Re: Actually looks like a great place to work at
Huh. Tell that to Valve.
313 publicly visible posts • joined 16 Jul 2013
While I won't argue that Broadcom isn't 50 shades of awful for this move, no vendor is under any legal obligation to continue renewing contracts in perpetuity unless that is explicitly stated in their earlier contracts (which no sane vendor would ever do unless offered a truly massive pile of money). The potential costs are ruinous.
The reason that companies don't normally acquire other companies and then Broadcom them is reputation hits. Yes, people used to actually care about reputation (a few of us weirdos carry on the tradition), but now everyone wants the government to renegotiate their deals if they go sour.
It's not like Broadcom doesn't have a reputation here. They've done this before. And they'll do it again, because people keep letting them. As soon as there was even a whiff of an acquisition, anyone with any sense of sanity and / or fiduciary duty started planning a migration immediately.
The IT administrations / CIOs / politicians / whatever in charge of these organizations that didn't migrate breached their fiduciary duties. If they locked in so hard on a single vendor that they couldn't migrate in a year, then they breached their professional duties.
Not responding to the real and imminent threat of your vendor being Broadcommed is an inexcusable and unforgivable management failure.
Not being able to respond to the real and imminent threat of your vendor being Broadcommed is an inexcusable and unforgivable management failure.
VMware's customers promptly fleeing would impose costs on Broadcom an order of magnitude more than all of the potential government actions combined, but too many people are too fucking stupid / lazy / incompetent to act in their own best interests. They could have made Broadcom choke hard on that acquisition price instead of profiting from it, but nah.
...and put it in charge of the really important stuff."
Am I the only one who foresees entire categories of fun and exciting new DoS attacks against network infrastructure monitored by agentic AI?
Maybe we should let agentic AI evolve beyond the point where any sane person's response to it popping up on a website is to immediately close the window first?
I mean, I try it maybe once a month just out of morbid curiosity and it never fails to exceed my expectations for disappointment. If anything, it seems to be getting worse.
But, no, Cisco and HPE and everyone else who desperately needs to churn their customer base is perfectly happy to drive us all straight off the cliff, full speed ahead!
The whole article is "we found one whole customer who's excited," then ends with ya'll better start FOMOing harder or, you know, you might experience the same kind of regret that companies that didn't push everything into the cloud are feeling now, what with having far more control over their systems' feature stability, costs, and downtime.
With stock markets addicted to (and therefore corporate incentives tied to) infinite recurring revenue growth, what else are they going to do?
They've pushed everyone and everything into the cloud or at least into SaaS. If the industry can't find a Huge New Recurring Value Add then their stonks will crash harder than a Starship test flight.
There are few cookies left in the jar such as moving orgs back from cloud to on-prem, but in terms of increasing adding value for customers there just isn't much left in the industry outside of some smaller under-served verticals (which don't scale hugely in revenue).
Performance requirements vs. the inviolate laws of physics are pushing not just towards soldered RAM, but SOC RAM.
The possible happy medium is external CXL RAM, which will be considerably slower than integrated RAM, but nowhere near as slow as flash. This will require significant OS-level support, but since it will be a server thing we may see it trickle down to the desktops (definitely for Linux, possibly for Windows, Mac users shouldn't hold their breath).
I agree - passkeys are about half a notch better than passwords, but are hardly the panacea they're promoted as.
They can still be stolen by end-device compromise
They're still vulnerable to every reset scam because 99% of end-users can't tell the difference.
They still leave all of the session cookies and whatnot vulnerable.
In the US and most of the world, courts have decided that the consumer liability for companies with sloppy IT security is basically zero.
I mean, it's not *exactly* zero, but close enough.
If this was set to a more appropriate number, behavior would change very rapidly.
SpaceX is out there level-grinding and building up more EXP than everyone else combined.
The difference is so pronounced. NASA notes a heat shield problem, has 100 different labs analyze it. They're not even planning a real-world test before having astronauts trust their lives to whatever fix is proposed:
https://www.theregister.com/2024/12/06/nasa_orion_heatshield_investigation/
SpaceX has a heat shield problem, they create an engineering fix, slap it on the next rocket (along with deliberately degraded tiles), and perform a real-world test within a few months. The heat shield will go through tests on real spaceships doing real re-entries through real atmosphere at least a dozen more times before humans are placed at risk.
And exactly how much money goes to the victims of the breech? Zero.
Maybe €251 million ($264 million) will go to the bureaucracy, but those affected get nothing.
This will be even worse than the US case where a major bank signed up customers for services without their consent. The government hit them with a massive fine, and then 5% of that went to the people affected (about $2.50 each).
Give the money back to the victims of corporate malfeasance.
You'd be surprised how awful a lot of this stuff is. I've seen gas pump terminals attached to the Internet with cellular modems with the admin interface over telnet (!) with a four-digit PIN. And not in Ye Bad Olde Days, in the last few years. I wouldn't be shocked if there were buffer overflows, input sanitization problems, etc. by the truckload in that software.
Ideally the question shouldn't be whether the tech is perfect, but whether it is better than a typical person.
But that question doesn't work because we have, through thousands of years of cultural training, a pretty good mental model of how humans fail and we're reasonably good at (sometimes, or at least not completely crap at) designing procedures and systems that deal with human failure.
The real problem with LLM AI failure is that it's so alien to our minds that we have no instinctive or even algorithmic understanding of how to cope with its various failure modes. Therefore, it's far more difficult to slot into business processes than even fairly sketchy humans.
I don't see this changing very soon.
“By comparison, the Vega-C mission marked the 351st launch by Arianespace in a considerably longer timeframe.”
Timeline perspective: SpaceX has launched roughly as many rockets in the past six months as Arianspace has in the past fifteen years, and has a planned cadence of more launches per fortnight than Arianspace has planned per year for the rest of the decade.
Naturally, they had to add the following legal disclaimer just in case someone was planning on eating or injecting this, or shoving it into their rectum:
"⚠️ WARNING: This item may contain one or more of 900 chemicals found by the state of California to cause cancer, birth defects, or other reproductive harm. The state of California requires this warning based on chemicals used in the manufacturing and/or decorating of some products. For more information, go to www.P65Warnings.ca.gov"
One of Trump's closest confidants right now is Elon Musk, who happens to be one of the largest players in AI. Musk is a moderate safetyist, although generally friendly with the e/acc types, and his views will probably have the greatest influence on the Trump administration (until one of them blows an ego stack).
It's also worth remembering, especially for developers, that Microsoft has a horrible reputation when it comes to following through on commitments to new platforms. This is where it destroyed every chance it had in the mobile space: they'd rush something out that was horribly conceived (WinCE), then pull the plug and push out something else that was poorly conceived, and each time they did so they'd shamelessly knife early adopter users and supporting vendors in the back. By the time they had a decent mobile version of Windows, everyone willing to give them a chance had been burned too many times and nobody trusted them anymore. We could run down the entire list of Microsoft's bold new strategies that they quickly abandoned leaving their supporters in the lurch (Zune and Fairplay come instantly to mind), but I'd probably hit a limit for comment length.
When Apple moved MacOS to ARM, they'd already been running a stripped-down version on iPhones and iPads for well over a decade and they knew exactly what they were doing. The key parts of the software stack and APIs had been dual-architecture for a very long time and were already mature and optimized. The hardware was bulletproof. Because Apple has a long history of sticking to its guns on major evolutions, vendors knew they weren't kidding and so anyone committed to the platform knew they had to be on board. The other key thing Apple did was provide emulation that worked extremely well. Even most action games were highly playable on the new hardware. It wasn't absolutely perfect; emulation never is. But showstopping glitches and even minor annoyances were reasonably few and far between.
Microsoft will never, ever hit this level of execution without a top-to-bottom corporate overhaul that obliterates their dysfunctional culture and that will probably never, ever happen. Windows users can expect this transition to remain a trash fire, assuming Microsoft doesn't decide to abandon it and the people who bought into it as Microsoft is wont to do.
Yes, there are many ways to back up your photos / videos / whatnot from your iDevices. I use Synology's private cloud solutions as primary and the photos app on my Mac as secondary, and there are plenty of other options.
Where Apple loses me is that when your iDevice starts running low on storage it will push you hard towards iCloud and if you're not careful in your clicks then iCloud it is. I agree that Tim Cook is the least creepy uncle in the cloud, but since I choose no creepy uncle I wish he'd fsck off as well.
And what happens when you accidentally choose the iCloud that Apple is constantly trying to ram down your throat? Turning it off deletes the files (and / or calendar / contact entries / notes / etc) from your device. What? Why? Who knows? Who cares? Yes, you can download them and then replace them, but this is a stupid amount of work.
It's not that Apple very aggressively pushes iCloud. This is annoying, but no worse than any of the other horrible vendors out there. Trapping customers like this is unethical and just being a bunch of petulant assholes. Apple used to be above that sort of thing, but the company has been running downhill since Cook took over.
If Microsoft pushes the install and the end user doesn't accept even a click-through EULA, I wonder how Microsoft expects to collect on this.
If Microsoft blocks functionality based on an action they unilaterally initiated, can the end-user sue?
An interesting legal problem.
...Qualcomm can just spin up a bunch of RISC-V cores that are optimized to anywhere remotely near where their ARM cores are like it's no big thing. That's probably a five-year, multi-billion-dollar road, assuming you have the personnel and resources to divert to it. This is roughly how long it took Apple to get to their first fully-built-from-scratch ARM cores, and their current chips have more than a decade of refinement past that. Every micro-architecture engineer you move to RISC-V has to be taken from somewhere else, and they don't exactly grow on trees. The elite ones are so rare that if you want them you just buy the company they work for (as both Apple and Qualcomm have done).
In the meantime, Qualcomm either has no products to sell or they (likely) still have to cough up the Arm money.
Considering that Qualcomm is notorious as being one of the most abusive chip vendors in terms of licensing agreements (they are to hardware what Oracle is to software), and they're well-known for treating their engineers like disposable garbage, I'm just going to enjoy watching them reap some richly-deserved karma.
First, AT&T is one of the most horribly abusive oligopoly telecoms in the world, so this is a mere fraction of the karma they're due. There's a reason their logo resembles the Death Star.
Second, as soon as Broadcom started sniffing around VMware we knew where this was possibly heading. It's not like Broadcom doesn't have a long and (in)glorious reputation here going back well over a decade. Any CIOs and IT directors that weren't at least checking the condition of the lifeboats on this soon-to-be-sinking ship are getting what they deserve.
In a well-run organization, this is a legitimate concern.
In a poorly-run organization, moving as much technology as possible out from under their alleged leadership can be a plus, even if the place it moves to is something many of us scoff at. And many organizations are poorly-run.
We were using them for endpoint management when Apple was making the final migration to 64-bit-only.
Kaseya didn't have a 64-bit MacOS monitoring client despite being given, oh, about five years notice that 32-bit mode was deprecated and going away. Anyone who did their normal MacOS version update (this is nowhere near the nail-biter that Windows version updates are) lost management capabilities, and this took about two months for them to resolve. Keep in mind that MacOS had been 64-bit for a very long time at this point; they weren't doing the Microsoft-esque separate 32-bit and 64-bit versions. It was only 64-bit with backwards-compatibility mode that was finally being stripped out.
According to Kaseya, this occurred because they were dealing with "other development priorities." For five years. This is not a serious company.
From the lawsuit:
"7. Additionally, the support services enable tens of thousands of agents at various
AT&T customer contact centers to assist roughly 1,000,000 AT&T customers every day with
their communication needs. Without the support services, it is not a question of if but when the
software will crash due to a software error, security issue, or lack of upgrades and maintenance.
When that happens, these 1,000,000 daily customers will find it significantly more difficult to
resolve issues with their accounts because thousands of fewer customer contact agents will be
available to assist them."
Having managed AT&T accounts with fairly significant monthly and annual spending figures, I'm darkly curious to find out how this could possibly be made worse.
And by lean, I don't mean cheap. I mean efficient. There's a difference. You could probably drop the management / administratvie headcount of most large companies by 75% and run them better. If you look at the ROI for the individual activities that many people are doing, it's outright negative: compiling TPS reports that nobody reads, etc. If you've sat through enough executive meetings you'll realize that at the end of the day they either wind up winging it or bowing to some milquetoast consensus that provides minimum personal career risk to everyone in the room - unless they're fortunate to have a badass or two with strong enough personalities to actually do the leadership thing *and* do it reasonably well. The badasses are driven by passion and commitment to actual product / service excellence as opposed to the words written in some statement of corporate values that produced mountains of infinitely detailed spreadsheets badly explaining poorly-gathered information. When you talk to the badasses they're usually tell you they're ignoring everything they learned at university. I haven't met one who said they were so thankful for their MBA, assuming they have one (in my field they usually start out as engineers who happened to have some extra talents).
I'm to the point where I think we should just scrap university degrees for business and start doing paid internships instead, with trade school-style classes for mathematics, writing, and whatever other ancillary subjects individuals might benefit from (or they can watch online videos or whatever works best for them).
Are there any CPU manufacturers other than Intel and Samsung that fab their own chips now?
Apple is probably the premier CPU company right now in terms of overall efficiency per watt. They contracts to TSMC and has probably considered doing it - they're one of the few comanies in the world who have enough money burning holes in their corporate pockets - but probably figures that it's not worth the hassle unless TSMC starts screwing up as much as Intel did.
AMD also contracts to TSMC, as does Nvidia.
Fabless is the rule now, not the exception.
To me the on-prem vs. cloud argument is a sideshow next to the real argument: control over the stability of your IT assets and infrastructure. If an organization's IT department is a rancid dumpster fire (and many are) there may be few downsides to losing this control. But for those of us who run and have run tightly-managed IT ships and put up strong availability numbers abandoning on-prem (which this is pushing) demands that we shave at least two 9's of availability off for the greater good of Microsoft's stock P/E ratio.
And it's not like it saves us significant hassle. Managing Microsoft365 + Azure is arguably worse than managing on-prem systems, and the full scale of it hits you square in the face even for SMB environments because they aggressively opt your users into everything unless you manage to slap their digital hands away first. No thanks. I've been conscientiously dumping Microsoft for years as they are simply no longer a foundational requirement for almost anything anymore. Even with the far less-mature management tools, we manage to do far better with Macs for desktops and laptops. Life is not perfect, but it's good. I no longer bite my nails every time we roll out security updates. Financially, we still come out ahead paying for the far more expensive machines because of the lower management and support overhead. Plus it makes a lot of the users happier, and that's a big thing when you actually see it happen - it makes the whole workplace better. I've been dreaming for many, many years of a viable corporate Linux desktop, but I still think it's a bridge too far. I'm all-in on Linux servers and especially NAS systems, which scale down the management requirements tremendously and have almost microscopic attack surfaces by comparison.