WEBVTT 00:00:00.000 --> 00:00:03.640 So imagine building a digital tool that is just 00:00:03.640 --> 00:00:07.799 so incredibly useful, so universally loved, that 00:00:07.799 --> 00:00:10.000 15 million people download it and use it every 00:00:10.000 --> 00:00:12.240 single day. Right, like you're basically the 00:00:12.240 --> 00:00:14.179 darling of the internet at that point. Exactly. 00:00:14.460 --> 00:00:16.859 But then, I mean, imagine realizing that this 00:00:16.859 --> 00:00:20.579 exact massive popularity is the precise thing 00:00:20.579 --> 00:00:23.780 driving your company into total financial ruin. 00:00:23.960 --> 00:00:26.879 Yeah, it's such a wild paradox. It really is. 00:00:27.079 --> 00:00:29.820 And today we're taking a deep dive into the source 00:00:29.820 --> 00:00:32.859 material behind a piece of software that lived 00:00:32.859 --> 00:00:36.310 that exact nightmare. It's basically the ultimate 00:00:36.310 --> 00:00:39.909 cautionary tale of the Web 2 .0 era. We're looking 00:00:39.909 --> 00:00:43.009 at a browser extension called X Marks Sync, which 00:00:43.009 --> 00:00:45.490 actually originally launched under the name Fox 00:00:45.490 --> 00:00:47.890 Marks. Oh, Fox Marks. Yeah. And if you're the 00:00:47.890 --> 00:00:49.609 kind of learner who really wants to understand 00:00:49.609 --> 00:00:51.509 the foundational architecture of the internet 00:00:51.509 --> 00:00:53.909 we use today, but, you know, without getting 00:00:53.909 --> 00:00:56.229 bogged down in endless technical jargon, this 00:00:56.229 --> 00:00:58.409 is definitely for you. Absolutely. Because their 00:00:58.409 --> 00:01:00.729 story, it isn't just about some old tool that 00:01:00.729 --> 00:01:02.890 saved your favorite websites. No, not at all. 00:01:03.070 --> 00:01:05.670 It's a master class. in the broader narrative 00:01:05.670 --> 00:01:09.510 of early cloud syncing, the frankly brutal reality 00:01:09.510 --> 00:01:12.230 of software monetization, and really the evolution 00:01:12.230 --> 00:01:14.370 of digital privacy. Yeah, the privacy aspect 00:01:14.370 --> 00:01:17.370 is huge, but to really grasp the stakes here, 00:01:17.709 --> 00:01:20.090 you have to mentally travel back to the digital 00:01:20.090 --> 00:01:23.930 landscape of like 2006. Oh man, 2006. Right. 00:01:24.290 --> 00:01:26.230 You probably remember what the early days of 00:01:26.230 --> 00:01:29.489 the modern web felt like. If you were using Firefox 00:01:29.489 --> 00:01:32.329 on your office computer and Internet Explorer 00:01:32.329 --> 00:01:35.670 on your home computer, those two ecosystems were 00:01:35.670 --> 00:01:38.670 totally, completely isolated. They were absolute 00:01:38.670 --> 00:01:41.950 walled gardens. Completely. Your data lived on 00:01:41.950 --> 00:01:44.730 the physical hard drive of the specific machine 00:01:44.730 --> 00:01:46.810 you were sitting in front of. There was just 00:01:46.810 --> 00:01:49.950 no concept of a continuous cloud -synced identity 00:01:49.950 --> 00:01:52.230 that followed you from device to device. Let's 00:01:52.230 --> 00:01:54.840 actually use an analogy here, because... managing 00:01:54.840 --> 00:01:57.480 your digital life back then was, well, it was 00:01:57.480 --> 00:01:59.760 a lot like maintaining a physical Rolodex. Oh, 00:01:59.819 --> 00:02:01.560 that's a great way to put it. Yeah. So imagine 00:02:01.560 --> 00:02:03.840 having this massive heavy Rolodex sitting on 00:02:03.840 --> 00:02:06.280 your desk at work. You find a fantastic website 00:02:06.280 --> 00:02:08.319 for your industry. You write down the address 00:02:08.319 --> 00:02:10.900 on a paper card and you slot it in. Very analog. 00:02:11.099 --> 00:02:13.659 Exactly. But then five o 'clock hits and you 00:02:13.659 --> 00:02:16.280 commute home. If you want to visit that exact 00:02:16.280 --> 00:02:18.439 same site from your living room, you either had 00:02:18.439 --> 00:02:21.460 to miraculously memorize the URL. Which nobody 00:02:21.460 --> 00:02:24.680 does. Right? Or you had to painstakingly duplicate 00:02:24.680 --> 00:02:27.919 that physical card into a second, completely 00:02:27.919 --> 00:02:29.919 separate Rolodex that you kept at your house. 00:02:30.120 --> 00:02:32.319 And the friction there was just immense. I mean, 00:02:32.460 --> 00:02:34.599 people would literally email themselves links 00:02:34.599 --> 00:02:36.860 just to bridge the gap between their own personal 00:02:36.860 --> 00:02:38.979 computers. I totally used to do that. Just an 00:02:38.979 --> 00:02:41.439 inbox full of random links. We all did. Because 00:02:41.439 --> 00:02:44.159 the browsers back then, like Internet Explorer, 00:02:44.400 --> 00:02:48.080 Mozilla Firefox, Apple Safari, they were designed 00:02:48.080 --> 00:02:51.020 by competing corporate entities. Right. Microsoft, 00:02:51.199 --> 00:02:53.599 Apple. Yeah. And they had absolutely no incentive 00:02:53.599 --> 00:02:56.340 to let their software communicate with a rival 00:02:56.340 --> 00:02:58.840 software. And that is the exact headache that 00:02:58.840 --> 00:03:01.419 a man named Mitch Kapoor decided to solve in 00:03:01.419 --> 00:03:04.199 2006 when he founded Foxmarks. And honestly, 00:03:04.539 --> 00:03:06.659 looking at the sheer scope of what they managed 00:03:06.659 --> 00:03:09.020 to synchronize, it feels like they pulled off 00:03:09.020 --> 00:03:11.830 a magic trick. It really does. Because they didn't 00:03:11.830 --> 00:03:14.650 just sync your bookmarks. They synced your passwords, 00:03:15.090 --> 00:03:18.270 your open tabs, and if you happen to be using 00:03:18.270 --> 00:03:21.150 Firefox, they even synced your entire browsing 00:03:21.150 --> 00:03:24.310 history across machines. Which is huge. And we 00:03:24.310 --> 00:03:26.289 really have to pause and look at the underlying 00:03:26.289 --> 00:03:29.310 mechanism here to appreciate why this was such 00:03:29.310 --> 00:03:31.590 a monumental achievement. Yeah, break that down 00:03:31.590 --> 00:03:34.710 for us. Today, companies like Google and Apple, 00:03:34.889 --> 00:03:37.949 they bake syncing directly into their operating 00:03:37.949 --> 00:03:40.250 systems, right? It's just a feature of their 00:03:40.250 --> 00:03:42.229 closed ecosystems. And you buy an iPhone, it 00:03:42.229 --> 00:03:44.870 just works with your Mac. Exactly. But back in 00:03:44.870 --> 00:03:47.969 2006, FoxMarks was a third party extension acting 00:03:47.969 --> 00:03:50.469 as this like universal translator between entirely 00:03:50.469 --> 00:03:53.009 different architectures. So getting Internet 00:03:53.009 --> 00:03:54.949 Explorer to somehow shake hands with Safari. 00:03:55.270 --> 00:03:57.669 Precisely. You're dealing with entirely different 00:03:57.669 --> 00:04:00.409 ways of storing data locally. Like one browser 00:04:00.409 --> 00:04:03.289 might use a proprietary database format, while 00:04:03.289 --> 00:04:06.009 another uses simple flat files or registry keys. 00:04:06.409 --> 00:04:09.210 So Foxmark's had to extract that data, securely 00:04:09.210 --> 00:04:11.969 transmit it to a central server, standardize 00:04:11.969 --> 00:04:14.610 it, and then perfectly inject it into a rival 00:04:14.610 --> 00:04:16.550 browser's architecture on the other side of the 00:04:16.550 --> 00:04:19.610 country. Yeah. It was a massive technical and 00:04:19.610 --> 00:04:22.129 cultural achievement to force these walled gardens 00:04:22.129 --> 00:04:26.009 into harmony. The users clearly recognized how 00:04:26.009 --> 00:04:28.350 revolutionary this was. I mean, the adoption 00:04:28.350 --> 00:04:30.850 rates detailed in the sources are just staggering. 00:04:30.930 --> 00:04:34.910 Oh, they exploded. By April of 2009, this little 00:04:34.910 --> 00:04:38.629 add -on was hitting 150 ,000 downloads every 00:04:38.629 --> 00:04:41.930 single week. That is wild for 2009. It really 00:04:41.930 --> 00:04:44.250 is. They eventually reached almost 15 million 00:04:44.250 --> 00:04:46.750 total downloads. Yeah, and the demand was so 00:04:46.750 --> 00:04:49.350 intense that, well, Opera wasn't officially supported 00:04:49.350 --> 00:04:51.470 by the company. Right, the Opera browser. Yeah, 00:04:51.629 --> 00:04:54.430 but users actually found a workaround using the 00:04:54.430 --> 00:04:56.970 Google Chrome extension just to force the software 00:04:56.970 --> 00:04:59.470 to function on Opera. I mean, people were desperate 00:04:59.470 --> 00:05:01.949 for this cross -browser harmony. And 2009 is 00:05:01.949 --> 00:05:03.829 also when they rebranded, right? Right. They 00:05:03.829 --> 00:05:05.990 dropped the Fox and FoxMarks since, you know, 00:05:05.990 --> 00:05:07.790 they were no longer just a Firefox tool and became 00:05:07.790 --> 00:05:11.680 XMarks. And they started rolling out sophisticated 00:05:11.680 --> 00:05:14.360 features, like suggested tagging for your bookmarks. 00:05:14.740 --> 00:05:17.519 Right. So they had essentially mapped out a core 00:05:17.519 --> 00:05:20.899 vital function of the modern internet years before 00:05:20.899 --> 00:05:22.699 the big tech giants got around to building it 00:05:22.699 --> 00:05:25.620 themselves. Exactly. But you would think 15 million 00:05:25.620 --> 00:05:28.459 active users is the ultimate victory lap. You'd 00:05:28.459 --> 00:05:31.240 think so. But if we look at the financial data 00:05:31.240 --> 00:05:34.279 in our sources, that exact number was basically 00:05:34.279 --> 00:05:37.540 a ticking time bomb. Which brings us to what 00:05:37.540 --> 00:05:39.879 I look at as the nine million dollar question. 00:05:40.120 --> 00:05:42.620 The infamous nine million dollar question. Yeah. 00:05:42.720 --> 00:05:45.779 We had this stratospheric high of user adoption, 00:05:46.139 --> 00:05:48.639 but then the company hits a sudden screeching 00:05:48.639 --> 00:05:51.540 halt. They basically slammed headfirst into the 00:05:51.540 --> 00:05:54.079 reality of server costs. I was looking at the 00:05:54.079 --> 00:05:56.420 financials in the source material, and it is 00:05:56.420 --> 00:05:58.959 honestly staggering to read in hindsight. The 00:05:58.959 --> 00:06:01.480 company was relying entirely on a voluntary donations 00:06:01.480 --> 00:06:04.399 funding model. Which is always a risky bet. Super 00:06:04.399 --> 00:06:07.879 risky. By October 2010, investors had sunk $9 00:06:07.879 --> 00:06:09.939 million of venture capital into the project. 00:06:10.110 --> 00:06:12.930 And apparently, there was no clear return on 00:06:12.930 --> 00:06:15.410 investment anywhere on the horizon. So on September 00:06:15.410 --> 00:06:19.750 27, 2010, Exmark makes this shocking announcement. 00:06:20.470 --> 00:06:22.709 They declare they're going to completely discontinue 00:06:22.709 --> 00:06:25.230 the service by January of the following year. 00:06:25.430 --> 00:06:27.410 Just boom, we're done. Yeah, they were just going 00:06:27.410 --> 00:06:29.629 to pull the plug. And the reaction from the tech 00:06:29.629 --> 00:06:32.389 community was an absolute explosion of panic. 00:06:33.370 --> 00:06:36.350 The CEO at the time, James Joaquin, noted he 00:06:36.350 --> 00:06:39.459 was I think his quote was pleasantly surprised 00:06:39.459 --> 00:06:42.279 by the massive volume of interest from both users 00:06:42.279 --> 00:06:45.279 and potential buyers, the second he announced 00:06:45.279 --> 00:06:47.939 the closure. Well, which leads to this incredibly 00:06:47.939 --> 00:06:50.199 desperate pivot. Yeah. They threw up a pledge 00:06:50.199 --> 00:06:52.379 page. Oh, the pledge page. Yeah, they essentially 00:06:52.379 --> 00:06:55.579 begged their user base, asking for 100 ,000 people 00:06:55.579 --> 00:06:59.160 to promise to pay $10 to $20 a year for a theoretical 00:06:59.160 --> 00:07:01.160 premium service. And they needed those pledges 00:07:01.160 --> 00:07:03.740 immediately, like... by October 15, just to keep 00:07:03.740 --> 00:07:05.600 the servers from going dark. Wait, let's unpack 00:07:05.600 --> 00:07:07.139 this for a second. Yeah. Because this is crazy 00:07:07.139 --> 00:07:09.439 to me. They had 15 million downloads. Right. 00:07:09.879 --> 00:07:12.579 They had $9 million in funding. Yet they still 00:07:12.579 --> 00:07:14.720 had to go hat in hand to their users for $10 00:07:14.720 --> 00:07:17.980 pledges. Was the internet simply allergic to 00:07:17.980 --> 00:07:20.740 paying for software back then? Well, it wasn't 00:07:20.740 --> 00:07:23.040 an allergy to paying. It was a fundamental shift 00:07:23.040 --> 00:07:25.339 in the architecture of how software operated. 00:07:25.860 --> 00:07:28.500 This is the ultimate trap of the freeware model 00:07:28.500 --> 00:07:31.089 when you apply it to the cloud. How so? We have 00:07:31.089 --> 00:07:32.949 to look at the difference between traditional 00:07:32.949 --> 00:07:37.550 90s desktop software and web 2 .0 cloud services. 00:07:37.889 --> 00:07:40.990 Like if I build a calculator application in 1998 00:07:40.990 --> 00:07:43.170 and you download it to your physical hard drive, 00:07:43.870 --> 00:07:46.970 my cost as a developer ends the very second your 00:07:46.970 --> 00:07:49.149 download completes. You can use that calculator 00:07:49.149 --> 00:07:51.490 a billion times and it cost me zero dollars. 00:07:51.649 --> 00:07:53.610 But cloud syncing is an entirely different beast. 00:07:53.870 --> 00:07:56.370 Exactly. With cloud software, the download is 00:07:56.370 --> 00:07:59.680 just the beginning of the expense. I see. time 00:07:59.680 --> 00:08:02.279 one of those 15 million users opened their browser, 00:08:02.639 --> 00:08:05.920 added a bookmark, or even opened a new tab, their 00:08:05.920 --> 00:08:08.759 computer sent a ping across the internet to the 00:08:08.759 --> 00:08:12.300 Exmark's servers. Wow. Every single tab. Yes. 00:08:12.699 --> 00:08:15.040 The server had to receive the data, process it, 00:08:15.160 --> 00:08:16.839 store it, and then push it out to all your other 00:08:16.839 --> 00:08:19.899 devices. So every single action taken by a user 00:08:19.899 --> 00:08:23.339 cost the company a tiny fraction of a cent in 00:08:23.339 --> 00:08:25.709 server bandwidth and compute power. And fractions 00:08:25.709 --> 00:08:27.490 of a cent multiply very quickly when you have 00:08:27.490 --> 00:08:29.889 15 million people generating them every single 00:08:29.889 --> 00:08:32.730 day. That's terrifying. It's the core lesson 00:08:32.730 --> 00:08:35.529 of SaaS software as a service unit economics. 00:08:36.250 --> 00:08:38.490 In cloud services, if you don't have a hardened 00:08:38.490 --> 00:08:41.990 monetization strategy from day one, massive success 00:08:41.990 --> 00:08:44.169 is actually a liability. Because you literally 00:08:44.169 --> 00:08:46.350 can't afford the popularity. Right. Every new 00:08:46.350 --> 00:08:49.110 user XMark celebrated meant more server space, 00:08:49.250 --> 00:08:51.649 more database maintenance, more bandwidth bills. 00:08:52.029 --> 00:08:55.299 They grew so incredibly fast that Their greatest 00:08:55.299 --> 00:08:58.159 triumph became the exact mechanism driving them 00:08:58.159 --> 00:09:00.879 to bankruptcy. So they were literally suffocating 00:09:00.879 --> 00:09:02.659 under the weight of their own popularity. Exactly. 00:09:02.879 --> 00:09:04.899 But the story doesn't end with a total blackout. 00:09:05.259 --> 00:09:07.720 Just before the clock runs out, on December 2, 00:09:07.860 --> 00:09:11.220 2010, the password management service LastPass 00:09:11.220 --> 00:09:14.580 swoops in and acquires X marks. Right under the 00:09:14.580 --> 00:09:16.779 wire. Yeah. And they end up launching that premium 00:09:16.779 --> 00:09:20.840 service a week later on December 9. So crisis 00:09:20.840 --> 00:09:23.960 averted, right? The tool is saved. Well, financially 00:09:23.960 --> 00:09:27.039 saved, yes. But salvation in the corporate tech 00:09:27.039 --> 00:09:29.139 world doesn't always translate to innovation. 00:09:30.500 --> 00:09:33.799 And that brings us to a period from 2011 to 2018 00:09:33.799 --> 00:09:37.000 that we can really only describe as the zombie 00:09:37.000 --> 00:09:39.679 years. The zombie years, yeah. Perfect name for 00:09:39.679 --> 00:09:42.480 it. The post -acquisition reality for X marks 00:09:42.480 --> 00:09:45.539 was pretty grim. According to the source material, 00:09:45.659 --> 00:09:48.320 the program was minimally updated. It received 00:09:48.320 --> 00:09:51.279 baseline bug fixes to keep it functional. but 00:09:51.279 --> 00:09:53.860 it did not get a single new feature added for 00:09:53.860 --> 00:09:55.899 the rest of its existence. It was essentially 00:09:55.899 --> 00:09:58.200 put on life support. I mean, LastPass likely 00:09:58.200 --> 00:10:00.440 bought them primarily for their massive user 00:10:00.440 --> 00:10:03.019 base and the underlying synchronization technology. 00:10:03.120 --> 00:10:05.159 Right, to integrate into their own stuff. Exactly. 00:10:05.220 --> 00:10:08.039 They could use it to bolster their own core password 00:10:08.039 --> 00:10:11.980 management ecosystem. But as a standalone, evolving 00:10:11.980 --> 00:10:15.440 product, Xbox was no longer the priority. It 00:10:15.440 --> 00:10:17.879 makes me think of someone buying a beautiful 00:10:17.879 --> 00:10:20.440 vintage classic car, but instead of driving it, 00:10:20.539 --> 00:10:22.539 they just park it in a garage under a heavy tarp. 00:10:22.919 --> 00:10:25.399 That's a sad image, but yeah. They go in there 00:10:25.399 --> 00:10:28.080 maybe once a year to give it an oil change, which 00:10:28.080 --> 00:10:30.620 is the bug fix, just to keep the engine from 00:10:30.620 --> 00:10:32.720 completely seizing up and rusting out. Right. 00:10:33.159 --> 00:10:35.120 But they never take it out on the open road again. 00:10:35.200 --> 00:10:38.139 It just sits there, slowly aging out of relevance. 00:10:38.299 --> 00:10:40.919 And the broader corporate timeline reflects that 00:10:40.919 --> 00:10:44.820 perfectly. In October 2015, LastPass itself gets 00:10:44.820 --> 00:10:48.000 acquired by a much larger entity, LogMeIn Inc., 00:10:48.000 --> 00:10:54.100 for $125 million. $125 million. Wow. Yeah. And 00:10:54.100 --> 00:10:56.179 at the time, LogMeIn released a statement saying 00:10:56.179 --> 00:10:58.899 they plan to continue supporting X marks. But, 00:10:58.980 --> 00:11:01.590 you know. The clock was always ticking. Right. 00:11:01.769 --> 00:11:04.509 On March 30th, 2018, LastPass sends an email 00:11:04.509 --> 00:11:06.909 to the remaining users announcing the final, 00:11:06.929 --> 00:11:10.490 absolute shutdown of XMarkz on May 1st, 2018. 00:11:10.789 --> 00:11:12.450 But wait, before we leave these zombie years 00:11:12.450 --> 00:11:14.690 entirely, I have to bring up a bizarre tangent 00:11:14.690 --> 00:11:17.210 from the source material that left me completely 00:11:17.210 --> 00:11:18.830 bewildered. Oh, I know exactly what you're going 00:11:18.830 --> 00:11:22.210 to say. It's so weird. In May 2012, while XMarkz 00:11:22.210 --> 00:11:24.269 is quietly sitting in this maintenance mode, 00:11:24.750 --> 00:11:27.490 the xmarks .com domain was suddenly blocked by 00:11:27.490 --> 00:11:30.080 major internet service providers in India. by 00:11:30.080 --> 00:11:32.700 a court order. Yeah, just out of nowhere. Users 00:11:32.700 --> 00:11:35.120 in India were just getting a generic screen saying 00:11:35.120 --> 00:11:38.240 access to this site has been blocked as per court 00:11:38.240 --> 00:11:41.059 orders. And according to the sources, X marks 00:11:41.059 --> 00:11:43.440 was caught up in a sweep that blocked various 00:11:43.440 --> 00:11:46.299 torrent and video sites like the Pirate Bay, 00:11:46.539 --> 00:11:50.299 Vimeo and Dailymotion. It's crazy. I just how 00:11:50.299 --> 00:11:54.200 on earth does a boring utility bookmark syncing 00:11:54.200 --> 00:11:57.399 tool get lumped in with global piracy hubs by 00:11:57.399 --> 00:11:59.750 a high court? I know. It sounds like a surreal 00:11:59.750 --> 00:12:02.110 administrative typo, but it actually exposes 00:12:02.110 --> 00:12:04.549 the mechanical reality of how internet censorship 00:12:04.549 --> 00:12:06.690 operates on a technical level. OK, I need you 00:12:06.690 --> 00:12:08.809 to walk me through this. So this is a prime example 00:12:08.809 --> 00:12:11.809 of the collateral damage caused by broad blunt 00:12:11.809 --> 00:12:15.169 instrument sweeps by ISP's internet service providers. 00:12:15.470 --> 00:12:17.649 But how does the internet mix up a bookmark tool 00:12:17.649 --> 00:12:20.450 with a torrent site? It comes down to IP addresses 00:12:20.450 --> 00:12:23.269 and content delivery networks. When a court orders 00:12:23.269 --> 00:12:26.370 an ISP to block access to a piracy site, they 00:12:26.370 --> 00:12:28.830 rarely have the search precision to block just 00:12:28.830 --> 00:12:31.750 the specific URL. Instead, they often block the 00:12:31.750 --> 00:12:34.190 entire IP address associated with that site's 00:12:34.190 --> 00:12:37.169 servers. But modern websites don't usually sit 00:12:37.169 --> 00:12:40.190 on their own isolated servers anymore. They use 00:12:40.190 --> 00:12:43.470 shared hosting providers or CDNs to manage traffic. 00:12:43.690 --> 00:12:45.909 Oh, I see. So it's like a massive digital apartment 00:12:45.909 --> 00:12:48.700 building. Exactly. That's it. Perfectly. The 00:12:48.700 --> 00:12:51.460 Pirate Bay might live in apartment 4B, and Exmarks 00:12:51.460 --> 00:12:53.919 might just be renting apartment 4C down the hall. 00:12:54.139 --> 00:12:56.600 Right? They share the same street address, the 00:12:56.600 --> 00:12:59.360 same IP block. When the ISP receives the court 00:12:59.360 --> 00:13:02.139 order, they don't knock on the door of 4B. They 00:13:02.139 --> 00:13:04.779 just demolish the entire building. Wow. They 00:13:04.779 --> 00:13:07.639 sinkhole the routing path for that whole IP block. 00:13:07.940 --> 00:13:10.019 So Exmarks wasn't targeted for its content. It 00:13:10.019 --> 00:13:12.419 was essentially just standing too close to the 00:13:12.419 --> 00:13:14.360 wrong crowd in the eyes of the network infrastructure. 00:13:14.679 --> 00:13:17.279 That is fascinating and honestly deeply concerning 00:13:17.279 --> 00:13:19.440 when you think about how fragile digital access 00:13:19.440 --> 00:13:21.559 really is. It really is fragile. But zooming 00:13:21.559 --> 00:13:24.980 back out to the eventual shutdown in 2018, it 00:13:24.980 --> 00:13:28.659 seems inevitable, right? I mean, Logman buys 00:13:28.659 --> 00:13:32.600 LastPass for over $100 million entirely to acquire 00:13:32.600 --> 00:13:34.480 its enterprise security business? Yeah, that 00:13:34.480 --> 00:13:37.779 was the real price. And by 2018, browsers like 00:13:37.779 --> 00:13:41.100 Google Chrome and Apple Safari had completely 00:13:41.100 --> 00:13:43.360 perfected their own native built -in sinking 00:13:43.360 --> 00:13:45.879 anyway. Right. To a massive corporation like 00:13:45.879 --> 00:13:49.360 LogMain, Exmark wasn't a pioneering revolutionary 00:13:49.360 --> 00:13:52.039 tool anymore. It was a redundant, obsolete piece 00:13:52.039 --> 00:13:55.200 of legacy code that literally just cost money 00:13:55.200 --> 00:13:57.279 to maintain. So pruning it from the portfolio 00:13:57.279 --> 00:13:59.039 was just a simple matter of corporate accounting. 00:13:59.360 --> 00:14:01.700 Exactly. It is so ironic, though, that LogMain 00:14:01.700 --> 00:14:05.059 viewed Exmark's as an obsolete piece of the puzzle 00:14:05.059 --> 00:14:08.379 by 2018, because if you look at how Exmark's 00:14:08.379 --> 00:14:10.559 actually handled user data back in their prime, 00:14:11.200 --> 00:14:14.000 they were pioneering concepts around digital 00:14:14.000 --> 00:14:16.980 privacy that we are still fiercely debating today. 00:14:17.019 --> 00:14:19.519 Absolutely. I mean, before data privacy became 00:14:19.519 --> 00:14:22.600 this massive regulated global buzzword, how did 00:14:22.600 --> 00:14:26.080 a company in 2006 handle holding the keys to 00:14:26.080 --> 00:14:28.559 the entire browsing history of 15 million people? 00:14:28.720 --> 00:14:30.779 Well, they offered a really fascinating mix of 00:14:30.779 --> 00:14:33.059 practices. Some were incredibly forward thinking 00:14:33.059 --> 00:14:35.200 and empowering for the user, while others were 00:14:35.200 --> 00:14:37.519 honestly the exact blueprint for the data mining 00:14:37.519 --> 00:14:40.039 economy we live in today. Let's get into that, 00:14:40.039 --> 00:14:43.000 because looking at the sources, Exmark's provided 00:14:43.000 --> 00:14:46.000 an option to encrypt bookmarks while they were 00:14:46.000 --> 00:14:48.759 in transit between the user's browser and the 00:14:48.759 --> 00:14:50.860 central server. Which was great security for 00:14:50.860 --> 00:14:54.320 the time. Yeah. And their privacy policy explicitly 00:14:54.320 --> 00:14:57.340 stated that individual users' bookmarks were 00:14:57.340 --> 00:14:59.659 kept private from other users. Like you couldn't 00:14:59.659 --> 00:15:01.940 just look up what your neighbor was reading. 00:15:02.200 --> 00:15:04.840 Right. The individual files were siloed. However, 00:15:05.059 --> 00:15:07.340 they did analyze all those private bookmarks 00:15:07.340 --> 00:15:11.179 in aggregate. Yes. That aggregated data was the 00:15:11.179 --> 00:15:13.879 fuel for their secondary public features. Exactly. 00:15:14.120 --> 00:15:16.240 They took millions of user bookmarks and fed 00:15:16.240 --> 00:15:18.860 them into features like suggested tags and this 00:15:18.860 --> 00:15:21.600 tool called Smarter Search. Oh, Smarter Search. 00:15:21.740 --> 00:15:23.799 Yeah, which actually actively enhanced Google 00:15:23.799 --> 00:15:26.580 web search results based on what Exmark's users 00:15:26.580 --> 00:15:28.860 were bookmarking. It was essentially crowd -sourced 00:15:28.860 --> 00:15:31.519 SEO. Right. Now, the privacy policy promised 00:15:31.519 --> 00:15:33.620 that the results of this analysis were published 00:15:33.620 --> 00:15:36.080 without providing any identifying information 00:15:36.080 --> 00:15:38.679 about individual users. Right. It was anonymous. 00:15:38.509 --> 00:15:41.590 But I want to push back on this model for a second. 00:15:42.889 --> 00:15:45.669 Is analyzing our private, personal bookmarks 00:15:45.669 --> 00:15:49.690 to feed a public search algorithm an early breach 00:15:49.690 --> 00:15:53.470 of trust? Or was Exmark's just pioneering the 00:15:53.470 --> 00:15:55.809 exact same data for convenience trade -off that 00:15:55.809 --> 00:15:58.750 we all blindly accept every time we download 00:15:58.750 --> 00:16:01.490 a free app on our phones today? It is definitely 00:16:01.490 --> 00:16:03.929 the latter, without question. The Smarter Search 00:16:03.929 --> 00:16:06.289 feature is the foundational economics of the 00:16:06.289 --> 00:16:08.250 modern internet. Give me convenience. Take my 00:16:08.250 --> 00:16:11.190 data. Exactly. You receive a highly convenient, 00:16:11.470 --> 00:16:13.750 frictionless service for free. And in exchange, 00:16:13.929 --> 00:16:15.970 the company aggregates your behavioral data to 00:16:15.970 --> 00:16:18.509 build and improve other products. Wow. It's the 00:16:18.509 --> 00:16:20.690 classic realization that if you are not paying 00:16:20.690 --> 00:16:23.230 for the product, you are the product. It's so 00:16:23.230 --> 00:16:25.549 true. Yeah. But the sources highlight another 00:16:25.549 --> 00:16:27.730 feature they offer that runs completely counter 00:16:27.730 --> 00:16:29.710 to that whole data mining philosophy. Yeah, the 00:16:29.710 --> 00:16:32.389 buy OS option. Yes. Yes. BIOS, bring your own 00:16:32.389 --> 00:16:35.029 server. When I read how this functioned, it genuinely 00:16:35.029 --> 00:16:37.009 blew my mind because it just feels impossible 00:16:37.009 --> 00:16:39.230 in today's tech climate. The BIOS feature is 00:16:39.230 --> 00:16:42.250 a magnificent relic of a much more open Internet. 00:16:42.309 --> 00:16:45.250 It really is. To alleviate the very real trust 00:16:45.250 --> 00:16:47.649 issues users had about sending their private 00:16:47.649 --> 00:16:50.690 digital maps to a third party corporate server, 00:16:51.570 --> 00:16:55.090 Exmark's actually allowed Firefox users to entirely 00:16:55.090 --> 00:16:58.299 bypass the Exmark's infrastructure. Okay, let's 00:16:58.299 --> 00:17:00.559 break down how that actually worked for the user, 00:17:01.059 --> 00:17:03.200 because bring your own server sounds incredibly 00:17:03.200 --> 00:17:05.859 technical. Well, it required some technical literacy, 00:17:06.119 --> 00:17:09.380 sure, but the mechanism was brilliant. You could 00:17:09.380 --> 00:17:12.880 configure the Exmark's browser extension to completely 00:17:12.880 --> 00:17:15.680 ignore the company's central cloud. Instead, 00:17:15.759 --> 00:17:17.940 you could point the extension to sync your data 00:17:17.940 --> 00:17:21.140 using your own personal WebDV or FTP server. 00:17:21.440 --> 00:17:24.380 Meaning I could set up an FTP file transfer protocol 00:17:24.380 --> 00:17:26.880 server on like an old computer sitting in my 00:17:26.880 --> 00:17:28.839 own basement. Yeah, exactly. I own the hardware, 00:17:29.039 --> 00:17:31.359 I control the Internet connection, and the Exmark's 00:17:31.359 --> 00:17:34.029 extension would just quietly route my data directly 00:17:34.029 --> 00:17:36.480 to my basement? instead of a corporate server 00:17:36.480 --> 00:17:39.559 farm. Precisely. WebDavy, which stands for Web 00:17:39.559 --> 00:17:41.859 Distributed Authoring and Virgining. Titty name. 00:17:42.059 --> 00:17:44.619 I know, right? It's basically an extension of 00:17:44.619 --> 00:17:46.880 the internet standard protocol that allows you 00:17:46.880 --> 00:17:49.220 to securely manage files on a remote server. 00:17:49.539 --> 00:17:52.259 By giving users the ability to input their own 00:17:52.259 --> 00:17:55.660 WebDavy or FTP credentials, xMarks was essentially 00:17:55.660 --> 00:17:58.960 saying, hey, we built this beautiful seamless 00:17:58.960 --> 00:18:01.500 software interface, but we don't demand to hold 00:18:01.500 --> 00:18:04.200 your data hostage. You can keep the actual files 00:18:04.200 --> 00:18:07.170 on your own hardware. Try asking a major tech 00:18:07.170 --> 00:18:09.990 giant today to let you use their pristine software 00:18:09.990 --> 00:18:13.150 interface, but demand that all your data be stored 00:18:13.150 --> 00:18:15.829 on a hard drive in your closet instead of their 00:18:15.829 --> 00:18:18.289 proprietary cloud. They would laugh you out of 00:18:18.289 --> 00:18:20.569 the room. They really would. We live in an era 00:18:20.569 --> 00:18:23.950 of mandatory walled gardens. Completely. Exmark's 00:18:23.950 --> 00:18:27.170 existed in this brief utopian window where software 00:18:27.170 --> 00:18:29.470 developers were actually still willing to hand 00:18:29.470 --> 00:18:32.430 users the keys to their own data infrastructure. 00:18:32.799 --> 00:18:35.559 It really demonstrates that the centralized data 00:18:35.559 --> 00:18:38.440 -hungry cloud we have today wasn't the only possible 00:18:38.440 --> 00:18:40.200 outcome for the internet. It was a choice made 00:18:40.200 --> 00:18:42.720 by the industry, mostly driven by the monetization 00:18:42.720 --> 00:18:44.880 problems we discussed earlier. So synthesizing 00:18:44.880 --> 00:18:46.640 everything we've pulled from the sources today, 00:18:47.000 --> 00:18:49.559 the story of XMarts really is the ultimate tragedy 00:18:49.559 --> 00:18:51.839 of the early web. It really is. It was a tool 00:18:51.839 --> 00:18:55.519 that brilliantly solved a massive universal friction 00:18:55.519 --> 00:18:58.980 point. It brought cross -platform harmony to 00:18:58.980 --> 00:19:02.599 our chaotic digital lives. But the fundamental 00:19:02.599 --> 00:19:05.680 unit economics of cloud computing meant it grew 00:19:05.680 --> 00:19:08.420 too fast to afford its own success. Success was 00:19:08.420 --> 00:19:10.500 fatal. Right. It was swallowed up by a larger 00:19:10.500 --> 00:19:12.880 entity that only wanted its underlying technology, 00:19:13.380 --> 00:19:16.119 left to languish in a zombie state, and eventually 00:19:16.119 --> 00:19:18.880 shut down once the major browsers just copied 00:19:18.880 --> 00:19:20.960 its functionality directly into their own code. 00:19:21.279 --> 00:19:23.980 It was the pioneer that painstakingly carved 00:19:23.980 --> 00:19:26.519 the path through the wilderness only to be, you 00:19:26.519 --> 00:19:29.500 know, paved over by the massive big tech hype. 00:19:29.319 --> 00:19:31.740 ways we all commute on today. That is a perfect 00:19:31.740 --> 00:19:33.920 way to put it. OK. Which leaves us with a lingering 00:19:33.920 --> 00:19:35.920 question for you to mull over as you go about 00:19:35.920 --> 00:19:38.460 your day, opening tabs and watching your devices 00:19:38.460 --> 00:19:41.220 seamlessly sync without a second thought. Yeah, 00:19:41.220 --> 00:19:44.339 think about this. If a universally loved, critically 00:19:44.339 --> 00:19:47.440 acclaimed tool like Exmark's, a company that 00:19:47.440 --> 00:19:49.700 literally held the map to the daily habits of 00:19:49.700 --> 00:19:53.039 15 million people, could not survive with a $9 00:19:53.039 --> 00:19:55.700 million investment. Right. What does that say 00:19:55.700 --> 00:19:57.940 about the true cost of the free internet we rely 00:19:57.940 --> 00:20:01.230 on right now? It's a scary thought. Is any application 00:20:01.230 --> 00:20:03.869 actually free, or are we all just enjoying a 00:20:03.869 --> 00:20:06.589 temporary illusion of scale, blissfully unaware, 00:20:07.289 --> 00:20:09.309 until the eventual shutdown notice arrives in 00:20:09.309 --> 00:20:12.269 our inbox? It certainly makes you view... the 00:20:12.269 --> 00:20:14.390 invisible infrastructure of your digital life 00:20:14.390 --> 00:20:17.089 for a much sharper lens. It absolutely does. 00:20:17.470 --> 00:20:19.390 Thank you for joining us on this deep dive into 00:20:19.390 --> 00:20:22.150 the source material. We love dissecting these 00:20:22.150 --> 00:20:24.329 incredible tech histories with you, peeling back 00:20:24.329 --> 00:20:26.750 the polished surface of the tools we use every 00:20:26.750 --> 00:20:29.430 day to understand the hidden mechanics underneath. 00:20:30.009 --> 00:20:32.430 Until next time, keep questioning the magic behind 00:20:32.430 --> 00:20:32.869 the screen.