Chris McGee (@seeteegee) header
Chris McGee (@seeteegee) profile picture

Dad, Cynical Geek, Tech and Software Enthusiast. Building namespaces and cutting teeth on edge computing. My thoughts come from my own distorted brain.

Ottawa, Ontario
Joined March 2010
Photo by Chris McGee (@seeteegee)Photo by Chris McGee (@seeteegee)Photo by Chris McGee (@seeteegee)Photo by Chris McGee (@seeteegee)Photo by Chris McGee (@seeteegee)Photo by Chris McGee (@seeteegee)Photo by Chris McGee (@seeteegee)
Chris McGee retweeted
Unpopular opinion: `--foo value` in CLI options is harmful. It should be (--)foo=value (equal sign required!) so that app wrappers have an easier (not to mention less brittle) time determining where options end and arguments start.
16
5
3
144
7,616
This one looks legitimate. It’s x86 specific and is a consequence of a bug in the JIT. Long story short, the java.util.Base64 decoder has an AVX-512 (SIMD) routine swapped in at runtime to speed up the decoding. Except there’s a bug in the routine which could lead to memory corruption. The primary concern here is that Base64 is used in a number of application servers for HTTP basic authentication, Kerberos, SAML, WebSockets, etc. You might not even use these features, but if the server decodes the string before rejecting it, it may be too late. Until a fix is in place the workaround is to add the flag -XX:-UseBASE64Intrinsics to your application startup to disable the AVX-512 routine.
🚨 BREAKING: A critical Java flaw that reportedly lets hackers hijack servers without a password leaked before its patch. Security firm Calif says it found the bug, dubbed Java4Shell, and will publish details today. Oracle fixes 11 Java flaws today but hasn't tied any to it.
10
44
5
430
63,271
Disappointed to see this become official as I had heard rumours months ago that Ontario’s SMR project was falling behind because of engineering delays at GE Hitachi. There were always risks associated with being a first-mover, and OPG/Ontario knew that. The commercial case for small reactors only makes sense if capital costs and timelines to build the next reactor fall a lot as a result of reproducibility. Otherwise, we may as well be building large reactors. GE could very well complete the project, but the difficulties and delays could render the BWRX300 commercially unviable for future sites. I hope they see the bigger picture. It also reaffirms my belief we should focused on developing Canadian nuclear technology should we ever do first-mover projects in the future. All the risk and the IP remains American? Nah.
OPG update on the BWRX300: "DNNP Unit 1 is no longer expected to achieve the previously disclosed commercial operation date target of October 2030 or be completed within the previously disclosed estimated cost of $7.7 billion"
12
5
1
106
6,498
Chris McGee retweeted
OPG Acknowledges the West’s First SMR is Behind Schedule Ontario Power Generation has acknowledged that “DNNP Unit 1 is no longer expected to achieve the previously disclosed commercial operation date target of October 2030 or be completed within the previously disclosed estimated cost of $7.7 billion.” OPG explicitly identifies unfinished engineering by GE Vernova Hitachi as the primary contributing factor, going on to say... “This reflects GE Vernova Hitachi Nuclear Energy's ongoing engineering for its first-of-a-kind SMR, which has impacts on construction execution and supply chain performance.” Detailed engineering is falling behind construction, causing costly delays, rework and lost productivity. These pressures are compounded by the departure of several senior engineers at a critical stage of the project, when GVH’s already stretched resources are needed most. Meanwhile, GVH is pursuing an expanding portfolio of projects, including Clinch River in Tennessee, where TVA aims to begin construction with a design more than 95% complete and benefit directly from lessons learned at Darlington. Ontario is thus absorbing the risks of pioneering the first BWRX-300 while the vendor prepares to deliver its successors. Having committed billions to pioneering an American reactor design, Ontario now depends on a foreign vendor whose scarce engineering resources are being pulled in competing directions. Ontario’s immediate priority must be getting Darlington back on track. GE Hitachi needs to hear an unequivocal message from Queen’s Park and Ottawa: completing the engineering required to deliver Canada’s first BWRX-300 must take precedence over the company’s expanding international ambitions. Canada is hardly powerless. OPG’s Integrated Project Agreement provides broad audit rights, performance-based financial incentives and contractual protections where project partners fail to meet the required standard of care. Ontario also retains substantial commercial leverage through three additional planned BWRX-300 reactors that have yet to reach final investment decision. The province should use this leverage to demand an independently verified engineering recovery plan, backed by dedicated senior personnel, binding delivery milestones and transparent reporting. Where contractual obligations have been breached, OPG should pursue the remedies available under its agreements. The larger lesson concerns the next generation of Canadian nuclear construction. Canada has deployed the full weight of its public balance sheets to establish the BWRX-300, an American reactor design, as an international commercial product. We cannot afford to repeat this naive arrangement with the Westinghouse Electric Company AP1000, committing billions of dollars and decades of energy planning to another foreign vendor whose scarce engineering resources and strategic priorities ultimately lie elsewhere. It is time to be Maîtres chez nous. Canada’s next reactors must be CANDU.
17
25
10
147
12,218
Chris McGee retweeted
Canada’s Real-Time Rail (RTR) is a new national payment infrastructure owned and operated by Payments Canada. It allows participating financial institutions, credit unions, and payment service providers to send and receive account-to-account payments that clear and settle in seconds, 24 hours a day, 365 days a year. It launched on November 9, 2026. How it differs from Interac e-Transfer Many people confuse the two, but they are not the same: •Interac e-Transfer is a consumer-facing service. You send money using someone’s email address or phone number. It feels fast for users, but historically the funds did not fully clear and settle between banks in true real time. •Real-Time Rail is the underlying national “rail” (infrastructure). It handles the actual exchange, clearing, and settlement of the money in real time between financial institutions. e-Transfer is expected to begin migrating its clearing and settlement onto the RTR in 2027. The familiar e-Transfer experience will largely remain, but it will eventually ride on the new infrastructure. Key features •True real-time settlement: Funds are available to the recipient almost immediately, and payments are generally final and irrevocable once settled. •Always on: Works evenings, weekends, and holidays with no cut-off times. •Data-rich payments: Uses the ISO 20022 messaging standard, so more structured information (like invoice numbers) can travel with the payment. •Built-in fraud tools: Includes centralized fraud services (such as confirmation of payee and real-time risk scoring) from day one. Canada is the first G7 country to launch a real-time system with mandatory centralized fraud controls. •Higher limits: System limit starts at $100,000 (financial institutions can set lower customer limits). Current status and rollout The system is live, but access is phased. Some smaller institutions and fintechs are already connected; the major banks are coming online later. The goal is for over 90% of Canadian bank accounts to have access by the third quarter of 2027. In short, the Real-Time Rail is the modern backbone that makes instant, final, data-rich payments possible across Canada. It is not a new app or a replacement for e-Transfer—it is the national infrastructure underneath (and eventually supporting) faster money movement.
Clearly, most people have zero understanding of what Real Time Rail is and how it’s not the same thing as e-transfer. If you can not distinguish the two clearly, you really have no business commenting on economic or financial matters.
132
147
28
1,017
113,242
Chris McGee retweeted
Microsoft chased using a DB as a filesystem for about 20 years. It likely all started with OFS around the Cairo timeframe. OFS was conceived as an installable filesystem that would: - Store OLE objects (not just raw files) with rich, application-accessible metadata. - Support multiple data streams per object, long filenames, ACLs, quotas, and journaling. - Index both content and metadata so items could be searched and queried across applications (not just by name or path). - Optimize for large numbers of small files. The broader Cairo vision included “Smart Folders,” representing non-file objects (printers, devices, etc.) in the same hierarchy, and fulfilling Bill Gates’s “information at your fingertips” idea. Public demos appeared around the 1993 PDC. Then of course there was Storage+ and RFS and WinFS. Even ReFS uses a B-tree database internally.
It’s so funny that SQLite is genuinely faster than most traditional filesystems. I wish there were more experimental operating systems using dbs for the backend. Not many people playing around in this space...
27
32
2
520
109,712
Chris McGee retweeted
The amount of Microsoft energy that went into trying out this theory—a database running locally as a file system—would be hard to calculate across Windows, Office, email. It is one of those ideas that seems interesting on paper. In practice it is never good enough for everything you want to do and the constraints do not help you to think of better or more clever ways to get things done. The constraints just got in the way of getting anything done. There's significant utility in having a local database available for use in the platform, but that too carries with it challenges in backup/restore/upgrade/security/privacy/etc. In hardcoresoftware.substack.co… I think I wrote about this adventure in four times (working on strategy for Bill, Outlook, SharePoint, Longhorn.) I don't know if there is another example of a technology aspiration that endured so long and lacked a solution that worked. And also caused so much division and consternation inside Microsoft. Perhaps more than anything it is the lesson of a big company thinking it can invent the future in isolation—all sorts of isolation or tunnel vision. Big companies, it turns out have a surplus of two things: 1. Hubris. There was no shortage of people starting at the top who become "obsessed" with delivering on an idea and believed the success Microsoft had recently achieved was a license to know how to build it. 2. Doubt. There was also no shortage of people who thought from the first days that the idea made no sense and believed the defined solution in no way benefitted the products that are being made, loved by customers, and pay the bills. This 1-2 punch is fascinating because it sets up a divisive dynamic where you are either an "innovator" who is "on board" with the future OR you are a "obstacle" who does not see the value in "big bold bets" on the future of the company. Along comes a book like "Innovator's Dilemma" [1998] and suddenly resisting the next big corporate strategy makes one throw out common sense engineering because you don't want to get disrupted. It is the other side of the disruption coin no one talks about—sometimes the new thing does not in fact work very well nor will it in any finite time. The lesson applies to the product and technology leadership of a company. Part of what happens in a big company as well is the ability for the company to speak broadly and enroll their ecosystem in their own excitement for the future. Microsoft described WinFS as early as 1992 and over the course of more than a decade it polluted our ability to talk about new products. The email product (Exchange, today the mail for 365) was famously the most important product to be launched and for the first decade of its life was always about to become strategic even while it was in fact the most strategic product the company made. The strategy nearly wrecked Outlook the face of the email strategy. It did in fact derail an entire generation of Windows. There was fallout across a range of products never released. All the while Microsoft was talking about the grand future of this product while customers were investing and buying into products built an entirely different way. A dynamic few are willing to speak about is another 1-2 punch: • Customers and the press talk all about waiting and seeing if the current product will adopt the all important strategy. The first signs of this are the sales force freaking out and escalating their concerns. • Customers actually ignore the strategy because "product market fit" dynamics take over and they can't buy and adopt the actual product soon enough. The first signs of this are visible to those in the real world who never get asked about the strategy because customer have a million feature requests. This is the second way the whole dynamic divides the company. The "obstacle" people start to look like they are focused on the past to the "innovators" who are clearly focused on the "future." You can find many examples of this near computing technology and broadly in other industries: • Intel Itanium (which Dave below knows well) was a chipset architecture that just never panned out. Itanium occupied Intel for more than two decades. • Copeland OS pursued by Apple muddled their decaying ecosystem and divided the company, cancelling it set the stage for Jobs return with NeXT becoming the future OS. • IBM "Stretch AND "Future Systems." IBM was such a defining company it had two generations of this dynamic in an effort to reinvent (or invent!) mainframes. Stretch was much like Microsoft Longhorn in that there were portions that went on to greatness in the 1960s. FS just got cancelled. • IBM also had the Micro Channel Architecture (MCA) which divided the hardware side of PCs at just the wrong moment. It just went away. Also IBM had OS/2 which I think still runs some bank ATMs :-) [There's a theme that maybe IBM invented this whole level of hubris which would follow from second-system-syndrome in "Mythical Man Month"] • Hydrogen Fuel cars have been surpassed by improved gas and now EV yet you can still see them being pushed in Japan though at last Detroit moved on. This one has been going on since I was a kid. • GE Predix was a broad SAP-like industrial information technology bet that was the technical undoing of GE IMO, though the 2008 financial crisis really did the work. Predix was Immelt's bet (his book details this) while cleaning up was left to a successor as often happens. There's a difference between a failed product and a failed all-consuming strategy. All companies have failed products essentially all the time. Technology companies seem unique at hubris required to broadly pursue—in public—a direction that doesn't work. They do so under the guise of preventing the decline of the company that makes it all the more painful. Likewise there is a difference between a successful product line that can't escape its own success gravity and failed attempts to reinvent it. The most common thread is the inability of a successful company to find a new strategy for an existing product. We are living in this moment now with AI as we see Intel which could not transform its chip architecture into something different while at the same time Nvidia's long term architectural bet ended up as the foundation of the the AI era. Had we found ourselves in some other chip architecture from a third company it might very well be that we'd be talking about Nvidia's GPU bet as a strategy initiative they just would not give up on. One of the things that happens over time is this corporate hubris can get rewritten into "we had the right idea but were just early." The problem with this is it separates the broad concept at 30,000 feet from the actual technology and implementation. I've watched this happen with everything from Clippy to TabletPC to Windows Phone. There's a huge difference between a concept and an implementation that makes the claims of being right. I love Clippy but there was no path where it went from what we did in 1995 to today's AI agents, any more than the clipboard Captain Kirk used was a product precursor to an iPad. And of course as always with product development, no matter what happens...someone always said it would. Which makes me sharing all the times I thought WInFS would fail all the less satisfying for me. Here are the places in Hardcore Software where WInFS was part of key decision making and controversy. 024. Discovering “Cornell is WIRED!” [1994] hardcoresoftware.learningbys… 053. Strategy Tax: Outlook Storage, First Attempt hardcoresoftware.learningbys… 065. SharePoint: Office Builds Our Own Server hardcoresoftware.learningbys… 069. Mega-Scale, Mega-Complexity hardcoresoftware.learningbys… 074. Outlook Pride, Finally hardcoresoftware.learningbys… 085. The Memo (Part 1) hardcoresoftware.learningbys… 091. Cleaning Up Longhorn and Vista [2006] hardcoresoftware.learningbys…
Microsoft chased using a DB as a filesystem for about 20 years. It likely all started with OFS around the Cairo timeframe. OFS was conceived as an installable filesystem that would: - Store OLE objects (not just raw files) with rich, application-accessible metadata. - Support multiple data streams per object, long filenames, ACLs, quotas, and journaling. - Index both content and metadata so items could be searched and queried across applications (not just by name or path). - Optimize for large numbers of small files. The broader Cairo vision included “Smart Folders,” representing non-file objects (printers, devices, etc.) in the same hierarchy, and fulfilling Bill Gates’s “information at your fingertips” idea. Public demos appeared around the 1993 PDC. Then of course there was Storage+ and RFS and WinFS. Even ReFS uses a B-tree database internally.
37
52
9
594
71,862
SQLite is not faster at being a filesystem. It’s faster at storing a bunch of blobs because that’s a very different task from being a filesystem. Database FileSystems were an area of heavy research in the 2000s with Mac and Windows both putting forward their solutions. IMHO, they both futzed it up. The Mac version wasn’t bad under the covers, but it exposed very little of the potential benefits. If anyone is interested in restarting this line of research, the source for the original ideas was BeFS. The author of that filesystem (Dominic Giampalo) wrote a book on the subject and later made it free to download from his website: nobius.org/dbg/practical-fil…
It’s so funny that SQLite is genuinely faster than most traditional filesystems. I wish there were more experimental operating systems using dbs for the backend. Not many people playing around in this space...
29
58
11
939
66,281
Chris McGee retweeted
Somewhere, in a universe next to ours, a man who looks very much like Patrick Stewart wakes up from a long strange dream and begins playing the flute.
87
805
49
7,651
86,304
Replying to @QiaochuYuan
ok my controversial, possibly morbid take on this... at least when it comes to the mushrooms... is that perhaps the kingdom of fungi is playing a long game to tenderize humanity? they are organic matter decomposers, which makes us their food. so they evolved to give us happy peaceful psychedelic experiences that help us embrace our own mortality in order to... make more tasty food for themselves. kinda like how when you're hunting, you want a clean shot so the prey doesn't get stressed / the meat quality is better.
7
3
44
941
What a heartbreaking week it has been. Sir Patrick Stewart was a friend, a colleague, and a magnificent actor. From the stage to the bridge of the Enterprise, he brought intelligence, grace, and a deep sense of humanity to everything he did. Captain Picard stood for the best of us, and so did the man who played him. I was filled with affection and admiration for him and his work. My heart goes out to Sunny and his family. He will be deeply missed.
2,493
25,560
817
188,694
2,788,632
Mark Carney say he wants Canada to become an associate member of the EU by June 2027 #cdnpoli Here are some of the concrete examples Canada and the EU gave as what that would actually mean - Canada joining Erasmus+, the students and youth exchange program - They signed a digital trade agreement, that Carney says will make payments between Canada and EU countries easier - They are exploring opening the banking, insurance and fintech sectors up. - Canada is negotiating to join Galileo, which is a global navigation satellite system - They are discussing potential new transatlantic and potentially transarctic data cables, as per Carney
217
199
61
1,191
176,533
Chris McGee retweeted
Canada is going to Eurovision 2027. It’s up to you to decide who takes the stage. Cast your vote by November 5. cbc.ca/eurovision
2,697
816
804
11,591
1,242,741
Chris McGee retweeted
In September 1992, Unicode had a problem on Unix. Storing characters as 16-bit values could introduce zero bytes that C programs mistook for string terminators. Earlier encodings had compatibility problems of their own. One evening, Ken Thompson and Rob Pike went to a diner in New Jersey. Thompson sketched a new encoding on a paper placemat. ASCII stayed exactly as it was. Other characters used multiple bytes, without accidentally introducing null bytes or Unix path separators. They went back to Bell Labs and started coding. By Friday, their Plan 9 operating system was running entirely on the new encoding. They called it UTF-8. Today, it powers over 99% of websites with known character encodings.
34
156
28
1,602
76,438
Chris McGee retweeted
Nice wind lull going on. If you follow the white Ontario demand line, we never stop running gas, so we are charging batteries with gas and then discharging that gas-generated energy during the peaks. The emissions savings of gas-powered batteries? 🤷🏻‍♂️ 😂
Rolling 14-day Ontario power mix🦇🎃🧙♀️👻:
3
8
2
15
691
Chris McGee retweeted
Remember when Mike Harris sold off Highway 407 for about a three billion and we all got a $200 cheque? That highway is now worth over 30 billion. Another boneheaded move by a Conservative leader. We need to protect our assets.
Toronto debate: Bradford just promised to reduce senior’s water bill by $300! How’s he going to do that?
22
337
2
1,800
14,684
Chris McGee retweeted
"We have essentially doubled back to the computer architecture of the 1960s through the 1980s, a period dominated by the mainframe and the so-called 'dumb terminal'. In a bizarrely ironic twist, generative AI has reverted us back to a computing system from 50 years ago." pcmag.com/opinions/welcome-b…
2
4
21
2,049
“I only like two oceans: the Atlantic Ocean and the Pacific Ocean. I’m not going to put Trump or America on any other ocean but these two.“ Yet another friendly reminder that the sitting President of the United States is insane.
47
262
10
1,820
36,581
Chris McGee retweeted
Full details of a remotely exploitable vulnerability in the core engine of *all* PHP versions will be released tomorrow, October 21st. If your organization has internet facing PHP applications then you need to be ready to patch the minute the update is released or *strongly* consider taking the application offline until you can: php.net/index.php#2026-10-20…
🚨 PHP warns of two critical vulnerabilities as emergency patches arrive tomorrow The PHP security team will release updates on October 21 at 14:00 UTC for two vulnerabilities, each rated 9.2 under CVSS v4.0. ⠀ Security researchers identified the flaws as CVE-2026-109618 and CVE-2026-109619 and described them as remote code execution vulnerabilities reportedly affecting PHP versions 5 through 8. PHP has not yet confirmed the affected versions, exploitation requirements or fixed releases. ⠀ The vulnerabilities were reportedly leaked before their planned disclosure. Researchers said they have not observed active exploitation, while PHP warned it may release the patches early if technical details become public or exploitation is detected. ⠀ Administrators should identify PHP installations now and prepare to apply the updates as soon as they become available.
81
563
165
2,964
1,030,532