I know I’m supposed to do this post in December, but here’s a better-late-than-never 2019 retrospective for this blog. It was my most prolific year of blogging since 2012, with 11 posts comprised of 12,103 words (25+ pages), and set new records for traffic (and hopefully readership).
Key takeaways for me: clicks very much follow a power law, with the posts on IBM/Red Hat and open source vs. the cloud accounting for a whopping 84% of traffic. Hacker News and Twitter were the top two sources of traffic. Fake press releases don’t get the attention they deserve, so I will continue to flog this format until it is fully appreciated (take note!). And as a purveyor of densely linked posts, it is disappointing how few click-throughs there are of all those links.
To review 2019’s posts:
This was probably my
post of the year, though I don’t seem to be accorded the same audience when I pop off about geopolitics that seemingly every political scientist gets when they talk authoritatively about the impact of AI. The conflict with China is much more than a trade war or nihilistic Trump thing: it is a fundamental geopolitical and ideological confrontation. The modus operandi of the last couple decades both in China and with China isn’t coming back. The tech industry still seems to struggle to understand the inflection points that have happened around China’s economic model, authoritarian leadership, and relationship with the rest of the world. I never would have guessed that after Hong Kong, Taiwan, the South China Sea, East Turkestan (Xianjiang), and Huawei, it would be an NBA tweet that would wake up the West that you periodically may have to go to bat for your values at the expense of TAM. South Park absolutely nails this issue, down to their apology. I got more traffic from America Samoa than China last year, so will assume I am a proud member of the Great Firewall blacklist. Their loss.
This post updates CAPEX spending by the big cloud providers through 2018, and an update with 2019 numbers is coming soon, I’ve been following CAPEX as a key tell for cloud success for over a decade, and it has proven to be a great way to separate the clouds and the clowns. Speaking of CAPEX clowns, both IBM (with the pivot to “hybrid cloud”) and Oracle gave up pretending in 2019. There are some other CAPEX-related topics I also want to explore this year.
This was a quickie post in response to the assertion that “open source eats the cloud”. I stretched the Y-axis for dramatic effect. The later “Dining Preferences of the Cloud and Open Source” post addressed this topic far more substantively.
This is a recurring theme I recycle every time the Fed gets a new Chair that has an audience of one: me. It is notable this idea has gone from shades of Skynet trying to wipe out humanity via quantitative easing to looking measured compared to the humans in the mix.
This post was probably the least satisfying in terms of its execution. It started well, and I really like the first eight paragraphs, but the back half was flaccid and rambling (and I got called on it by Dave Winer, who restated it far more succinctly).
This was my most widely read post of the year and perhaps coincidentally the longest. I put a link at the end asking people who made it that far to click, which less than a third of readers did. This post made it to #2 on Hacker News.
This is another recurring post that I pretty much mail in at this point every time Gartner drops the annual magic quadrant for cloud. Your cloud computing landlord is in Seattle: please make sure you get your rent in on time.
This was the snarky fake press release take on IBM’s acquisition of Red Hat, playing off IBM selling once-powerful software franchises like Notes to HCL for a song. We’ll revisit this in 2022 and see how IBM does with Red Hat. They’re off to a strong start to killing Red Hat by removing Docker from RHEL, which is already endangered in the cloud world, and replacing it with their own, proprietary container technology. But I”m sure IBM will tell everyone how open they are (and try to sell them a “blockchain”).
I was early to this idea (in September), and political advertising became a major miasma topic after Twitter announced they would not accept political ads at the end of October. But I”m inclined to agree with Stratechery that limiting political speech creates more problems than it solves, though that won’t stop me (or anyone else) from scoring cheap and easy points at Facebook’s expense.
I got less grief than I expected for my open source heresy in this post. Fortunately, people saved their grief for the horrific piece the New York Times later wrote on the same topic (and it seems pretty obvious who hand-fed them that story). This was my second most read and second longest post, so confirmation grows for my belief there is a correlation between popularity and length. This one made it to #4 on Hacker News.
After seeing one-too-many serious and self-serving VC posts touting their consumer portfolio companies, I was triggered. The world needs more B2B and esoteric infrastructure gift options.