00:00:07 ◼ ► I thought you meant like you cut something out of the show last week and you weren't pleased
00:00:15 ◼ ► Well, so the problem with cutting the tip of your left thumb is that, well, that's my phone
00:00:30 ◼ ► And so it was like, first of all, the entire left half of the phone keyboard when typing
00:00:37 ◼ ► And so I mostly just have to use the, you know, the other hand and then trying to navigate
00:01:00 ◼ ► I don't care about that, but that I can deal with, but I can't switch the side of the command
00:01:19 ◼ ► And I'm almost out of this national nightmare, but man, but there is no worse, like finger
00:02:17 ◼ ► Uh, and first of all, I'd like to say a thanks to the people who reached out and, uh, said
00:02:24 ◼ ► Uh, I'd like to, uh, congratulate you maybe, or thank you, Marco, uh, for doing such a good
00:02:30 ◼ ► You know, there's a lot of times that, uh, maybe not as much, John, but certainly I will
00:02:39 ◼ ► You know, something that we've talked about a few minutes before to set up, set up a joke
00:02:52 ◼ ► Uh, which on the one side is a little frustrating because, you know, I want to be along for the
00:03:01 ◼ ► So this is the sacrifice we make in our mutual friendships, uh, for you, the listener, you're
00:03:07 ◼ ► This is why I can do things like set up an entire data center worth of Mac Manus without
00:03:11 ◼ ► The only, the only person who knew about all this stuff as it was happening, I told underscore
00:03:20 ◼ ► We both had a lot of fun, like, you know, anticipating like when you guys would finally find out about
00:03:25 ◼ ► it, but yeah, and it's, you know, like when I bought the restaurant, I didn't tell anybody on
00:03:30 ◼ ► the internet because it's like, well, this, this is, and it's, it's only cause like, oh, this is
00:03:44 ◼ ► Uh, but anyway, I, I, I'm getting on a tangent, but the, what I was trying to say, all, all
00:04:06 ◼ ► Uh, first of all, how is it, Marco, that you had never seen a data center before you worked
00:04:17 ◼ ► So this was a question posed on, on the overcast Reddit, you know, it's for people who don't,
00:04:21 ◼ ► who kind of aren't in this world or maybe are in a different part of this very large world.
00:04:25 ◼ ► Um, it might've seemed ridiculous that I would have operated Tumblr for its first, you know,
00:04:34 ◼ ► And the reason why is because during those four years, it was 2006 to 2010, Tumblr didn't have
00:04:50 ◼ ► But what we had instead, you know, what people did back before compute instances, uh, which
00:04:56 ◼ ► used to be called VPSs or virtual private servers, what we all did before VPSs was dedicated servers.
00:05:03 ◼ ► So what this is, is now in the business, I believe they now call this bare metal servers.
00:05:08 ◼ ► You go to a hosting company on the internet and you just lease a server, you know, through
00:05:16 ◼ ► And you have a server allocated to you that you never see, you never deal with the physicality
00:05:23 ◼ ► They deal with it, but you know, and you have to do things like, you know, put in like a
00:05:29 ◼ ► And if one fails, either their monitoring or your monitoring will have to tell somebody,
00:05:47 ◼ ► Cause like, well, you can, you know, you can copy the data off yourself and copy it onto
00:05:51 ◼ ► But you know, the, the concept we have now, like instances that can be resized, like that
00:05:55 ◼ ► didn't exist, you know, very little concept of like managed images and stuff like that.
00:06:05 ◼ ► Um, at that time, it wouldn't have made any sense for us to buy a, like to have our own data
00:06:16 ◼ ► Most of the time we were on, we were on two servers, you know, maybe up to six or eight
00:06:25 ◼ ► They were dedicated servers and that's what everyone did before, you know, AWS, EC2 and
00:06:30 ◼ ► before, um, compute instances, virtual instances, things that we, things that we use now.
00:06:37 ◼ ► It's called bare metal servers, but I think it's a lot less of the business now than it
00:06:48 ◼ ► Uh, some, a bunch of people have asked this, uh, including Matt King, uh, how do you handle
00:06:59 ◼ ► There was like, I don't know how the RSS about podcasts works, but there was like an enclosure
00:07:07 ◼ ► The, um, the podcast 2.0 spec people defined a transcript tag, like a podcast colon transcript
00:07:21 ◼ ► And when Apple podcasts launched their transcripts, you know, whatever that was a year and a half,
00:07:26 ◼ ► two years ago, they said, okay, well, we will do our own transcripts, but if you want to supply
00:07:40 ◼ ► And the reason I haven't actually tested it is that very, very few podcasts so far actually
00:07:57 ◼ ► So I do plan to show transcripts from people's podcast transcript tags, uh, from SRT and web
00:08:14 ◼ ► you then cannot use dynamic ad insertion because your timestamps will be all messed up.
00:08:18 ◼ ► And there, there is not a well-supported standard to embed a transcript in a file that any podcast
00:08:35 ◼ ► One of the, one of those tags is an ID three tag to specify time synced captions to be displayed.
00:08:48 ◼ ► Um, so if you were to support like podcaster supplied transcripts with dynamic ad insertion,
00:08:55 ◼ ► something like that, that comes with the file and therefore can be adjusted with the file
00:09:04 ◼ ► Um, but given that these platforms don't even like the DAI platforms don't even support
00:09:21 ◼ ► And because almost all podcasts of medium to large size now are monetized with DAI, I can't
00:09:36 ◼ ► And another thing they could do is they could supply a transcript with timestamps that just
00:09:41 ◼ ► don't include any ads, but then the real time display of it will break it in the player.
00:09:46 ◼ ► Like, cause if they, you know, if they, if their transcript is just their content and doesn't
00:09:50 ◼ ► have any of the ad content in it, which it wouldn't in this context, as soon as you reach
00:09:55 ◼ ► a point in the show where they've inserted an ad for, you know, say two minutes from that
00:09:58 ◼ ► point forward, all of the timestamps in their supplied transcript will be off by two minutes.
00:10:03 ◼ ► And so the utility of that I think is always going to be significantly less than one that
00:10:10 ◼ ► was dynamically generated in such a way that, that it can, it can synchronize itself back
00:10:18 ◼ ► So in that way, I think my transcripts are probably going to provide more functionality
00:10:28 ◼ ► Cause basically, yeah, if you supply a transcript and you have dynamic ads inserted, that's not
00:10:33 ◼ ► So even if you supply your own, I, I haven't fully made this decision yet, but I think even
00:10:40 ◼ ► if you supply your own, I think I should probably also offer mine and maybe just have like, you
00:10:46 ◼ ► know, like a, a, a tab picker at the top of the screen that says, you know, my transcript,
00:11:01 ◼ ► Do you, so even if you accept these transcripts that are offered, which it sounds like you're
00:11:07 ◼ ► not too enthusiastic about, there's no real standard for noting like individual timestamps
00:11:13 ◼ ► Like, you know, you had said that you theoretically have the resolution by which you could do
00:11:18 ◼ ► the like Instagram thing of highlighting individual words as they're being spoken and that presumably
00:11:26 ◼ ► Um, well the, so the, the transcript tag, the podcast transcript tag, it supports these, the,
00:11:39 ◼ ► They do just timestamp line because they're, they're geared towards, uh, subtitles that will
00:11:45 ◼ ► So it's like at this timestamp, show this text on the screen and then three seconds pass
00:11:50 ◼ ► Well, it, it is usually, but I don't think it needs to be necessarily like there's nothing
00:11:55 ◼ ► stopping the authors of those files from having like one line per word and having, you know,
00:12:01 ◼ ► having each, I mean, it would look ridiculous in the file context, but you know, but I think
00:12:05 ◼ ► ultimately though you are right, Casey, that like in practice, the way these files are usually
00:12:11 ◼ ► authored, the way they are usually used, um, because they come from the world of subtitles
00:12:26 ◼ ► I think that almost any producer who makes SRT or VTT files would give word level timing.
00:12:34 ◼ ► And then I was going to ask, you know, would you create a spec, you know, kind of in the
00:12:37 ◼ ► spirit of, um, X callback URL, but it sounds like you don't really need to, you could just
00:12:49 ◼ ► And, and, you know, the reality is like, I can try to do a lot of features around these file
00:12:55 ◼ ► formats, but again, because the world of DAI does not really support them, there's just not
00:13:03 ◼ ► I will support them in the sense that I will download them and parse them and have a way
00:13:10 ◼ ► I think that that would probably be the right move, but I don't think I'm going to have them
00:13:15 ◼ ► And honestly, I would be very surprised if most of my users ever saw a podcast that actually
00:13:24 ◼ ► Uh, a couple of people wrote in, including Drew Stevenson with regard to how you're managing
00:13:30 ◼ ► And Drew wrote for background cues, did Marco consider something like a factory F A K T O R Y.
00:13:36 ◼ ► My jobby job uses it for processing media in the background on some Mac minis and it's rock
00:14:14 ◼ ► So what this queue is, is Beanstalk D, uh, which is very old, very boring, and it works
00:14:33 ◼ ► Um, I do use Redis's cues for feed crawling because they allow me to, the way overcast is
00:14:41 ◼ ► Basically, I have these crawl servers that run the go process that I wrote a million years
00:14:55 ◼ ► But if, I mean, I think today, if I ever needed to add anything to it, I would just have AI
00:15:01 ◼ ► Um, but anyway, the way those work is they crawl the feeds, they stuff the content of a changed
00:15:08 ◼ ► feed into Redis, and then they add to a big Redis queue so that other processing PHP processes
00:15:19 ◼ ► Um, Redis though is even heavier duty for this case than Beanstalk, which is Beanstalk D
00:15:25 ◼ ► is like, what if Memcached D, but for queues, Beanstalk is where I put lighter queue work
00:16:14 ◼ ► Like again, like the, what overcast processes overall, you know, in its job queues, especially
00:16:26 ◼ ► Any kind of like managed, managed queue service, when you actually price out what it would cost,
00:16:36 ◼ ► I just run it myself and it costs nothing because running Beanstalk D is incredibly lightweight.
00:16:50 ◼ ► And you know, I managed it through Beanstalk D and I send them and it's fine because I send
00:17:01 ◼ ► Cause like the, the processing time, the server resources to send those is as close to zero as
00:17:15 ◼ ► you know, just X cents per thousand notifications, you know, whatever their pricing ends up being
00:17:21 ◼ ► When you're doing millions of things a day, that could be like tens to hundreds of dollars
00:17:28 ◼ ► And when you spread that over the, you know, 12 years I've been running overcast, like that's
00:17:34 ◼ ► So if I can write, if I can write it myself in less than a day and, and have it work entirely
00:17:45 ◼ ► I don't have to worry about like what happens if this service gets merged with some other
00:17:53 ◼ ► And now, now they are, now it's into enterprise and costs 10 times more for my use case.
00:17:57 ◼ ► Like all those things that happen constantly in our business, I don't have to worry about
00:18:03 ◼ ► Like the boring old way with a few simple processes running on some Linux servers that don't
00:18:27 ◼ ► I looked up for the show notes, the Beanstalk D website, and it is exactly the kind of website
00:18:34 ◼ ► So, uh, yeah, this tracks, but, uh, you did another Marco thing, which is you wrote yourself
00:18:57 ◼ ► Like this is the app that transcribes things that pulls the jobs from the servers and actually,
00:19:11 ◼ ► I select, I assign each one like which, because the, the Apple speaks transcription API only supports
00:19:25 ◼ ► Uh, but anyway, so it supports six languages, but only supports three of them being installed
00:19:59 ◼ ► And it reports its stats of how many jobs it has done, uh, in the last minute or, or whatever.
00:20:12 ◼ ► So if for instance, one of the Mac minis doesn't respond at all, it will alert me and I can go reboot it or something that doesn't happen very often.
00:20:19 ◼ ► Um, but also if one of the Mac minis is reporting like a suspiciously high or suspiciously low job count per minute or, or transcription minute rate per minute, I can then go look, take a look and see like something might be wrong there.
00:20:33 ◼ ► So if for instance, it can't get jobs for a long time for more than a few minutes, or if a job never completes and it never refills its jobs, uh, after a certain amount of time, it will quit itself.
00:20:54 ◼ ► Like if, so if something weird has gotten wedged somewhere, this kind of thing doesn't happen at very high rates, but when you're running 48 instances of them 24 seven for months, you know, they do occasionally need like weird things like that, you know, sometimes happen.
00:21:13 ◼ ► And I'm glad you provided a screenshot, which we'll put a link to that in the show notes.
00:21:21 ◼ ► While they each only have 16 gigs of RAM, it is possible to spread LLMs over multiple Macs these days.
00:21:27 ◼ ► We've all probably seen the videos of four Mac studios clustered over Thunderbolt, but you can also do it over ethernet just a bit slower.
00:21:46 ◼ ► Uh, the meme was, and I'm surprised there was not a, like, know your meme, uh, page for this.
00:21:59 ◼ ► It was a system where you could just take a bunch of commodity servers and then distribute a job across them.
00:22:03 ◼ ► And so you'd run Linux on these random PCs and you'd be like, but now I have five PCs and if I have a job and I can break it up into five pieces, these five PCs can work on it.
00:22:15 ◼ ► And any time there was any kind of hardware story on slash dot, uh, I don't know when it started, but people started posting, imagine a Beowulf cluster of these.
00:22:27 ◼ ► And that just became a meme where people would say it about things that are not computers or whatever.
00:22:34 ◼ ► And so Aaron Dippner basically just wrote, imagine a Beowulf cluster of these about your Mac minis.
00:22:38 ◼ ► Like you could take one job and distribute it to all 48 and they'd all work on it at the same time.
00:22:43 ◼ ► Um, and what he's referring to, we'll put a link to this in the show notes is, uh, this thing where Apple gave a bunch of YouTubers, a stack of Mac studios.
00:22:53 ◼ ► I think it was like four Mac studios with like a little, little miniature, little cute, little, uh, Mac studio rack kind of.
00:22:59 ◼ ► Um, and it's to promote their RDMA over Thunderbolt technology that was added in Mac OS 26.2, uh, we'll link to Apple's developer site, uh, about that tech.
00:23:09 ◼ ► Um, what they say about it is RDMA over Thunderbolt enables low latency communication between Thunderbolt five hosts for use cases, including distributed AI inference using MOX.
00:23:18 ◼ ► And so what the YouTubers did with this stack was they got like the biggest model that they could get that wouldn't run on a single Mac studio because it didn't have enough RAM.
00:23:26 ◼ ► And they would run it on the four Mac studios combining their RAM and, you know, do performance numbers or whatever.
00:23:33 ◼ ► One from Alex Ziskind, uh, the title of his video is I ran a trillion parameter AI on a Mac.
00:23:40 ◼ ► And then Jeff Geerling also has one that says Apple didn't have to go this hard, but there were others.
00:23:45 ◼ ► So if you want to see an example of a technology that Aaron is talking about, imagine a Beowulf cluster of these.
00:23:52 ◼ ► It's like, can you distribute something across multiple computers that wouldn't fit on one?
00:23:58 ◼ ► And obviously doing it over Thunderbolt five with machines that are right next to each other is really fast.
00:24:02 ◼ ► Doing it over ethernet is probably a little bit slower, but that's what he's pitching here.
00:24:07 ◼ ► And so the main, the main downside of this for my use case is that, um, yes, as John said, speed of the communication is everything.
00:24:15 ◼ ► Uh, and, you know, ethernet between these Mac minis is just one gigabit because these are the base models.
00:24:26 ◼ ► I also don't have, don't have a giant 10 gigabit switch there, but you know, that could be remedied.
00:24:29 ◼ ► Um, but they don't, they only have one gigabit ports and these Mac minis only have Thunderbolt four because Thunderbolt five on the Mac mini currently requires the M4 pro chip, which like triples the price of the Mac mini and does not triple the performance for my use case.
00:24:47 ◼ ► So because I have the base models, I could only do it over ethernet and that would be probably slow enough.
00:24:56 ◼ ► And the pitch for this is that you would use a somehow use some bigger model than what you're using.
00:25:06 ◼ ► Like the pitches, well, what if you wanted to run some more sophisticated model, um, that wouldn't fit in the Ram of a single one.
00:25:13 ◼ ► You could, you could take lop off a portion of your 48 and say, this is like this cluster of four is the, the big beefy, uh, transcriber for the top 10 pop, most popular podcasts.
00:25:25 ◼ ► But right now, um, this is something that like, I'm, I'm going to look at these kinds of uses in the future, especially cause like once I'm done transcribing the back catalog for as many podcasts as I can, these are going to be at like, you know, a third to half capacity.
00:25:44 ◼ ► And I mentioned before, I'm, I would definitely want to move some of my heavier tasks from the main servers that are, you know, the main Linux servers.
00:25:51 ◼ ► I'd love to move a lot of those tasks to these, you know, to whatever degree that's reasonably possible.
00:25:59 ◼ ► Um, so I will have to maybe get creative on like, what do I do with this computing power that I have here?
00:26:06 ◼ ► And, uh, you know, ideally like things that can stay within one machine are better because I, you know, it's easier to scale that it's easier to share code between the iPhone and the servers.
00:26:19 ◼ ► Um, you know, so ideally I would just keep doing more things with Apple's foundation models and we'll see like, you know, this summer at WBDC, there might be updates to the foundation models.
00:26:30 ◼ ► And maybe they get things like larger context windows, which should make them more useful.
00:26:38 ◼ ► Right now they are being criminally underused, uh, in the sense that they, uh, let's see my current, my memory use.
00:26:48 ◼ ► Um, one of them that's under like, you know, full transcription load right now, um, is using about seven gigs of its memory of its 16 gigs.
00:27:03 ◼ ► That will change once I've moved image resizing to them, but it won't change that much.
00:27:07 ◼ ► So I, I think I'm going to have a lot of CPU or a lot of GPU capacity that I could use and a decent amount of RAM.
00:27:18 ◼ ► But right now I'm still going through back catalogs and maybe we'll see what the summer brings.
00:27:46 ◼ ► but it just became a meme of whenever you take a bunch of stuff and connect it together
00:27:53 ◼ ► Coming all the way back around to Aaron Dibner, who started this whole conversation, Aaron
00:28:08 ◼ ► I think this was done via ethernet, if I'm not mistaken, but you can look and we'll put
00:28:16 ◼ ► What I love about Squarespace is that they can host any kind of website you need to host and they do it so easily with so little effort and no like nerdery or nerd skills required.
00:28:43 ◼ ► So if somebody comes to you in your life and says, hey, you're a computer nerd, can you help me make a website?
00:28:52 ◼ ► Go to squarespace.com slash ATP and you'll get 10% off your first purchase of a website or domain by using that special URL.
00:28:58 ◼ ► But you can build the site there yourself and that's way more powerful than than somebody coming to you as a nerd and having you like do it for them or have to go through you all the time to keep it updated when they can do it themselves.
00:29:10 ◼ ► And what's great about Squarespace, too, is it whether they're trying to build like, you know, a portfolio site, a simple site or a whole business site.
00:29:17 ◼ ► You can you can host entire storefronts, entire businesses on Squarespace, whether you're selling physical goods, digital goods, things like time slots.
00:29:27 ◼ ► If you're, say, a consultant or a coach, you can sell things like electronic content, PDFs, you know, e-books and things like that.
00:29:35 ◼ ► Private membership areas, newsletters that are paid or not paid like they support all of this stuff on Squarespace.
00:29:44 ◼ ► And I strongly suggest when you or someone else needs to build a new website, go use their free trial because you can build the entire site in trial mode.
00:30:01 ◼ ► Anything a modern business might want, analytics, SEO, campaigns, you know, there's so much in Squarespace.
00:30:11 ◼ ► Start your free trial and use offer code ATP at purchase to save 10% off your first purchase of a website or domain.
00:30:49 ◼ ► I'd like to leave aside the announcement from today, yesterday, whenever it was, for just a moment.
00:30:54 ◼ ► But based on what you have read or maybe you haven't had the time to read anything, what would you say is your summarization of what you've learned?
00:31:02 ◼ ► So what companies do that have to manage large fleets of Macs, of course, is not setting them all up one by one like I did here.
00:31:13 ◼ ► What companies do is they have this thing through what I believe until yesterday was called Apple Business Manager or something like that.
00:31:26 ◼ ► You can configure machines to be set up for your business and locked to your business with MDM, which is the framework they use to manage, like, you know, corporate iPhones and iOS devices and Macs.
00:31:37 ◼ ► You can buy these things through your Apple Business Rep, through Apple's business channels, or through certain authorized resellers, and have the machine's serial numbers automatically enroll to your company when they connect to the internet for the first time.
00:31:51 ◼ ► So you can, they call it, like, zero-touch deployment, I think, so that you can, like, ship a laptop to a remote employee, and when they plug it in and set it up, it gets activated to the company's management system through Apple.
00:32:14 ◼ ► So you can do things like push software updates and stuff like that, and, you know, run, you can run programs, you can change settings, you know, whatever it is, like, you can have all that centrally managed.
00:32:24 ◼ ► And the only downside is when you already have the machines pre-existing, like I do here, everyone said, like, in order to enroll them into this kind of management, you need to wipe them, and you need physical access.
00:32:38 ◼ ► So it would be a process of, you know, I would, like, you know, go, you know, go to the data center one day, bring a big cup of coffee, you know, keep it below the Mac minis, John, and just go through, like, one by one, all right, reset this one, reset this, you know, just take them down one by one, reset them, wipe them, put them in the thing.
00:32:57 ◼ ► But before this announcement of Apple Business, which happened, you know, yesterday, the day before, yesterday, before this announcement, there was, like, the second part of it, which was, okay, Apple Business Manager would have, like, the component of, like, registering the serial numbers, associating them to a business that had an MDM thing.
00:33:15 ◼ ► But then what everyone else did was have their own MDM providers through companies like Jamf, former sponsor of the show a long time ago.
00:33:23 ◼ ► This is what most big businesses do is they have one of these MDM apps that the problem with those from my use case, and, you know, they provide a huge amount of functionality, because usually it's not only providing, like, the basics of MDM configuration and stuff like that, but they also usually have additional services they layer on top of it, like security packages, stuff like that.
00:33:45 ◼ ► The problem is that all of those MDMs, or at least most of them, you're paying something like $6 to $12 per Mac per month, and, you know, again, when you have 48 Mac minis that literally don't have users, that's not a great selling proposition for me.
00:34:05 ◼ ► Like, all of the features of these MDMs that do, like, you know, user security management, making sure your users have good passwords, and making sure they don't browse weird things on the internet, like, that's all stuff that businesses do need for their computers, but I don't need for these computers.
00:34:22 ◼ ► These computers have no user, they run one app, you've seen it now, like, it's a very simple arrangement, and what I'm doing right now to keep them up to date, their software I don't really keep up to date yet, because that hasn't really had a need yet, I guess I'll get there.
00:34:56 ◼ ► So, the way I access these machines, because they're all behind the router at the data center, so they all run Tailscale, and that allows me to say, you know, SSH, you know, into Scribe23, and it just gets me there.
00:35:13 ◼ ► So, what I do now, when I have to update my transcriber app, is I do, you know, an archive build on Xcode, and I export the binary after it's signed, and I just use SSH to SCP it over to each of these Macs, and like, you know, part of the SSH script is like, you know, have LaunchD stop the app, SSH the new one into place, have LaunchD start the app again.
00:35:37 ◼ ► And it just goes through the 48 servers, and just does that to each one in succession, and I can push an app update to all of them in about 30 seconds.
00:35:44 ◼ ► So, I've kind of already built a lot of what an MDM would give me, and in a more nerdy way, but what would be even better is if I didn't have to run Tailscale on all of these.
00:36:00 ◼ ► So, if the MDM situation can do things like, you know, remote desktop, or, you know, remote viewing, remote control, if it can do those things and give me network access to them directly without me having to run something like Tailscale on them, that's great.
00:36:15 ◼ ► I haven't had time to look that much, but for the most part, most of the features that the MDMs offer are mostly things I don't really need.
00:36:32 ◼ ► And, getting to the announcement now, coincidentally, Apple announced yesterday that they've merged all of their various, like, business backend services into the new thing just called Apple Business.
00:36:44 ◼ ► Yeah, so, Apple Business is apparently a thing where they're, I guess, doing their own MDM or providing their own sort of solution.
00:36:52 ◼ ► I'm a bit fuzzy on it, but let me read a couple of things that John was kind enough to pull out of the newsroom announcement.
00:36:56 ◼ ► Apple today, on March 24th, announced Apple Business, a new all-in-one platform that includes key services companies need to effortlessly manage devices, reach more customers, equip team members with essential apps and tools, and get support.
00:37:08 ◼ ► Apple Business will be available on April 14th as a free service in the U.S. and 200 other countries.
00:37:23 ◼ ► It sure seems to me like if I were you, I would at least kick the tires on this first and see what it's got.
00:37:32 ◼ ► Yeah, that's my current plan is, like, now that Apple is about to launch, I think they said April 14th, yeah.
00:37:40 ◼ ► Now that they're about to launch their own first-party, like, better, more featured version of this, I would rather just use that.
00:37:56 ◼ ► But I will look at this, especially, like, as I approach this summer's beta season, if I do end up wanting to run betas on some or all of these, that's going to be a burden.
00:38:07 ◼ ► So, you know, having, like, you know, centralized software management will be a lot nicer.
00:38:15 ◼ ► One suggestion, and a lot of other people wrote in with a suggestion as well, is, like, it does seem like you need something probably from Apple that can handle OS updates and stuff like that.
00:38:23 ◼ ► Although the various reports we got about the quality of Apple's previous Apple Business Manager software and their MDM stuff is not all glowing, but, you know, and this is a new product, Apple Business, supposedly.
00:38:37 ◼ ► But the other part of it, the part where you're SCPing a file onto a bunch of Macs, like, everyone, this is where the second piece of advice we got is, like, Unix nerds are saying there's a million, you know, free open source command line packages for configuring a whole bunch of Unix-y servers.
00:39:01 ◼ ► And if all you have to do with, if all you're doing is essentially something that you can do with SCP now, rather than doing SCP in a loop and dealing with that and everything, there's a million software packages, most of which probably run on the Mac that you can try.
00:39:23 ◼ ► They're, you know, they're Unix-y command line things, and you have to figure out how to get that installed on the Macs.
00:39:29 ◼ ► But if the only job of that part of the system is, oh, and by the way, there's something that I want to do that I can do from, like, a shell script or whatever, just you can do that outside of Apple's things.
00:39:41 ◼ ► And the reason I worry about it, having not used these products, is, like, you have a new version of the transcription app that we just talked about before.
00:39:50 ◼ ► I'm sure Apple Business Manager has a way to get the new version of your transcription app to all the machines.
00:40:11 ◼ ► Yeah, even if they offer that as an option, it's just, like, that's not what those things are made for.
00:40:17 ◼ ► Like, you probably never even want that in a situation where you're deploying to 2,000 Macs.
00:40:23 ◼ ► So maybe, depending on if Apple Business Manager or whatever, if Apple Business has some gaps, rather than saying, oh, now I need one of those commercial MDMs that charges me an amount per Mac per month, at least look at, like, Ansible and some of the other Unix-y stuff that basically are a fancy distributed way for you to run Unix commands on a fleet of servers without you having to, like, deal with the plumbing.
00:40:47 ◼ ► I mean, so my argument, so that would be, though, like, I already built that with SSH and a shell script.
00:40:55 ◼ ► I know, but, like, it's just, you know, how good is your error checking and how, how parallelized is it?
00:41:04 ◼ ► People, things like Puppet sprang into existence because people used to write scripts that would loop over things and distribute tar balls and untar them and do, you know, like, they, and then they made products out of that because doing it manually kind of sucks.
00:41:16 ◼ ► Yeah, I mean, and again, like, this is the kind of thing, like, what your needs are determines a lot about what the, you know, what the solution here should be.
00:41:24 ◼ ► My needs are so simple here that if all it's going to save me doing is, like, you know, right now, like, I have just a simple, you know, text file that's about 30 lines long that is my instructions for setting up a new Mac mini.
00:41:41 ◼ ► You know, it's like, I said, I changed, like, five settings, and, you know, mostly about power management, like, stay on, really stay on, no, really, always stay on.
00:41:50 ◼ ► And, you know, give it my SSH public key, and say, all right, well, you know, accept anything from this, go, and, oh, and turn on remote desktop.
00:42:03 ◼ ► If something is like, well, you know, you can, you can have more control over it by installing this, this different product.
00:42:10 ◼ ► Okay, but right now, I install no different products, and I have the, I have all the control I need.
00:42:16 ◼ ► So, to me, I would say, like, that additional complexity, even if the product is free, you know, setting aside any kind of cost, the additional complexity of having another thing being involved, that's another moving part, that's another company that can go out of business, or another thing I have to keep updated, or that might break in three years with new, you know, OS update that Apple gives, or whatever.
00:42:38 ◼ ► Every one of those moving pieces, I evaluate, like, is this, is this giving me enough benefit over my current situation to be worth its additional risks and complexity?
00:42:48 ◼ ► And, because my needs for these are so simple, the answer to that for a lot of these things is no.
00:42:54 ◼ ► Now, not for everything, and again, like, software updates are the big thing that, that I definitely would like to be automated better.
00:43:01 ◼ ► But any, adding any other layers, any other packages, any other products, they all have to pass that test before they're going to be worth it for me.
00:43:09 ◼ ► And they're going to be worth it for a lot of businesses before they're going to be worth it for me.
00:43:13 ◼ ► Yeah, I wasn't suggesting you get rid of your little 30-line thing that you run, I was suggesting you just get something else to be the plumbing to run your 30-line thing.
00:43:20 ◼ ► Because, you know, again, if you just, it'll do it more in parallel, they can monitor if it's been successfully done on all the machines.
00:43:27 ◼ ► If you update the machines, it will immediately redo it, because it'll see that it hasn't been done, that type of thing.
00:43:33 ◼ ► It's basically a fancy, remotely run a bunch of commands on a bunch of machines for me, and make sure that if they, if those settings ever get changed, or if anything changes in the machines, that you set them back to the way they were.
00:43:45 ◼ ► And it's also a fancy plumbing way to say, oh, and if there's some files you want to be on those machines, like your app, just tell me where it is, and I'll make sure it's, so the latest version is always on all the machines.
00:43:59 ◼ ► So there is, you know, like, there's, there's something to be said for doing it all yourself.
00:44:04 ◼ ► But if it turns out to be tedious, and you're spending a lot of time dealing with that, those unique, I would, what I'm saying is I would look at those unique things before I would look at one of the fancier MDMs.
00:44:18 ◼ ► No, I think I'm just, I'm working through all the, you know, bugs and reports and feed.
00:44:23 ◼ ► The other thing is, I said that, you know, that I was, I had spent a huge amount of time last summer getting an algorithm to fingerprint the audio so that I can align it around dynamic ad insertion.
00:44:37 ◼ ► I did not realize that Shazam Kit was an API, and it offers exactly that functionality.
00:45:00 ◼ ► But sometime soon I'm going to experiment with that to see, like, should I basically start using their alignment instead or in addition to mine?
00:45:08 ◼ ► But that was kind of a little bit like, oh, no, did I, did I spend months doing something from scratch that Apple has an API to do?
00:45:30 ◼ ► But we'll see, like, so part of, part of my fingerprinting algorithm is it has to be very, very low granularity of time.
00:45:43 ◼ ► Like, if, if the, if the block size it's working on is like a 10 or 15 second chunk, that might not be good enough for what I'm doing here.
00:45:52 ◼ ► Like, what my, my fingerprinting, the sample resolution for it is something like a quarter of a second.
00:45:59 ◼ ► And it also has to be a small enough data set that, like, whatever the fingerprint that you generate for an entire podcast, it has to be a reasonable size.
00:46:09 ◼ ► It can't be, like, many megabytes of data if you're going to do that resolution across a three-hour show, you know.
00:46:15 ◼ ► So, I'll see, you know, I don't know what, what exactly their data is stored and what resolution it is.
00:46:21 ◼ ► So, that's all going to be part of the experimentation phase and see, like, you know, should I switch to that or not?
00:46:26 ◼ ► And in a final piece of follow-up for today, AirPods Max and their popularity in New York, you had commented that everyone and their mom has a AirPods Max in the New York metro area these days.
00:46:39 ◼ ► How is that possible, especially when some of these people are kids, you know, like, 15, 16-year-old kids or whatever?
00:46:44 ◼ ► Apparently, all of them, or maybe most of them anyway, are fakes, which I did not know.
00:46:49 ◼ ► Yeah, I, so many people have written in to say either they, either they suspect they're fakes or, like, they live in Manhattan and they see that they're fakes.
00:47:02 ◼ ► Uh, I do see, like, you know, the guys on the streets that sell all, like, the fake luxury goods on those little tables.
00:47:13 ◼ ► And every, like, all, like, the store windows that are, like, the, like, the random little electronic stores where you look and you're, like, I bet most of the stuff in there is probably knockoffs.
00:47:36 ◼ ► Like, people say, like, oh, you can, you can tell, they look cheaper, their finish is wrong, they have a plastic crown.
00:47:44 ◼ ► But at the distances that, you know, I'm not going to, like, you know, hover over someone's shoulder and start eyeing the details of their digital crown to see, like, hey, is that plastic?
00:47:59 ◼ ► Um, so, I, I assume, I'm sure lots of them are fakes, but I also think a lot of people are probably buying them because they are, like, visually iconic and Apple is still selling them.
00:48:10 ◼ ► And, you know, they, if Apple wasn't selling these, you know, in any meaningful quantity, they wouldn't have even given them the two half-butted updates they did give them.
00:48:30 ◼ ► So, if you've ever shipped an app internationally, you know that localization pretty much goes one or two ways.
00:48:36 ◼ ► You can ship your localizations via strings or, God help you, a spreadsheet to some faceless company that has 80 million salespeople and three translators.
00:48:46 ◼ ► And then, forever later, you get translations that sound, I don't know, something like an office printer's mad at you.
00:48:53 ◼ ► Or, option two, and you're probably going to want to go with option two, you work with Babylon.
00:49:01 ◼ ► You can talk to them in Slack, deliver strings during a sprint, and get answers quickly when something in your UI breaks.
00:49:18 ◼ ► I said to them, hey, I've already got all my localization files in this public GitHub repo.
00:49:31 ◼ ► And I figured, to be honest with you, that I was getting the white glove treatment because of the whole, you know, ad read that we're in the middle of and all that.
00:49:40 ◼ ► And not only that, when they started looking at these translations in the context of the test flight, after I added them, of course, there were a handful of times that they said, oh, you know what?
00:50:20 ◼ ► So, if your app is ready to go global, or if you're tired of fixing translations after the fact, Babylon will help you ship localizations that won't make your international users feel gross.
00:50:52 ◼ ► Anyways, suffice to say, June 8th through 12th, apparently in episode 673 back in January, which for my purposes is like 700 years ago, we talked about what we were looking forward to in 26th.
00:51:09 ◼ ► I'm sure there are things, I mean, there are definitely things I'm looking forward to, but nothing is like leaping out at me right this moment.
00:51:29 ◼ ► It's by far the most likely scenario that the Mac Studio will be updated at some point in the future, and at that time, they will discontinue the Mac Pro.
00:51:41 ◼ ► It's not as if the Mac Pro is filling some need waiting for the Mac Studio to replace it.
00:51:48 ◼ ► Anyway, what I remember from the January episode is we were talking about, like, tech things for the entire year.
00:51:55 ◼ ► Narrowing it to WWDC, I think the subset of things that I talked about on that episode, are still basically the same.
00:52:00 ◼ ► I'm still looking forward to M5 Ultra or whatever new chip that they haven't yet announced.
00:52:05 ◼ ► And presumably that will be in a Mac Studio, and I'm looking forward to it both because I'm interested and also because I'm probably going to buy one.
00:52:10 ◼ ► So I'm kind of, you know, was looking forward to it and now kind of a little bit scared about how much it's going to cost to get that 8TB SSD and the amount of RAM I'm going to want.
00:52:20 ◼ ► And then the other thing that I also talked about, I think, in the January episode was, let's see how Apple does with their AI stuff.
00:52:26 ◼ ► Like, they were supposed to, like, roll out a bunch of new Siri stuff in supposedly 26.4, which was just released, but notice it didn't arrive in 26.4.
00:52:34 ◼ ► In fact, we had an item in the show notes that just kept getting pushed off forever that was basically like, hey, remember Apple's dates that it set for itself?
00:52:40 ◼ ► That they were going to do some serious stuff in 26 point something and then more in 27?
00:52:58 ◼ ► Even if it's not, if there's no Ultra at all, just the M5 Max Max Studio and see how that's priced.
00:53:20 ◼ ► On the one side, I'm not sure what it is that I want them to tell me with regard to the AI story.
00:53:26 ◼ ► But then again, Apple is so very good at telling me, oh, you never knew you needed blank, but you did.
00:53:36 ◼ ► That's what we're all waiting for is just because almost I always feel like everything else, like the world is passing them by and all the other areas.
00:54:01 ◼ ► With regard to development tools and things like that, I think there's a couple of things that I'd really look forward to, but I'm skeptical that any or all of these will happen.
00:54:39 ◼ ► I mean, I don't think they would ever come out and say, hey, we're actually going to write documentation now.
00:54:46 ◼ ► And again, in their defense, they've been getting better and better over the years, but there's still a long way to go.
00:55:01 ◼ ► Speaking of documentation, like just one of the things that's been bothering me lately, because I've been really like sort of digging down to some really minor stuff in my apps because I'm not adding major features and just looking for like really obscure bugs and stuff.
00:55:14 ◼ ► God, can you imagine if Apple wrote in its documentation what threads things happen on in the functions that they provide and the APIs they provide?
00:55:23 ◼ ► Because it's really important for me to know, hey, hey, what thread is that callback going to be called on?
00:55:34 ◼ ► But if you check at runtime, it's like it seems like this is always called on the main thread.
00:55:41 ◼ ► That would be an example of where you would add documentation or give us a source code.
00:55:47 ◼ ► Another thing I'd like is, please, Apple, please, for the love of all that is good and holy in the world, can you make tab management in Xcode make any f***ing sense whatsoever?
00:56:04 ◼ ► Please, for the love of God, anything would work other than what you've got right now and the seven tries you made before it because, oh, my God, tabs in Xcode, they're the bane of my existence.
00:56:16 ◼ ► I don't think a lot of people are very enthusiastic about this, but it drives me nuts literally every time I do my job.
00:56:39 ◼ ► The only thing I do with tabs is I have my own custom keyboard shortcut, which I'm really regretting because they have no way to sync these.
00:56:45 ◼ ► But anyway, I have my own custom keyboard shortcut that I have to constantly change as they mess with the tab system, which adds a second pane with the same file that I'm editing to my right.
00:57:00 ◼ ► And then, you know, obviously, just in general, new APIs are fun, especially if they're, you know, mostly fully fleshed out.
00:57:09 ◼ ► But that being said, I would also love for Apple to just tell us, you know what, we've gone through and we have improved all the existing APIs.
00:57:16 ◼ ► You know, with the things that aren't async await friendly already now are, you know, I would love the proverbial snow leopard year for APIs as well.
00:57:40 ◼ ► Like what they added last year, you know, like the first year of, quote, Apple intelligence was for developers, nothing like we had nothing that we could possibly do for iOS 18.
00:57:50 ◼ ► Last year with iOS 26, they gave us a lot of APIs that I think overall look pretty good.
00:58:02 ◼ ► And I certainly haven't for the most part because we have liquid glass really took up a lot of our resources and a lot of our time.
00:58:09 ◼ ► And, you know, the one big one, obviously, that I did use was the speech transcriber API, which is great.
00:58:17 ◼ ► Like if you've if you've seen a podcast that uses the word billion in a sentence, how how the speech transcriber transcribes the word billion is one zero zero zero zero zero zero zero zero.
00:58:38 ◼ ► So you see this one giant, you know, this one followed by nine zero in the middle of a sentence.
00:58:46 ◼ ► I just got a notification today that said you're getting a call from one billion two hundred thousand.
00:58:55 ◼ ► Yeah, that's actually that's one of my longstanding gripes is like, you know, oftentimes I'm wearing AirPods when a message comes in and it will announce the message.
00:59:04 ◼ ► And it never knows what numbers are even like it'll be reading an address to me and it'll read the it'll read like the zip code as like, you know, one thousand five, you know,
00:59:15 ◼ ► And but the best is when it reads out tracking numbers from shipment notifications to you.
00:59:24 ◼ ► Anyway, those those whatever whatever part of machine learning or Siri is being used to announce messages to you in AirPods is really primitive.
00:59:38 ◼ ► This is a diversion from my other main topic because that whatever that API is that's generating that speech is really dumb and also unreliable.
00:59:48 ◼ ► Like still I will still like I'll have a message, you know, so and so has sent you a message, but I can't read it.
00:59:59 ◼ ► What what something broke or it'll say so and so sent you a message and it'll start reading it and it'll just stop like halfway through a sentence and just fail.
01:00:16 ◼ ► Anyway, but what I'm what I'm looking forward to for WVDC continued progress on the AI related APIs for developers last year was a great start.
01:00:29 ◼ ► We still are nowhere near the vision of Apple intelligence originally with like the app intent powered system to do all that stuff that we are we are constantly rumored to get it.
01:00:39 ◼ ► It's like next year is going to be the year of Linux on the desktop next year is going to be finally the the OS release that we get the app intent based Siri intelligence system.
01:00:49 ◼ ► You know, there's a lot still to do in terms of integrating AI with the system, giving developers ways to integrate AI into our own apps and making the AI that's tying all this together better.
01:01:09 ◼ ► So as that gets better, we are going to be able to make much better apps and our customers will be able to have much better experiences and our apps will be able to do better things with the system and with other apps through AI integrations and everything.
01:01:29 ◼ ► What else can we now do with AI on Apple's platforms without like bundling in our entire, you know, bundling in a three gig model with our apps, you know, like what if we just use Apple's platform things?
01:02:01 ◼ ► We're not going to, they're not going to throw it all in the trash and give us something that looks like iOS 18 again.
01:02:09 ◼ ► What I want is revision, improvement, iteration on the, on the concept of liquid glass.
01:02:17 ◼ ► The number one thing I want to see is there should never be a common pattern in Apple's stock UI that puts blurry text under other text I'm trying to read.
01:02:41 ◼ ► Navigation bars in iOS parlance are the bars on top of a screen with the title bar and a couple of buttons and text that scrolls under it.
01:02:47 ◼ ► Toolbars are the same thing at the bottom of the screen where you have buttons in the bottom.
01:03:01 ◼ ► And when most people have trouble with liquid glass, either conceptually or really like in terms of reading it, that's like the number one thing on the hit list is the lack of solid or mostly solid bars and having to blur text under other text.
01:03:22 ◼ ► With the design language they have chosen, that's going to require a decent amount of design to really to go back to some kind of defined bar with a defined edge.
01:03:35 ◼ ► And then the bar itself maybe is made of frosted glass or something, you know, and pretty heavily frosted, hopefully.
01:03:44 ◼ ► Like if they just made bars a defined area with a border that ended and a background that is mostly solid, you fix this problem.
01:03:56 ◼ ► But if you look in Mac OS 26.4, they did do something that you can see in system settings app.
01:04:03 ◼ ► Lots of people used to post, you know, screenshots of the system settings app, which has a search field on top of the left sidebar.
01:04:10 ◼ ► And as you scroll the left sidebar, the left sidebar has text items like general sound, whatever, that text would scroll underneath the search field.
01:04:18 ◼ ► And as soon as you scrolled anywhere other than at the very, very top, where basically where there was any text below the search field, the search field has placeholder text that says search.
01:04:27 ◼ ► And then you could also read whatever the item was that was underneath it from the sidebar that had scrolled up there.
01:04:37 ◼ ► In 26.4, the search field is a lot more frosty and you can no longer read the text underneath it.
01:04:48 ◼ ► You can see a black shadow scroll by as the text goes underneath it because it's real important that you do that for some reason.
01:04:57 ◼ ► So that's my guess of what they'll do in 27 is not actually have a defined bar, which I agree would be way better.
01:05:03 ◼ ► But my hopes are my expectations are set at the 26.4 level, which is like, no, they're not going to bring back bars because they're too stubborn and it would be too hard.
01:05:11 ◼ ► But what they will do is crank up the frost dial and make it so that you can't literally read text that's underneath controls.
01:06:00 ◼ ► And and you can keep like so much of the liquid glass design language is not necessarily at odds with that.
01:06:09 ◼ ► You could keep so much of how controls look toolbars, buttons, the animation, the way they blob up and blob back together.
01:06:17 ◼ ► And like so many of those things are totally doable on a mostly solid bar with a border.
01:06:28 ◼ ► Like there's so many different ways you could do it and you could make the bar frosted.
01:06:45 ◼ ► It just takes the will and maybe the humility to some degree to realize like, OK, this this direction of liquid glass has too many problems in real life.
01:06:57 ◼ ► So anyway, that's that's what I would hope to see is those those two big pillars, better AI APIs and liquid glass being more usable.
01:07:06 ◼ ► And I think I'll get probably both of those things, but probably neither one done to the level I want to be done.
01:07:18 ◼ ► I think to some degree there's, you know, what is the term term of phrase from Godfather?
01:07:40 ◼ ► But the thing that drives me nuts the most often is that I am looking directly through a toolbar and don't even know it.
01:08:10 ◼ ► DeleteMe makes it easy, quick, and safe to remove your personal data from hundreds of data broker sites at a time when surveillance and data breaches are common enough to make everyone vulnerable.
01:08:20 ◼ ► So if you search for your own name or anyone else's name on a common search engine on the web, what you're going to get is a listing of a bunch of sites offering to tell you that person's address, phone number, maybe even, like, social security number.
01:08:40 ◼ ► And what these sites are are data brokers, and they, you know, hoover up as much personal data as they can on as many people as possible.
01:08:46 ◼ ► That way, if somebody searches for their name on a search engine, they might come by and give them, you know, 30 bucks or whatever.
01:08:51 ◼ ► What DeleteMe does is automate the opting out and takedowns of your info from hundreds of these sites.
01:09:01 ◼ ► And so one person would never be able to keep up with all the hundreds of data broker sites out there and all their different opt-out procedures.
01:09:09 ◼ ► So what you do is you sign up for DeleteMe, and they, on an ongoing basis, will go around and make sure that your info is opted out and taken down of as many data broker sites as they can find.
01:09:20 ◼ ► As new sites crop up, they figure out their opt-outs, and they add them to the procedure.
01:09:28 ◼ ► I personally went around looking for one of these services a few years ago, and I found they were highly recommended.
01:09:32 ◼ ► In fact, the New York Times Wirecutter named DeleteMe their top pick for such services.
01:09:41 ◼ ► So take control of your data and keep your private life private by signing up for DeleteMe, and now at a special discount for our listeners.
01:09:47 ◼ ► Get 20% off your DeleteMe plan when you go to joindeliteme.com slash ATP and use promo code ATP at checkout.
01:09:54 ◼ ► The only way to get 20% off is to go to joindeliteme.com slash ATP and enter code ATP at checkout.
01:10:12 ◼ ► We haven't had the opportunity to do that in a little while, and we're starting with something that could have been written by Casey Liss, but was actually been written by Philip Miller, who writes,
01:10:23 ◼ ► When they first emerged, I took a wait-and-see approach, and that's where I remain today.
01:10:34 ◼ ► What are the biggest pitfalls of, or excuse me, what are the biggest benefits of Passkeys for someone like me, who's a faithful user of tools like 1Password for managing traditional passwords and two-factor codes?
01:10:44 ◼ ► Am I being foolish to worry about scenarios where my Apple devices are all lost in a house fire?
01:10:49 ◼ ► With 1Password, everything is backed up in the cloud, and I can keep my emergency logging credentials on a single sheet of paper in a safe deposit box.
01:11:02 ◼ ► My understanding is that you can share Passkeys, or maybe not originally, but certainly now.
01:11:07 ◼ ► But there's a lot of stuff that, you know, one of the great things about 1Password is that it has family vaults, or you can get a family account with a family vault, where, you know, Aaron and I will share a bunch, if not most, of our passwords.
01:11:23 ◼ ► I don't know, Marco, if you have thoughts on this, and let's start with you if so, but I'd very much like to hear John's perspective as well.
01:11:33 ◼ ► The way I've embraced Passkeys is when a website asks me, hey, you want to make a Passkey?
01:11:45 ◼ ► I have also found that, as I am still bouncing between Apple Passwords and 1Password, although more 1Password these days, since I'm still using Chrome as my default desktop browser, the way both of them handle Passkeys is fine.
01:12:03 ◼ ► With Apple, you go to a website, if it wants a Passkey, it'll show you the standard Touch ID, Face ID prompt, and you authenticate that way, and it gets you in.
01:12:12 ◼ ► With 1Password, if you have the browser extension, if it's unlocked already, it'll just show a little thing in the upper right corner, and it's like, do you want to use the Passkey?
01:12:26 ◼ ► The 1Password one is, I think, a little more convenient, just providing the security in a different way, but they're both fine.
01:12:32 ◼ ► My main issue with Passkeys is that it's kind of like that old XKCD comic that's like, we have too many standards.
01:12:47 ◼ ► But from the actual user experience of using them in practice most of the time, it's just another thing that gets added to the pile of things that we are supposed to keep track of and, you know, have in a password manager or something and, you know, try to share maybe with mixed success with family members or coworkers or whatever.
01:13:09 ◼ ► It can be done very well in an ideal implementation of Passkeys, both on, like, you know, the platform side and on, like, the website side.
01:13:21 ◼ ► The main downside with Passkeys so far is that the way that I see them being implemented by websites and services is not as a replacement for passwords, but as a replacement for two-factor.
01:13:35 ◼ ► So it's actually just taking this other complex thing and replacing it with a different complex thing.
01:13:42 ◼ ► Now, the security of a Passkey is way better than, you know, most of those other systems or alternatives to it.
01:13:59 ◼ ► Like, ideally, you could just have it, like, automatically, if you have a Passkey for the site you're on and it shows you a login screen, it just prompts you to use it and it logs you in all the way.
01:14:15 ◼ ► So that's kind of my kind of mixed opinion of them is that they are a great technology that solves a lot of problems.
01:14:28 ◼ ► And they are not really making logins on websites that much easier in practice, not because of their inherent technology, but because of the implementation decisions of those websites.
01:14:42 ◼ ► I was always trying them out whenever they were offered or whatever, but now having implemented them or having asked coding agents to implement them for me, more precisely, and seeing all the details there and all the different tradeoffs and everything.
01:14:59 ◼ ► I still think implementing them is the most annoying part of Passkeys because the input, like, if they want to get better adoption on websites and stuff, they should really, you know, streamline that and work harder to give, like, friendly libraries that are easy to use with good documentation, Casey.
01:15:22 ◼ ► They have a bunch of security advantages that passwords and two-factor stuff are never going to be able to match just because they're inherent in the technology.
01:15:29 ◼ ► The first is that when you use a password on a website, the server stores something on its side that could potentially compromise you.
01:15:41 ◼ ► You hope they don't just store your password in plain text, but as we've seen, some websites do that.
01:15:46 ◼ ► And you have no way of knowing if the website is incompetent in storing your passwords in plain text.
01:16:05 ◼ ► And all this is relevant because websites get hacked, and what people leak from them is account information, which is usually your email address, which, whatever, people don't like it to leak, but, you know, your email address is not a secret.
01:16:27 ◼ ► Maybe they can't get it now, but three years from now, they'll be able to do it with some NVIDIA GPU or something that can crack it.
01:16:37 ◼ ► Again, you're assuming that they're implemented in the same way, but I can tell you that PASCIs don't involve the server storing something that is secret.
01:16:52 ◼ ► You can give it to the world, and they use that to verify through this protocol that you're being signed in or whatever.
01:16:58 ◼ ► So if a website gets hacked, the only thing they can leak about you is your email address, which, again, I don't think is actually secret because spammers just spam every email address in the entire world.
01:17:14 ◼ ► Second thing is you don't have to make a human decision about where to enter something like a two-factor code or a password.
01:17:25 ◼ ► Yes, you're still vulnerable to, like, DNS poisoning or something else, but the point is social engineering is how people get hacked.
01:17:31 ◼ ► They throw up something in front of them that looks like a web page for the website they're familiar with, and they just blindly enter their password in there.
01:17:38 ◼ ► And, yes, password managers, like 1Password, which, by the way, is a current and former sponsor.
01:17:42 ◼ ► And the reason we recommend them is you should be using a password manager because they will automate a lot of this for you.
01:17:47 ◼ ► But you can always manually, and people know how to do it because they're like, oh, I don't, you know, the password autofill is not working or whatever, and people learn you can right-click and do autofill.
01:17:56 ◼ ► I've seen even technically non-savvy people go find their password in whatever their password thing is, usually Apple passwords or whatever, and copy and paste it into a text field.
01:18:06 ◼ ► People do that today because they have to, because it doesn't, you know, autofill isn't everywhere, or they don't know you can right-click and autofill on the Mac and recent versions of Mac OS.
01:18:16 ◼ ► So they make human decisions about where to put things, like their two-factor code and their password, which is why every two-factor code comes with an all-cap screaming at you, saying,
01:18:35 ◼ ► It's automated based on the identity of the server as verified by TLS, blah, blah, blah.
01:18:50 ◼ ► So this whole realm of social engineering is eliminated because humans don't, can't, literally can't, and don't decide where to put their credentials.
01:18:59 ◼ ► And that prevents you from screwing yourself when you get, like, phished or whatever, and you think it's a regular login screen.
01:19:06 ◼ ► And then finally, I think this is one of the big advantages that starts to shade into the social, which is, for most accounts these days, like, if it's something important, it's like, okay, you have a username and you have a good, you know, generated, long, complicated password that's stored in a password manager of your choosing, right?
01:19:29 ◼ ► And that second factor, sometimes they make it to SMS, which is incredibly insecure and websites demand that you do it.
01:19:35 ◼ ► But a lot of places will let you use an authenticator app, like Apple Passwords or Authy or 1Password or a million other apps that will, that are better than SMS.
01:19:44 ◼ ► And they need those two factors because there are so many other vulnerabilities in terms of phishing and getting your password cracked or reusing passwords and all sorts of stuff like that.
01:20:15 ◼ ► But multiple steps are annoying, which is why when I did all my personal sites, I use pass keys only because it is a one-step login process.
01:20:24 ◼ ► There's no second screen after you log in with a pass key where it should ask you for a two-factor.
01:20:27 ◼ ► Now, to Marco's point, lots of websites say, oh, no, we're going to use the pass keys as a second factor or we're going to make you pass key login and then ask you for an SMS code.
01:20:38 ◼ ► But the potential is, because they are secure enough for all the reasons that I said before, they can be the only factor which simplifies login.
01:20:46 ◼ ► So every website, every web service that I use that lets me only use a pass key to log in, I'm like, thank you.
01:21:00 ◼ ► You can choose to use a password, a pass key, or, you know, neither one of those, or an email login like you do all those things.
01:21:06 ◼ ► Now, there's still the question of like, okay, but if you can do email resets for forgot password, doesn't that make pass keys less secure, blah, blah, blah.
01:21:21 ◼ ► I'm embracing it with the websites that I maintain, that I build for myself, for my little toy things.
01:21:29 ◼ ► And I, you know, don't like it when they're not the only factor to log in, but they can be.
01:21:54 ◼ ► So you can have multiple pass keys to a website and it's, you know, everything's great.
01:21:58 ◼ ► And in terms of like the failure scenario, losing stuff, like, like the name one password, again, current and former sponsor, you can have one password that you actually remember to unlock your password vault.
01:22:15 ◼ ► Maybe that's the one password that you should actually memorize and have it be complicated.
01:22:29 ◼ ► This is one of the complaints I hear about pass keys is like, I can't just write it down like a password.
01:22:35 ◼ ► Well, first of all, you can write down your like Apple ID password, which is the keys to your, you know, iCloud keychain.
01:22:43 ◼ ► You can write down your one password password and put it in a safe deposit box or whatever.
01:22:49 ◼ ► But second of all, most sites, even if they do like pass key is your only login, will also give you backup codes,
01:22:56 ◼ ► which is like print these out and in case of fire, like put these in a safe deposit box in a remote location on a piece of paper.
01:23:03 ◼ ► And a lot of them will force you to prove that you wrote them down somewhere by asking you to enter them or whatever later.
01:23:17 ◼ ► I want to see that like the giant base 64, you know, like the pass key systems hide that from you as much as possible.
01:23:27 ◼ ► The reason they hide it is because they never want a user to be able to get at that because if users can get at that, people will fish it.
01:23:37 ◼ ► Go grab your private key from your password manager and paste it into this text field and people will do it.
01:23:42 ◼ ► So that's why they make it basically impossible for a regular person to get at that data.
01:23:47 ◼ ► But websites with accounts almost, especially if they're important and they're technically decent, almost always have backup codes, print out your backup codes, store them off site somewhere, have multiple copies of them in physically secure locations.
01:24:01 ◼ ► So if everything burns down and you lose every single one of your devices and you're like, how am I going to restore my digital world?
01:24:13 ◼ ► But if you really get hacked or something and you can't get your email, backup codes, backup codes are the answer.
01:24:24 ◼ ► But again, like I said, now having implemented them a few times and using them more every time I find a site that uses one like, yes, yes, more of this, just this one thing and nothing else.
01:24:40 ◼ ► But if the site you use makes you use it as a second factor or something like that, you just have to wait or maybe send them an email and say, hey, I wish you did something different.
01:24:48 ◼ ► John Fabridius writes, do you think that the eight gigs of RAM constraint in the MacBook Neo helps keep or make macOS lean a self-imposed limitation that can prevent lazy bloat?
01:25:07 ◼ ► I will say that I'm friends with a handful of Apple engineers and without disclosing, and they've never told me anything secret,
01:25:24 ◼ ► Like, you know, just, oh, I've been working on some perf stuff lately, and that's all they'll ever tell me.
01:25:33 ◼ ► And if you think about it, you know, if things perform well, and maybe not specifically around RAM, but if things perform well, that can also reduce, like, battery drain and strain and so on.
01:25:48 ◼ ► But I think Apple just genuinely really cares about this stuff, irrespective of the fact that they have low RAM devices still floating around in the world.
01:25:56 ◼ ► Yeah, I think the main thing to keep in mind here is that when you look at where RAM goes on a Mac, Mac OS is not what you have to worry about.
01:26:08 ◼ ► What you have to worry about is all the third-party apps, especially apps that use things like the web frameworks, like Electron and stuff like that.
01:26:24 ◼ ► And then all of the, you know, all the Electron-based apps, they just gobble up RAM like crazy.
01:26:34 ◼ ► You know, the OS does its best, but for the most part, that's what you have to worry about.
01:26:40 ◼ ► And now, that being said, like, you know, when you look at the, when you're designing an app like that,
01:26:45 ◼ ► you look at the installed base of what people actually use, and if computers that many people use are going to run that kind of app very slowly,
01:26:56 ◼ ► and it's going to bloat their computers up, you don't care, because the entire industry doesn't care.
01:27:02 ◼ ► You want to think that they would look at, you know, the entire field of computers out there and be like,
01:27:21 ◼ ► Not a single software developer ever cares about that kind of stuff when making things like, you know,
01:27:33 ◼ ► Yes, the 8GB RAM constraint in the Mac with Neo keeps macOS lean, but macOS wasn't your problem.
01:27:50 ◼ ► But the real question is, and this was the question when the Neo came out when we were discussing it.
01:27:53 ◼ ► If you're going to roll out a brand new computer in 2026 with 8GB of RAM, how long will macOS run on that computer?
01:28:04 ◼ ► Like, I don't think their Apple will feel constrained to keep macOS lean because they need it to run on the Neo.
01:28:12 ◼ ► I think they'll decide, based on wherever macOS is going, how long the Neo will be supported.
01:28:17 ◼ ► And I'm sure it'll be supported for years because, like, not being able to run on the Neo.
01:28:25 ◼ ► But, like, there is a certain point where they'll add some feature to macOS 31 or something.
01:28:41 ◼ ► And I think because the Neo has so little RAM, it is definitely in danger of being the first line of Macs to be dropped from a software revision.
01:28:52 ◼ ► Obviously, like, with the processor transition, like, the Intel ones get dropped or whatever.
01:28:57 ◼ ► Like, how far back they go on the MacBook Pros for various versions of macOS versus the MacBook Airs versus the Minis and stuff like that.
01:29:15 ◼ ► And I think there is a potential for the Neo to fall off the end of the supported list longer than, let's say, its contemporary M5 MacBook Air with 16 gigs.
01:29:41 ◼ ► But I have to think some of them have to because it's like you have all these grand vision.
01:29:45 ◼ ► And, you know, when we got the big Apple intelligence dividend where all the Macs went to 16 gigs, like, finally, finally, we broke out of that 8 gig jail and we fell back into it.
01:29:56 ◼ ► But part one of the things that is going to hold back, especially this first gen Neo is it arrives at a time where 8 gig is the especially with the RAM crisis and everything.
01:30:06 ◼ ► It arrives at a time where 8 gig is it's the only choice for that machine, given the prices and given the current RAM scenarios.
01:30:16 ◼ ► And currently, 8 gig is a little bit under what you would want from even a low end machine.
01:30:23 ◼ ► Now, when the next one comes out, presumably with 12 and the A19 Pro in it, 12 is a little bit better.
01:30:31 ◼ ► But it doesn't like I feel I feel like if Apple continues to use the A series chips and does not adjust their A series chips to be up to 16, like to try to keep, you know, because it's possible they'll do that because we all know the Macs get stuck at the base RAM for way too long.
01:30:55 ◼ ► In 10 years, we're all going to be complaining that the MacBook Air has 16 gigs of RAM in the base model because Apple just does not change the base RAM to keep up with the times.
01:31:12 ◼ ► Is it the case that the A series chips will end up with 18 while the Macs still have 16 because the phone just outraces them because the Macs stay static for so long?
01:31:22 ◼ ► But, like, we're currently in a situation where the A series has half of what the base Macs have, and that's probably the worst, I imagine, the worst it will ever be.
01:31:31 ◼ ► And I'm hoping that gap will narrow as Apple, again, if Apple keeps doing the A series chips on these, I think they'll start designing the A series chips, like, for the three years from now with the Mac in mind for, like, peripherals, like USB and stuff like that.
01:31:58 ◼ ► How long until we see an iPhone with an M chip or an iPad Pro with the ability to plug into a Thunderbolt dock or a studio display and present Mac OS to you?
01:32:11 ◼ ► Does the MacBook Neo now mean that Apple will likely never allow the iPhone to plug into a monitor and be used like a computer the way you can with an iPad or a Samsung Dex?
01:32:21 ◼ ► I just, I think this would be awesome, except I don't think it would be as awesome as I think it would.
01:32:46 ◼ ► I know, obviously, that's different now, but they certainly steadfastly refused it for years.
01:33:16 ◼ ► But it's weird that, like, it's a phone when it's in your hand, but it's a Mac when it's not.
01:33:26 ◼ ► And you kind of need to have power being supplied when you're hooked up to the monitor, too.
01:33:34 ◼ ► Even though this is exactly what I pitched in one of my early Macworld back page articles,
01:33:50 ◼ ► either they just want, like, an ultra-portable, like, Mac-like experience that goes everywhere with them.
01:34:05 ◼ ► What they're asking for, I think, is what Apple would hopefully deliver, well, I don't know,
01:34:31 ◼ ► And if you have that experience, you know, if that's working well, you don't need to plug your phone into a monitor.
01:34:38 ◼ ► All you need is to be able to just, like, log into a guest account on that monitor thing and just get all your stuff and use it kind of, like, more like a thin client or whatever.
01:34:51 ◼ ► And, like I said, I kind of agree with them because it's just weird to run iOS and macOS off the same device.
01:34:57 ◼ ► Setting aside the amount of disk space you're wasting and having to divide it up or whatever, it's just, it's not an efficient use of resources.
01:35:05 ◼ ► And it's a difficult product for people to, like, if you had this and you used it, you'd be like, yeah, but it is kind of weird that I run macOS and then I run the iPhone OS.
01:35:33 ◼ ► And Apple is pretty good, in most cases, Vision Pro, from keeping products out of the market that would end up that way.
01:35:58 ◼ ► This week on Overtime, we're going to be talking about how Apple is going to add touch support to macOS.
01:37:09 ◼ ► So I've been, I don't know if quiet quitting is really the right term of art slash phrase for what I've been doing,
01:37:41 ◼ ► If I can, I'll put a link in the show notes, but it's, I don't think it's made anymore.
01:37:50 ◼ ► But I have been working on trying to set myself up to have the sonology go away and be okay with it.
01:38:08 ◼ ► I think I have something like 10 to 15 terabytes worth of stuff, of active, like, stuff on the sonology.
01:38:14 ◼ ► You know, and then another, like, and about that much again in free space at the moment.
01:38:48 ◼ ► And honestly, that is the executive summary of what I'm about to tell you, which is that,
01:38:52 ◼ ► you know, a couple of years back, and we talked about it on the show, but a year or two back,
01:38:56 ◼ ► they said, okay, all future Synology, you know, NASAs, disk stations, as they call them,
01:39:03 ◼ ► they will be forced to use first-party Synology hard drives, which are themselves, you know,
01:39:12 ◼ ► The details honestly don't matter, even if they are legitimately their own first-party drives.
01:39:18 ◼ ► But this thing that was always and forever, you could put whatever drive you wanted in it,
01:39:28 ◼ ► But they've announced that, oh, no, no, no, no, no, you must use our first-party stuff.
01:39:39 ◼ ► It's far easier to guarantee that it'll work the way they want it to, and thus the way you want it to.
01:39:48 ◼ ► But I feel like, even though they are not taking it away from legacy users like myself,
01:40:07 ◼ ► In the same way that, all things being equal, I would put third-party RAM in a Mac, hi, John,
01:40:29 ◼ ► And to be clear, I think this will happen, but I don't think it's going to happen anytime soon.
01:40:52 ◼ ► And that's in no small part, because I think what I am likely to replace the Synology with,
01:41:10 ◼ ► I mean, obviously, there is a computer in there, but not really from a user's perspective.
01:41:32 ◼ ► Which, by the way, even despite having added third-party RAM to my Synology as an add-on rather than a replacement,
01:41:49 ◼ ► And maybe, you know, you listening might disagree with me, but that's kind of my perspective on it.
01:41:53 ◼ ► It's supposed to do basic serve-y stuff, but not be, like, your one server for your small business or for your home or whatever.
01:42:02 ◼ ► And so, what I've done is, I feel like I've talked about a little of this on the show, so forgive me if I'm repeating myself.
01:42:07 ◼ ► But if you recall, way back when, a year or two ago, I got a little tiny NUC, a little tiny Intel PC,
01:42:29 ◼ ► By scooping them up from the air and, you know, using the Absolutely Excellent Channels app,
01:42:41 ◼ ► That didn't end up working out, because my friend lives far enough into the middle of nowhere in Connecticut,
01:42:47 ◼ ► which is a thing, despite what you'd think, that his signal from his house was basically useless.