearly muffle that sound out. You see Shelly Palmer in IEEE Spectrum writes “As for the 100,000 jobs the project is supposed to create? Some construction jobs will be created as the data centers are built, but many more (millions more) will be created as the data centers come online. We’ve never had a compute cloud like this—there’s literally no way to calculate the economic impact of this amount of AI compute. It will be massive.” I actually don’t know about that. The idea that “there’s literally no way to calculate the economic impact of this amount of AI compute” is as I see it bogus. For 500 billion ($500,000,000,000) I expect more. But at present it comes across like a huge NSA data collection hub. Come to think of it, We could (optionally) get some data from the NSA, Google or IBM. They have experience with really big data centers. So what are those costs? What is the return on investment? And there is the setting of the value of collected data and that will not even have value until lots of data is collected, so lets say by 2030 and all those billions need to show investment value and at present the big-tech market lost over 1 trillion dollars a few days ago. So where is the ROI of all this?

Then we get “There are many tech skeptics, and it has become fashionable to denigrate and vilify big tech. To me, the Stargate Project is the first step in securing the future of the U.S. economy as well as our digital and cyber security. Every business will benefit from the power and promise of AI, and—like it or not, believe it or not—warfare will be dominated by AI. Today, the U.S. has a clear lead. The Stargate Project will help ensure it stays that way.” My issue is that there are always skeptics, I am one to some extent and the words “the power and promise of AI” fills me with dread. It is the included word “promise” and warfare isn’t dominated by AI, the setting pf properly programmed deer machine learning is. It is not AI and it is unlikely to show until somewhere in early 2040 at best (as I personally see it) but the 500 billion is coming out of ‘our’ pockets now. Yes, I know what they say that corporations will push the bill. Yet when this goes pear shaped. They will al put in in a bad bank account and relinquish the debt as a write off, so you, in the end still pay the bill in some way.

Then there is the sentence “Today, the U.S. has a clear lead” do they? DeepSeek is Chinese and their setting blew the rest away, you want to find out what a two-nil for China looks like? You are about to see that in very unrespectful terms. And as everyone is on that so called AI horse no one is investigating it, the media least of all.

In the meantime I will reengineer games. There is at least some revenue in that. And as I saw the reengineering options for ‘Infamous: Second Son’ The Sony firms could get some more coins from an 11 year old game on the PS4. And now there is an option to get it upgraded to PS5. Consider the gaming population. Whomever played in to PS4 (early days PS4) would like the setting on PS5, I tried that original game on PS5 and it plays well. A few minor glitches but that is what happens. The storyline could be upgraded and with linearity removed the game would get a much tougher stance. Then add the ‘cleaning’ of Seattle and we get a more complete game. With the setting to an optional change to Smoke-TV-Neon sequence the game alters a fair bit, and in this the game could also encase the stealth option in the game. Take with that the option to go back to the beginning to free the people from concrete affliction the good and the bad will also alter to some degree and it isn’t merely the good and the bad setting, the larger stage of animosity could reverberate through the game. And I am now looking to a few more games. A setting that I believe is great for Sony in the immediate future. 

Can’t stop a creative mind puzzling on how to make something better, a trick that isn’t possible with Deeper Machine Learning and LLM’s. Have a great Thursday.

Leave a comment

Filed under Finance, IT, Media, Science

A changed setting

That is where I found myself a few days ago. The realisation that things weren’t what they were supposed to be. Now, it is not really new. Settings change, but for the most it is up to the makers to herald a certain stage of doing business. This is a strange telling, because I believe in the Robocop setting that Kurtwood Smith handed to us “Good business is where you find it” and for the most I believe this is true. The stage was handed to us by Satya Nadella when on December 26th 2024 he gave us “the era of SaaS as we know it is coming to an end, giving way to integrated platforms where AI becomes the central driver. This transformation is poised to disrupt traditional tools and workflows, paving the way for a new generation of applications.” Not only do I not believe him at present. He is paving the way for people to set doubt in a place and push them all towards Azura (i’ll get to this later). Still, this is a weird statement from Microsoft when we got on July 22nd 2024 ‘Microsoft joins forces with Austrade to help its Australian SaaS partners go global’ (at https://news.microsoft.com/en-au/features/microsoft-joins-forces-with-austrade-to-help-its-australian-saas-partners-go-global/), seems like a strange setting. And with the statement “Microsoft has today announced a new program in collaboration with the Australian Trade and Investment Commission (Austrade) to help local partners that offer software-as-a-service (SaaS) solutions accelerate their international growth” It almost sounds like the Asian joke “Two Wongs don’t make a Write” (or something like that). 

You see, as I personally see it, Microsoft is in trouble. It hatched its eggs too widely and too many of them are not paying off. There is only so many losses you can book and not take a massive hit. And as long as people are ‘dependent’ on Microsoft Nadella can sing whatever he wants. And that is where the shoe becomes a tight fit (and not in a good way). There is a cluster of people reposting and optionally with their ‘own’ insights as why it is such a stellar move. But there are issues.  You see, the first is that SaaS is a good solution for a lot of people, but as the Indian indie developers are gaining in that field Microsoft needs to haul exceedingly into another field where it is just them and their ‘agents’. And Microsoft will get a percentage for EVERY deployment we face.

The second setting is that SaaS goes together with IaaS and PaaS, but with the Microsoft setup all PaaS becomes Azure. It was the Microsoft solution to get from the statement “It is very possible to link single service of IaaS, PaaS and SaaS on 3 different cloud providers.” We got this answer three years ago and that never worked for Microsoft. You see, Microsoft wants it all. They failed too many times (in several fields). The need it all to survive and if enough are connected Microsoft (as I see it) prevents collapse. As I see it the AWS (Amazon) and the Oracle’s Platform as a service are vastly superior to Microsoft. As such Microsoft is dwindled down to size and they do not like it. I also think that Googles PaaS service is better than. Microsoft, but that is a more personal view then evidence driven. As such Microsoft needs to change speed and I reckon that the impending death proclamation of Software as a Service was Microsoft’s way to go and that is what Satya Nadella went with. The issue in this is an additional stage. In the 5 days of Christmas it is all that LinkedIn went with. I was torpedoed with these ‘news casts’ and opinionated settings from hundreds of sources (not only on LinkedIn) and these millennial sales screw ups all wanted a piece of that pie. They want it all whilst the getting was good and it is Christmas, wasn’t it? 

It is at this point when I wonder what Huawei has in store with their cloud solutions. It is the media appeasement of Microsoft that I wonder what the ‘enemy’ will bring us and that is where the setting stalls. The attack on our senses is almost infinite and some are deciding where we are able to (or allowed) to look. And we are all in the setting that we want to know where we can go and places like LinkedIn will not give us the full news making them propaganda channels for people like Microsoft. So when will we get the real deal of how to avoid Microsoft? I wonder what Oracle and/or AWS will bring to the table, them and Google would make a good replacement for Microsoft. But will we see that given to us, or is the influencer scene of Microsoft drowning it all out?

I cannot say for sure because the others are seemingly staying silent. Have a great day you all.

Leave a comment

Filed under Finance, IT, Media, Science

The Apple folly

Last week I saw a video on the new iMac, so much to tell, so much to show and they give it. Yet they forget about the number one issue. Storage! I wonder if that is the achilles heel of some Mac products. And this isn’t new. I gave clear warning over three years ago.

But let’s take this from the first square.

The new iMacs are equipped with the new M4, 32GB ram and 2TB storage at C$3,449.00. You think it is the bomb, but the MacBook Pro with the near same specifications costs you C$5,149.00 and with 4TB it is $5,899.00, new new Mac Pro with only an M2 processor is C$9,499.00 and with 4TB it becomes C$10,249.00. As such in this day and age it would have been prudent to include a 4TB setting to the iMac. You see, the other options are more than $1000 more expensive. You see, all these influencers, vlogger and photographers. 2TB doesn’t hack it. 5K video’s grow in demand getting a quick 128GB per video and 8 video’s get us one terabyte in space. I warned about this a year ago and no one at Apple had the notion of taking heed a simple equation. Then there are the photographers who get into camera’s with 80MP or more. That gets us file sizes of 480Mb if saved to 16 bit tiff. With a photoshoot easily surpassing 50-100 shots, the drive becomes too small. As such either these people cannot consider the new iMac and they are forced to get either another Mac with 2 to 7 thousand dollar more or consider a PC as a solution. So we have 64 million YouTube creators or vloggers that cannot consider the new iMac, how is that for jollies?

And for the people slow of mind, I saw this coming a year ago and I wrote about it in Adaptation 103 on the 19th of October 2021 (at https://lawlordtobe.com/2021/10/19/adaption-103/). 

So does anyone consider how Apple made this error. Do they want to push you towards the cloud? There is the real danger that whomever goes to the cloud, there intellectual property is possibly transgressed. As was reported this year in April “It was reported that 45% of breaches are cloud-based, and 69% of organisations admitted to experiencing data breaches or exposures due to multi-cloud security configurations.” In that atmosphere you want to push people to the cloud? That was my issue for years and with this iteration it seems that Apple might have lost the plot (as I personally see it). 

And I get that not everyone needs 4TB, but these groups (the vloggers) are millions where the new iMac is no longer an option. So how much business will Apple lose? Do you really think that the M4 chip and their so called AI version will fix that? I personally don’t think so. But you. Could make up your own mind. Personally I am not a vlogger, but I would like to be but I cannot do that on a MacBook Air, I can not afford a MacBook Pro and a Mac Pro is out of the question at close to 11 thousand. And that is before you get the Adobe solution, which is what pretty much every vlogger needs. So consider, has Apple made a booboo? I think they did and why be so happy about upgrading RAM from 24GB to 32GB and ignore storage needs? Jut a few simple question where we could surmise that Apple is nothing more than a new Microsoft (less error prone than Microsoft mind you).

Have a great Sunday, Monday is merely 80 minutes away here.

Leave a comment

Filed under IT, Science

Reengineering an old solution

I was bending my mind over backwards to stay creative. And as I was mulling over something I read a year ago, my mind started to race towards an optional solution. You see, the idea is not novel but it has been forgotten. So if Tandon never renewed their patent, you get the exclusive option to rule there. If they have, you could file for an innovative patent, giving you still a decent payment for your trouble. 

Going back 34 years
Yes, it was the height of the IT innovative time and this age had plenty of failures, but it also had decent blockbusters and whilst they all wanted to rule the world, they clamped down on their IP innovations. Tandon was one of those.

As you can see in this image the drives (both of them) look like space hoarders, it was the age of Seagate with their 20MB or 30MB drives. The nice part was that these drives could be ejected. It was a novel idea where the CFO could put its drive with the books in the vault.  

Why is this an issue?
Well, last year I saw an article that well over 70% of all cloud accounts were invaded on. To see this we need to look (at https://www.cybersecuritydive.com/news/cloud-intrusions-spike-crowdstrike/708315/) where we see ‘Cloud intrusions spiked 75% in 2023, CrowdStrike says’ it comes with the text “Organisations with weak cloud security controls and gaps in cross-domain visibility are getting outmanoeuvred by threat actors and struck by intrusions” And this is not all. Captains of industry lacking IT knowledge will happily accept that free 1TB USB drive at a trade show, not realising that it also creates a backdoor on their servers. They shouldn’t be too upset, it happened to a few people at the Pentagon as well (as they are supposed to know what they are doing). So the cloud is a failing setting of security. So consider that, as well as Samsung putting their stuff online because they didn’t realise how to operate OpenAI. Just a few examples. So what is to stop their research or revenue results to be placed on a drive like the pre-cloud days?

You think I would put my IP in the cloud? Actually I did, but I have a rather nasty defence system that is a repeated action I learned in 1988 and no one has a clue where to look (and I never put it with the usual suspects), but this is me and I will not give you that trick because all kinds of people read my blog. 

So back to Tandon. In stead of this big drive, consider a normal drive space and in stead of that big box. Consider a tray with enough space to fit an SDD with the connector inside the tray, going to a plug on the outside of the tray. With a simple kit that can be purchased if more than one drive is used. Now see the Tandon solution as it could be. An ejectable drive solution for many. Yes you can connect just a wire and use an external SSD, but it becomes messy and these wires can also malfunction. There is even the option of adding AES256 that could be added in the drive on one side, so even if they steal the drive (optionally with computer) the thieves lose out as a dongle could be required. It merely depends on how secure you want the data to be. A CFO might rely on his safe for the books. An IP research post might need more security. So consider if you want to be the optional victim staged in the 75%, or do you need your data to be secure. 

So whomever take the idea and reengineer it (with optional extras), you are welcome and have a nice day. I just completed 12.5% of Monday, time to snore like a lumberjack.

Leave a comment

Filed under IT, Science

A gap preceding a gap

This is a difficult article to write. A lot of it is speculation, optionally presumption. Yet if someone tells me that presumption, due to my lack of knowledge is inappropriate I would accept that. My mind wandered and I saw a weakness in a program called Final Draft. It is not really a weakness, but it could be THEIR opportunity. This site is debatable for them, but to be the more rounded solution would enable them to create a larger gap behind whomever are following them. 

My mind saw Ryan Reynolds (yes, the Canadian actor) starting talks with some guy named David Thomson. Not entirely sure what the conversation was about. But Ryan saw a new opportunity that grew from his insight in MNTN (www.mountain.com). You see the narrative of advertisement in changing. In that view we all focus on the advertisement themselves, but the larger stage is now becoming Development, production, deployment and so on. As such Final Draft might optionally be best to offer a new setting on Final Draft v13. 

You see, the missing part of project management. Even as there are some solutions, they are generic, for these people having one that is specific for Final Draft with timelines (somewhat similar to Monday dot com) would benefit Final Draft greatly. 

And it isn’t just for large companies, indie developers and script writers would be able to use it as well. Especially in scripts that have more than one season you want to see time lines, to test whether the scrutiny holds up and in that way advertisement companies could use the same setting. I reckon that the larger companies all have their own solutions, but Final Draft is used in over 60 countries and when you think of it, there is a larger pressing need to use project management solutions catered and tinkered to media and advertisement. Excel (not the worst solution) will no longer suffice in these places and as media corporations have a more distributed purview, that purview requires management with timelines, with options to see where pressure points arise and even more important when idle time and zero hours become too abundantly visible. We all have idle time at times, but when you have an entire crew on 40-80 hour idle time the cost will start to show and that impacts the margins and profit stages. 

For the simple developer (like myself) seeing the timelines and the completion times are also important. Not the fact of pressure, but to see where the lag is and whether certain parts were overlooked. It is the stage where the working project is set to a 90% complete versus 95% complete starts to show. If we take the old rule that completion of 5% past 90% has the same timeline as the first 60%, that gain is easily suppressed when you have a project file seeing what you optionally overlooked and in advertisement that part is even more clear, they work against (at times irrational) timelines and deadlines we see that Final Draft has an opportunity to grow its solution in another direction as well. I reckon that no matter where the project is set up, it needs to be deployed on a local level, because no executive will trust its data to the cloud, not with IP of this nature and also clouds fail (see Microsoft, September 2023). A distributed system might be the way to go, and independent developers (that one person studio) might not benefit with the cloud. And that is before you realise that there are 38,000,000 active VLoggers, they too need to up their game, the competition is cutthroat and murder as well. All options that show the possible opportunity that Final Draft has coming their way. I haven’t touched Final Draft since I tested version 6 decades ago, but I just went through the PDF of version 11 and I saw this gap and covering the gap preceding the gap makes the remaining gap passable and that is also where MNTN could find itself in the next 2 years. As they started the new setting the wannabe’s will come in soon and the better prepared they are, the stronger their foothold on the market will be and with all the others failing in obvious places MNTN could chisel out the niche that is safe, secure and almost spectacular (a subjective term). As such Final Draft has an even larger foothold it could create, but that is just me. I could be wrong after all. 

Another practical Tuesday filled. Time to rewatch Bullet Train and snore a little too. Enjoy the day.

Leave a comment

Filed under Finance, IT, Media, movies

The dangers we ignore

That is the setting we are confronted with, or perhaps better stated the danger that Microsoft exposed itself to. Now, I have been happy to snap at Microsoft at every option I see. Them souring the gaming world gives me ample reason to, or at least that is how I see it.

Yet the poll at LinkedIn gives me another view that I am not alone and yes, as you see I see Azure the biggest intrusion danger of the others mentioned. It is not the only setting that people face and I have issues with some of them. 

You see, there has been a larger issue with Microsoft and they are all about buying their way into other streams at the cost of $69,000,000,000 and we see very little issues on RESOLVING safety and security issues. There is (as I personally see it) a massive architectural problem with the Azure setting. Now, I have NO evidence that this IS the case, but the listings are starting to add up.

July 2023: How a Cloud Flaw Gave Chinese Spies a Key to Microsoft’s Kingdom
June 2022: 6 ‘nightmare’ cloud security flaws were found in Azure in the last year.
Mar 2022: Source code for giant’s web browser app, virtual assistant allegedly leaked

That list goes on for a while and the examples are all out there in the media and online. Yet, instead of setting resources that can fix and redesign that part we see too much spin and not enough fixing. Or perhaps what one fix achieves, it also opens other ‘windows’ into a blue blue data pool.

Now this is speculation from my sider, but the sources as I set them out were never mine. Microsoft is losing and shedding marketshare. This brings me to the article that partially sets this article off.

It was the Verge (at https://www.theverge.com/2023/10/5/23904375/uk-cma-microsoft-amazon-cloud-investigation) that gave us ‘Microsoft and Amazon face UK regulator investigation over cloud services’. In this my issue is sen with “It’s part of a fresh investigation into public cloud providers in the UK, after telecoms regulator Ofcom “identified a number of features in the supply of cloud services that make it more difficult for customers to switch and use multiple cloud suppliers.”” The stupidity of ‘that make it more difficult for customers to switch and use multiple cloud suppliers’ is the delusional setting of some wannabe. You see, you cannot have multiple mainframe operating services running next to one another, you cannot have more than one operating system for a SERVER to run together. You might have two servers and they may have different data settings, but that requires a specially designed API to exchange information, which is a massive security risk, which any corporation does not need. The interesting part is that this same danger would be a case with IBM and Google too, but they are not in that mess are they? Azure and AWS are the larger players and someone wants to cut them short (for whatever reason). A stage made optionally by stupid politicians, optionally with friends that have a solution no one wants (a speculation from my side) and no one is drilling into the claim that we see from the Competition and Markets Authority (CMA). I want to see the complete documents and the sources who investigated both Microsoft and Amazon. And the link we see in the article that relates to “Microsoft recently restructured the deal to transfer cloud gaming rights for current and new Activision Blizzard games to Ubisoft”. From my point of view Ubisoft after the next failure to bring a good product (AC Mirage raked at 78%) makes Ubisoft willing to bend over backwards to survive another year. 

As a character from ‘Who framed Roger Rabbit’ states: “this whole thing smells like yesterdays diapers”. And we are all in a stage to accept parts of this, but the political side is seemingly lacking in a larger stage of cloud systems and the amount of transgressions due to Microsoft failures are not met with official investigations and that is before they will block (as one might expect) any investigation into their shortcomings. 

Should you wonder about this, consider the 90’s and mainframes, or perhaps mainframes today and wonder how easy it is to switch those services. Yes, it might be possible, but consider the amount of dollars needed make such a switch non-realistic to say the least and that is on ALL providers. I feel uneasy to say that this should be possible, but I understand that it might have been an essential future issue. Yet, when we see the dangers of cloud services and the way that they are transgressed on. It might be that IBM and Apple clouds are the safest, or they are too small to get any representation and they are both in the other section, which is only 8%, as such the idea of either being a mere 4% against Azure scoring 50% must be some kind of hell for Microsoft and the amount of visibility of their issues are gaining strength all over the media. The Verge is not alone in any of this. 

No matter how people, media and Microsoft are spinning this, they have a problem and them diversifying in fields they do not understand for the mere setting of greed (as I personally see it), is a stage we should have been able to avoid and we are not, because the political parties in too many countries are willing to let too many Microsoft issues slide and that is one of the problems we all face. Is too much of what I write here speculation? That would be a fair question. Yet what actions have political parties taken to keep their national corporations safe? I am asking that question. You see, there is no top-line data from any media on that simple given part. The media seemingly doesn’t want that, Microsoft definitely does not want that and there we see a dangerous setting of ‘advertisers’ versus informing the audience. The setting that I have referred to in the past as the connected stakeholders. Yes, I could be wrong, but I have been in the IT business since 1979. I have seen a lot and I have a long memory, as such there is plenty of evidence all over the field. So why am I the only one seeing this? Yes, again, it could merely be me. However, is that the case? 

I will let you mull this over and draw your own conclusions. Enjoy the day, the week is almost over.

Leave a comment

Filed under Finance, IT, Media, Science