The Development Disaster behind macOS

The Development Disaster behind macOS

April 30, 2019 48 By Bernardo Ryan



coming in at second place at nearly 13% of the desktop computer market Apple's Mac OS is a reasonably popular operating system with plenty of dedicated fans behind it financially Apple is doing better than it ever has before but in the 90s that couldn't have been further from the truth between increasingly unpopular products like the Newton in the Pippin and the continued rise in popularity of Wintel computers it seems like all of their product plans for collapsing at once and even apples cash cow the Macintosh wasn't immune to really understand the issues with the Mac OS though we have to go back at least 10 years earlier to the introduction of the Lisa being the first computer Apple released with a graphical interface the machine was quite the powerhouse with a 5 megahertz processor dual internal floppy drives in 1 megabyte of RAM of course the machine also came with a $10,000 price tag over $25,000 today needless to say the Lisa did not do well but a very similar project in Apple was able to see some success the Macintosh was designed to be a more affordable Lisa and had a quarter of the price it sold almost as many units as the Lisa had in its entire lifetime in just one month of course to drop the cost as much as the Mac did a few corners had to be cut the Mac for example had only a slightly faster processor than Louisa a single disc drive at less than half the capacity of the Lisa's twiggy drives and probably most notably only a hundred and twenty eight K of RAM to pull this off the Macintosh operating system needed to be pared down as well so features like protected memory and multitasking had to go even a few years and a few increasingly powerful Macintosh models the Mac had for the most part caught up hardware wise with the Lisa but with the exception of the incredibly limited desk accessories the Mac OS wasn't really designed to run more than one program at a time the solution hacked together by Andy Hertzfeld was inspired by a similar program found on the IBM PC switcher eventually called multi finder would allow the user to load and unload programs from memory so that by clicking in the top corner of the screen the user could quickly switch from one running application to another it didn't require much modification to the Mac operating system but still managed to achieve true multitasking Apple included it with system 5 and onward it's worth noting that in the mid to late 80s Apple is constantly pushing out updates to the Mac operating system to the point where a new version was essentially an annual occurrence besides having things like multi finder though the updates from mostly bug fixes and adding support for new hardware pulling the release of system 6 in 1988 things began to slow down and it would be another three years until 1991 when system seven a major overhaul to the Mac OS would come up introducing among other important features a cooperative multitasking model one of the consequences of a cooperative multitasking model is that the programs in and have to well cooperate that is the operating system gives complete control of the computer to a program with the expectation that that control will eventually be given back theoretically this works as each program takes turns running on the processor but this model completely falls apart at the moment one program refuses to give up control usually because it's crashed before it could return to do s because the computer is stuck in the crashed program and the OS can't regain control none of the other programs can run either so the moment one program freezes it brings the entire system down with it this wasn't much of a problem at first but as programs became more complex and better Hardware allowed users to run more programs at once the chances of a system crashing bug occurring became greater and greater the Mac OS wasn't the only graphical operating system on the market by this time either let alone the only multitasking one Windows had started out in 1985 as a crude graphical shell to dos and while the second version and even more so Windows 3 in 1990 expanded it into a more fully featured operating system there were still plenty of inconsistencies unintuitive controls and legacy specifications that made it look pretty clunky compared to Mac OS Microsoft recognized this of course and after catching up significantly with Windows 3.1 hope to close the gap completely in the next few years with a little project called Chicago what would eventually be known as Windows 95 management at Apple was an oblivious to these issues though and as far back as 1988 a team had gathered together to discuss the future evolution of the Mac OS easy short-term improvements to the OS like adding color were written on blue index cards while more advanced features like adding a new more stable pre-emptive the tasking system went up to the pink index cards was extremely long-term plans going onto the red cards the pink and blue goals were eventually given to the pink and blue teams which will work in parallel to produce the next two versions of Mac OS all the ideas from the blue team became the massive overhaul system seven that I mentioned earlier but things didn't quite go so easily for the pink team in 1991 the same year that system 7 had shipped CEO John Sculley have made an offer that would have largely been considered heresy ten years earlier and reached out to IBM to work together on Apple's next-generation operating system a project that came to be known as telogen since the pink team in wisdom the next Mac OS had moved mostly outside of the company not much was being done regarding the future of the operating system with most improvements being through patches by the blue team that continued distress the already unstable system 7 of course letting the OS become less and less stable wasn't much of a concern since telogen would fix it soon enough right wrong the telogen project itself was a disaster specifically because of fighting between the Apple and IBM employees over the suspicion that IBM would end up taking controller for the project from Apple what would become a self-fulfilling prophecy as Apple backed out of the deal in 1995 an IBM finished telogen took less than mediocre reception in there Apple was with an increasingly unstable operating system on one hand and therefore years of backup work down the drain they needed a replacement OS fast and in doing so cooked up an infamous little project known as Copeland code named after the composer Aaron Copeland claiming to have essentially a complete OS rewrite out for developers in 1995 ready for public release the following year the task seemed impossible and that's because at face value it pretty much was instead the first stage of Copeland was just to introduce a kernel to the OS adding in pre-emptive multitasking instead of the unstable in-place cooperative model at the same time to avoid having to rewrite all the existing system code and libraries all old Macintosh code will be run in a virtual memory location called the blue box after that release ship I would rewrite all the blue box code and transition fully to the new platform of course just like with Talan things weren't quite so simple but this time the problem was mostly within Apple Copeland was the project to be on at Apple so it wasn't uncommon for developers from side projects that had started in Italian days to want to leave those projects to be involved with Copeland the managers in interest of self-preservation attempted to counter this practice by adding their projects to Copeland that way nobody left and the developers would still be happy to be working on a piece of Copeland as 1995 went on this practice became more and more popular as the scoop of Copeland became more and more bloated the next year the Worldwide Developers Conference around the time Apple would plan to have Copeland publicly released they showed a demo of McIntosh system 8 it wasn't pretty most of the OS liked basic interactions like text editing the system was so fragile demo staff had to be on hand just to replace hard drives that the system corrupted during a crash said one commentator it was incredible they even let us see the beast as if Copeland wasn't already a victim of feature creep from internal politics at Apple the fact that the system was already late prompted Apple to justify that by promising more features but the then CEO Gil Amelio in the middle of the conference allegedly promising symmetric multitasking a major design change by the end of the year of course to implement those features more time will be needed and with each delay I would promise even more features by some miracle a developer release of Copeland went out a year late at the end of 1996 but the software was so buggy that the OS could just crash from sitting idle there were suggestions that Copeland would be out by 1997 though officially the release date was just some time it was apparent to Emilio at that point that the fragment development of Copeland was unlikely to ever coalesce into one revolutionary operating system and so he called in a friend Ellen Hancock who had worked with Gale at his previous job bringing national semiconductor to profitability if anyone was going to get to Copeland dev-team back on track it was her Hancock's first decision upon seeing the state of Copeland was to call it off there was no future for the US currently being developed and Apple should instead consider buying an existing operating system and reworking it into a Mac OS a few offers were brought up like be created by Apple veteran Gianluigi sa which was too expensive Solaris a favorite of Hancock but largely seen as unworkable to Emilio an apples version of a deal with the devil Windows NT avoiding the almost certain backlash of what little Mac users remained Emilio went with his one last option next step since 1985 the company next had made a name for itself by selling so-called 3m computers to higher educational markets 3m standing for one megabyte of memory 1 megapixel display resolution in one mega flop of processing power next never was the most popular brand for the general public but it was tied to some pretty famous successes in computing such as being the first web server as well as having the first web browser being written on it the other big thing behind next was its founder Steve Jobs seeing the next step was exactly the operating system Apple was looking for they bought out next completely in December of 1996 and brought jobs back into the company he had co-founded as an advisor as our Copeland the project was officially canceled in favor of porting next step over to the Macintosh the process of doing so was going to take a while though and in the interim Apple bought up a bunch of third-party extensions for system 7 to incorporate them into the operating system as well as it working on improving stability in system 7.6 given some more time features like customizable themes and a multi-threaded finder originally slated for Copeland would make their way into Mac OS 8 even later Mac OS 8 did see some simplistic support for the pre Multi Testing system that have prompted the development of Copeland to begin with Mac OS 9 released in 1999 brought more improvements to various elements of the Macintosh operating system but more importantly it marked the start of Apple's transition to their new next architecture introducing libraries like carbon that would be used in the new Mac OS subsequent versions were designed with virtualization in mind so that compatible programs could still be run on the new Mac OS architecture and in 2001 it happened Mac OS 10 was ready for its debut and after nearly a decade of planning apple's next generation of us after the failed deals with IBM after the disastrous internal politics of mid-90s Apple and after buying out next in shaking up the leadership the OS had finally shipped and the general response was yeah Apple had made great strides with revamping the US with the Darwin kernel beautifying the user interface through the new aqua system theme adding new UI elements like the dock all while still supporting Mac OS 9 software through virtualization at the same time though OS 10 had poor support for external Hardware out of the box some found the UI changes to be less intuitive the Mac OS 9 and the flashy new aqua theme and effects tended to make the OS somewhat unresponsive on older hardware some were skeptical of the improvements of Mac OS 10 with jef raskin an original Mac team member calling it nothing more than a face light generally many users decided to wait it out until the OS mature Jobs himself had to counter this concern by throwing a mock funeral for Mac OS 9 as a 2002 conference urging developers to move on to only OS 10 development he worked tirelessly on our behalf always hosting our applications never refusing to command always at our beck and call except occasionally when he forgot who he was and needed to be restarted given a few years of evolution and more importantly more powerful generations of Max to support the increasingly demanding operating system OS tank code on receiving nearly annual updates to this day Copeland was far from just a failure and os10 was far from just a rewrite in fact many of the side projects shoehorned into Copeland did eventually become programs on her list app like video chatting which would become FaceTime and desktop search programs like spotlight even the Apple Newton a portable product at the same time as Copeland was being felt would be reworked into the handwriting recognition program Inc WA Jobs had said at the launch of OS 10 that the technology behind the new Mac OS would be able to support it for 20 years to come nearly 20 years later that prediction has become more or less true while the OS has gone through 13 iterations already at the end of the day it's still Mac OS 10 and if you think about it OS 10 has actually been around and supported for longer than the original first nine versions of Mac OS combined you