Ask Paul: May 2 (Premium)

Happy Friday! This has been a momentous week, for our industry, but also for me. So let’s get the weekend started early: We fly home from Mexico City in the morning and there’s still so much to do.

? Yet another subscription service

helix2301 asks:

Paul there has been a lot talk about different subscriptions on site this week due to pocketcast drama. I know myself I’m still grandfather in pocketcast. I’m alpha member on thurrott still grandfathered in at $42 dollars year. My question is have you ever thought of changing your pricing models. I know you offer a few different pricing models. Any thoughts changing site and moving to something like memberful just asking not saying the site should change. Memberful would made it easier if it was around when thurrott launched remember all the crazyness with the premium FRD in the beginning and feed not working. Good times 🙂

I haven’t looked at Memberful specifically, but yes, I’ve been thinking a lot about the future and what this might look like. I can’t claim to have specific plans here, but I feel like continued change is inevitable.

To start at a high level, I’m sure I’ve written at some point about retirement and how this isn’t really a concept that applies to me. I’m a writer, not a coal miner, so I should be able to write for as long as I want, and I feel like being engaged in this fashion is healthy long-term. I’m not going to sit on a chair on the beach with a drink in my hand. But I may sometimes sit on a chair on the beach with a laptop and get some work done. That’s the nice thing about this job. I can do it from anywhere.

Lower-level, when we transitioned Thurrott.com from George to me (or, for legal reasons, to a new company called Thurrott LLC that is basically me), I made a number of changes tied to streamlining the backend we have, which lowered the cost of doing business and made it more viable as a standalone operation. Since then, I’ve done nothing in regard to marketing or sales or whatever and have pretty much coasted in a business sense. Not out of laziness, but because I’m fundamentally just a writer and not a business person, and this isn’t where my head is at.

This year, we’ve endured two episodes that speak to a common dynamic in business in which one typically reacts to problems as they occur rather than somehow magically predicting them and working to prevent them ahead of time: YouTube without warning or good reason took away our access to the service back in January, and then we had the fun little site outage yesterday and were down for the better part of a day for the first time in, I think, years. Years ago, I suffered some data loss on a PC and observed that people typically get backup religion when they lose data. These two events were reminders of that.

In the case of the site outage, we lost about a day and a half’s worth of uploaded content (articles and perhaps forum posts), and then there was a related issue in which comments–which are in OpenWeb and thus not stored with the rest of the site–were misaligned with now-missing articles that were reposted but had different underlying article IDs. This is semi-acceptable, not a disaster. But Robert, who is a genius I trust explicitly, correctly decided to increase our rate of backups so that we have less manual work to do in the future should something like this ever happen again. (What that something was remains a bit of a mystery, but there is a lot of finger pointing at “upstream” infrastructure levels, and this apparently impacted multiple sites/machines. I feel like I’m never going to get the full picture there.)

In the case of the YouTube lockout, I got lucky and got back into the service, and I subsequently downloaded all of my content in there, for now, to a local hard disk and soon to the NAS(es) I’ll be getting this summer. And as I do with other things, I now have backing up YouTube on my to-do list for quarterly activities just in case. That’s what happens: You get a scare, and you fix things as much as you can.

As I look to the future, it’s about simplifying further and, where possible, saving more money in doing so. And this is where I don’t have specific ideas but do have some rough long-term goals. I mentioned retirement up front because this is the conversation most would have had in the past, but given the nature of my life and the way the world’s gone, I’ve always framed it like, people like me don’t get to retire, but I would like to do less over time. And “doing less,” to me, starts with minimizing all the business backend stuff. I can’t stand it, it’s always something that interrupts what it is that I want to do. But it’s also a business. And so I can’t just ignore it.

What I can do is simplify it. And I have looked into what that might look like. This isn’t something that happens today or this year. It’s just back of my mind stuff right now. But my wife and I have had such a positive experience with Substack, which we use for Eternal Spring, that it’s only logical that I would consider using such a thing, or perhaps something similar like Ghost, for Thurrott.com as well. This is not obvious or easy, there are complications like Premium memberships and comments, and just migrating content will be a nightmare. But the cost savings would be enormous. And I feel like Thurrott.com would be “better” in meaningful ways, in the sense that it would be easier for me to not deal with backend issues and focus on writing, and that segregating content between those who pay and those who do not would be easier too. So it’s a thought.

That said, it also moves from the back of my mind to the front of my mind when I’m sitting there all day long yesterday wondering why the f#$k my website is still down. I have to remind myself how unusual this is, but I also balance that with the understanding that this should never happen. So I don’t know.

To actually answer your question, yes, I think about changing the site and how memberships and access to certain content works all the time. I go through spats where I actively investigate this, and see what it would look like to make this major change. It’s daunting. In many ways, I like what we have now and don’t want to upset that unless I’m forced to.

I do feel like that’s how it will happen because that’s how life works. I never wanted to leave BWW, just as I never wanted to leave Penton before that, but sometimes life has other plans, and then you look back and wonder why you didn’t do these things sooner. But whatever. My goal is to keep doing what I’m doing for as long as I can, and not throwing up roadblocks between me, the content, and readers. And simplifying is a key way to make that possible. I don’t have specific plans, let alone a schedule. I’m open to other ideas. And we’ll see what happens.

? Speaking directly

gg1 asks:

We haven’t had a new version of DirectX since 12.2 in 2021, with no announcement that I can see of a successor or new interim release in any form. Why do you think we’ve now gone the longest without a new version of DirectX? Has Microsoft moved away from DirectX versions or are other market dynamics at play?

I hadn’t really thought about this, though as I read this, I tapped Windows key + R on the laptop, typed dxdiag and hit Enter, in a random example of latent memory. I wish I could remember the source, but someone at Microsoft once observed that DirectX was the last “pure” Windows API in the sense that it’s basically native and has been updated continuously over many years. I know it was added to the Windows SDK at some point–Windows 8, I just rediscovered–and that it can be used for apps, too. But also that there are interim releases not tied to specific Windows versions, similar to the work Microsoft did with the Windows App SDK.

Anyway. On the surface, it is kind of odd that we’ve not seen a new version of DirectX in so long. Perhaps this is aligned with console, and if you consider how much effort Microsoft has put into PC gaming in recent years, it’s impossible to think they’re ignoring something important here. (DirectStorage came out of console and is now part of DirectX for PCs, for example, and I think that’s an interesting example for future work because of its device-like requirements.)

So I looked, specifically to see how and when ray tracing, as an obvious example, was added to the platform. And it was a while ago, actually: DirectX Raytracing (DXR) was added to the platform in 2018, and it was just updated to version 1.2 a month ago. It may just be that these technologies are improved, and that the bigger push right now in this space is tied to portable gaming PCs (Steam Deck-style) and Windows 11 on Arm/Snapdragon X, and that the push there is more about AutoSR (Super Resolution) and platform compatibility. Traditional gaming-class PCs are pretty well understood at this point.

That DXR 1.2 announcement links to a DirectX State of the Union speech from GDC 2025 that is likewise very recent, and that talk addresses the work that team has done in the past year, and it’s all pretty advanced stuff, with mesh shader-based graphics pipelines, etc. But among the topics in there is DirectSR, which is a “platform-agnostic GPU-accelerated upscaling tool,” obviously tied to AutoSR. Advancing both ends of the game graphics spectrum–with ray tracing at the high end and SR capabilities for lower-end/new platform PCs–certainly makes sense.

I suppose part of the question here is about branding. Will there be a DirectX 13 (or 14, if we superstitiously skip that version number) and, if so, when? My guess is that this will be tied to a next-generation console, most likely running on Arm, and that it will be largely focused on making that platform a first-class citizen for gaming. Timing is tough. I expect second-generation Snapdragon X chips and new PCs at IFA this September, and non-Qualcomm chipsets announced for Windows 11 on Arm sometime this year. And that the Xbox/Windows portable gaming device that Phil Spencer can’t stop talking about will be tied to one of those things. And so maybe that’s the rough timing.

? Xboxed in

christianwilson asks:

I’ve got Xbox on the mind this week. I continue to try Xbox Cloud Gaming. It has improved dramatically since the early days, results vary on device, but my overall experience is still poor overall. Latency is still a problem. Video/Audio starts stable for me but that falls apart the longer I play. The video resolution and color depth are a step behind where they could be.

This is my experience as well. I have seen definite improvements, I get that it’s never going to be identical for advanced games like the latest Call of Duty, but I also feel like those kinds of things are basically unworkable on Cloud Gaming. But older games, and single player experiences, are OK. I was just streaming the original Quake, for example, and that works nicely. Granted, it’s over 25 years old, too.

I know Microsoft is at least somewhat dedicated to Xbox Cloud Gaming. They keep pushing the Xbox app to different platforms like TVs and Fire TV. On paper, their cloud service is ideal for the kind of player I am these days. I can justify a Game Pass Ultimate subscription and use Cloud Gaming as my “platform of the future” for a number of reasons… if the technology fulfilled on the promise. Today, for me at least, it can’t.

I feel like Xbox Game Pass is a component of that “meet gamers where they are” strategy, and that it’s a way to get around some of the issues we see with platform makers like Apple and Google overcharging for fees and having locked-down infrastructure. It’s a way to get Xbox games out of Xbox, which is a limited thing by nature, and onto more devices, in this way getting around licensing issues or just technology issues, where there’s a console game that will never be ported natively to mobile, PC, or whatever else. It’s not a key component of the strategy. But it’s a nice bullet point item on a slide, a “complaint negater,” if that makes sense, that might prevent a fan from leaving the platform.

It’s not the end game. There are reasons Xbox Cloud Gaming is not a standalone offering and is instead a feature of the highest-end and most expensive Game Pass tier only: It’s expensive to operate, and there will always be latency/bandwidth issues that make the experience of playing certain games impossible or lackluster. (There are probably other reasons, too.)

Cloud streaming is always going to be a complex problem to solve because of all hops, client hardware configurations, etc. I can only hope Microsoft will continue to improve the streaming quality. Latency, on the other hand, can absolutely be improved with a cloud controller. Stadia and Luna both proved that. The Xbox marketing leaks from a year or two ago showed a cloud connected Xbox controller. Do you know if that is still in the works?

It’s odd to me how much that was in those leaks hasn’t materialized. Phil Spencer cautioned about that at the time, but I almost feel like some of it is deliberate, as if to make that point. But tied to the DirectX bit above, we’re in sort of a holding pattern with Xbox right now. There’s a third-party Xbox/Windows portable device coming this year, but I feel like the one Microsoft brings to market will be Arm-based, and that the shift–either to other silicon makers, like Nvidia, or to new-generation Snapdragon X with better graphics–is just held up for now. But it would make sense for a Stadia/Luna-style cloud controller to accompany this coming PC-based Xbox portable platform, since it would help solve problems (like AutoSR does for graphics quality/perceived resolution and frame rates when running emulated games on Arm).

? Backup v. sync

wright_is asks:

I know you use Google Drive and OneDrive as part of synchronising files across devices and you plan to use a pair of Synology NAS devices going forward as well. You wrote that in the past you used to use external drives to back up your devices and swapped them out every month off site (at your mother’s?).

Do you currently use anything or any service to backup your data?

For the most part no, but that will likely change when I move to NASes, assuming that experience is positive. That is, I will still use Google Drive, OneDrive, and/or Apple iCloud to whatever degrees, but one goal with the NASes is to move from a cloud service-first system (where the primary copy of whatever data is in one cloud and then replicated or backed up elsewhere too) to a NAS-first system where the NAS-based data is the primary source. And then I back up/sync to whatever cloud(s).

I wrote “for the most part” there because I do technically back up data. For example, I have copies of my work data in both Google and OneDrive, at the least, I work day-to-day mostly in Google Drive, and then on a monthly or quarterly basis, I copy data from Google to OneDrive (and in some cases, move it). But I would like this to be more automated.

A lot of this is based on how these technologies have changed. I used to cart around backup CDs/DVDs when I traveled so I could recover on the road if needed. Microsoft improved Windows to the point where recovering a PC almost always works locally and, if not, can work over the Internet. And it improved OneDrive sync to be real-time and seamless, and I don’t have to worry too much about losing data. So that’s all been streamlined, and I would never do something like make system image backups of whatever PCs these days. There’s no point. I wrote about this a few years back in Roll Your Own Windows Time Machine (Premium).

I have most of my data on OneDrive, and I am slowly swapping it to iCloud as I don’t use any of the M365 tools in my family subscription any more and want to drop that subscription – I use iPhone, iPad and Mac, so iCloud makes more sense. But I still had backups to external drives or my NAS, but that is too loud for me now, so I have recently switched to Backblaze for backups, separate from cloud storage, which is a replication of the data, at best, but no backup. I did have Carbonite running on my old PC and on the NAS, to secure the information on the NAS, in case that died.

I learnt the lesson of not relying on synchronised data stores from my old CTO. He bought a pair of synced servers and declared that we didn’t need to run backups any more! I told him that wasn’t how synced served worked, but he didn’t listen, until he corrupted the main company database. “No problem,” he declared, and went to the synced backup server, only to find that, you guessed it, the sync function had synchronised the corruption in the database! After that, we got our backups back.

Yep. This will be an interesting experiment with the NASes. If this works properly, and there are no guarantees, I can work locally on data synced to a PC, it will sync between the two NASes, and it will be backed up–or some part of it will be–to one or more cloud storage services. If that doesn’t work properly, I would just continue doing what I’m doing for work data and backup to the NASes. But I need to experience this first and see how it does.

I do feel like you need a primary source of truth, what we used to call the master copy. And that in my example with a NAS, that it would sync in one direction to the second NAS. You can call that a backup, I guess. And that it would also then be backed up to the cloud. You can handle the corruption/hardware issues with redundancy–two to four disks in each NAS, in my case–but these things need time. You have to make sure it’s all working. There are so many stories of people backing up religiously and then belatedly realizing the backups are garbage (as with your replicating database corruption example). This happened to me as long ago as my first co-located server in the mid-to-late 1990s, which had a tape backup that never once worked properly but was always running.

We’re old enough and have had enough good and bad experiences with data/data loss that we can at least try to be proactive here. You can’t catch every eventuality. But that was the nice side effect of the blue screen in the Windows Recovery Environment (WRE) I experienced: By the time this happened, I was working in such a way that I wasn’t going to lose data. I just lost half an afternoon.

Anyway, I’m kind of excited to get started with a NAS, if only to see how this all comes together. Or doesn’t.

? Efficiency

brettscoast asks:

Paul this question might he more appropriate for a future what I use article but here we go since you have transferred to a more mobile PC setup for your business/work needs would you say your productivity output has increased over this time or has it dipped? I’m trying to get a sense of whether transitioning from traditional immovable desktop PC to a more mobile PC setup ( 15″ plus sized laptop connected to a larger external display.) makes sense and is it cost-effective? thanks.

The other day, I saw a headline in my feed that read “I replaced my laptop with this mini PC for a week, and here’s what happened,” and I literally laughed out loud. Because I can tell you what really happened: This person spent one week working from one room in one building in one location instead of being free to move around. And that’s the real advantage of a laptop if your work requirements permit it. You can work from anywhere. That’s why I think of it as a more mobile setup. You’re literally more mobile.

So that alone makes you more productive. I routinely move around with my home, whether it’s here in Mexico City or back in PA, and in both places, I have a desk with a dock-based setup with at least one large display, I have a nice laptop table I can use from the couch (which I’m using now as I write this), and I can go lay on the bed. I can take the laptop with me when I leave the building and I can work from anywhere.

From a cost perspective, it depends, but I feel like a laptop almost always comes out ahead if you are actively using a PC each day and especially if you’re going to have a laptop anyway and need to work elsewhere at least some of the time. Throwing together a dock setup, with an external keyboard, mouse, display, and maybe a webcam and microphone, is simple enough. (My wife and do this in both homes.) Using a single cable connection to the laptop is nice, you can just unplug it and go.

There are all kinds of ways you can do this. I do have a mini-PC setup in PA right now, and I’m going to look at that Snapdragon Dev Kit mini-PC back there when we get back. But I can move back and forth easily enough. And one could make a good argument for repurposing an older PC of whatever form factor for a desk-based setup and then using a separate laptop otherwise.

Modern laptops are so much more reliable than before, the performance is off the charts for most use cases, and you can even play games on these things without needing dedicated graphics. I just expect them to come on immediately when I open the lid now, because they do, mostly, and those that don’t are the exception. And are undesirable.

Everyone is different. But I very much prefer working on laptops. In large part because it makes me more productive and efficient.