5 ms·

I noticed the trend with Snow Leopard. Also the tendency towards iOS-ification of Mac OS, so I decided to switch back to Linux. I think Apple's best hour came w

by nextos 11y ago

I noticed the trend with Snow Leopard. Also the tendency towards iOS-ification of Mac OS, so I decided to switch back to Linux. I think Apple's best hour came with Tiger and Leopard. Simple applications that did one thing well and were robust. Now there's a lot of feature bloat. A nice Linux set up with minimal software (e.g. xmonad, mutt, emacs or vim) is a joy to use, but takes time to set up and learn to use. So it's not for everyone. Interestingly, I've experienced the same issue Apple is suffering with Ubuntu. Edgy Eft (6.10) was incredibly simple and nice. Like OS X Tiger. Now there are dozens of services running and something always gives trouble. I guess the old adage applies, make things as simple as possible but no simpler...

wwweston 11y ago
IIRC, Snow Leopard was actually the best release -- essentially Leopard but with a focus on stability and performance, plus the code signing stuff. It's the last OS X release that felt like an improvement to me, and one of the last OS X releases that I trusted. Every one since has actually had feature regressions (let's make a formerly visible folder invisible!) or stability issues. I still have a 2008 Macbook Pro running Snow Leopard and it's more stable than a 2012 MBP running Mavericks and performs about as well.
- fuzzywalrus 11y ago
  Mavericks was tough to give up, stable, fast. No discoveryd. IMHO, its the best release of OS X. Maybe everyone has amnesia but pre 10.3, OS X wasn't really usable full time. Each 10.2.x release was a whirlwind of changes. Early iOS had wonky reset issues too.
  - VonGuard 11y ago
    Very true statements. However, those were growing pains. I feel like Apple's problem now is that it has completely given up on its overarching philosophy: to make software simple. Instead, they're just offering us their own versions of things just because they can, not because their's is any better. Maps being exhibit A, and Apple's mail client not changing in 10 years (aside from some minor adds and removes) are examples of this. Apple used to mean software for people who were not technology savvy. Now it means phone software for people who are not technologically savvy, and computer software that's just like everyone else's. Frankly, I can never figure things out on an iPhone. They're too confusing in their interface. Especially compared to my beloved Newton 2000.
    - pyre 11y ago
      > Apple's mail client not changing in 10 years (aside from some minor adds and removes) are examples of this How is Mail.app supposed to change though? It still completes the task that it was designed for. I'll agree that Mail's stability has been spotty across releases, but how are you expecting them to change such a core app?
    - pivo 11y ago
      Regarding Mail, I don't really want it to change. Other than the improvements in stability and speed that I've noticed over the years, Mail.app is pretty much exactly what I want in a mail reader on OSX. Sometimes Apple's changes seem purely for the sake of change and result in reduced usability (e.g. Photos), I don't want that to happen with mail.
  - shampine 11y ago
    I'm so glad both my work and personal machines are still on 10.9. I've been contemplating updating, but everything I've ready continues to say I shouldn't.
  - pyre 11y ago
    Don't worry. I remember. cifs.kext is hella more stable nowadays than back then. I used to have the entire OS come to a grinding halt if a mounted Samba (or NFS) share disappeared (e.g. mount local share at home, put laptop to sleep, wake laptop at school). I remember that in (I think 10.3) they "fixed" the issue by having a timeout dialog popup and ask if you wanted to disconnect... only it was too sensitive and would popup (and then go away before you could react) if there was any jitter in the network connection.
- mergy 11y ago
  Absolutely. Snow Leopard was the pinnacle. I could run for months without a reboot and did a tremendous amount of heavy work on that old Macbook Pro. Starting with 10.7 and seeing the decline coming, I was happy that Ubuntu ran very well on that laptop and that held me over until I left Apple hardware entirely.
tachyonbeam 11y ago
Ubuntu has been doing the same thing as Apple with their new UI, trying to jump in on the tablet bandwagon, with this belief that tablets are the future. This is fairly stupid, considering that Ubuntu's marketshare was pretty much 100% PC, 0% tablets. They gave their main demographic a slap in the face, and they never really did acquire any tablet market. Now, according to DistroWatch, Ubuntu is below Mint and Debian in terms of popularity. There's something to be said about reinventing the wheel just for the sake of reinventing it, change for the sake of change. Constantly redesigning UIs that were working perfectly well, chasing after fads. In implementing a new UI, Ubuntu acquired a lot of new bugs, broken features and reliability issues. That's normal. Newer code is buggier code. You might think this "old" code is crufty, it might not be designed in the ideal way you like, but by throwing it away, you throw away years of testing and fixes too. That's an argument in favor of incremental evolution and refactorings IMO. Maybe I'm biased. When I was 16, I used to find colorful desktop backgrounds and fancy UIs cool. Now I'm 30, and I just want the damn thing to f'ing work reliably. I don't need rounded corners, or transparency, or animations or even a desktop background. I'd be OK with a bland UI that looks like Windows 98, so long as the machine can do all I need it to reliably and fast.
- agumonkey 11y ago
  Immaturity leading to think that following market trends instead of believing in (and even 'still understanding') you're own quality, will be the only way to sustain your business. Jitter.
  - slgeorge 11y ago
    Every business (and I'd argue project) needs to grow and change with how the world is developing. The Palm Pilot software was perfectly fine in 2004(?) but wouldn't stand a chance serving users-needs in 2016. The Linux distro's actually serve multiple 'customer' segments. On the client side there are at least two customers (at a minimum), one is end-users, the other is the OEM's who pay for Linux to be preloaded. The traditional OEM's need a solution to the PC market shrinking while the devices market has eaten their lunch. It's true you have to determine the difference between a short-term 'trend' and a long-term shift. I'm sure you don't think that the mobile market is a short-term 'trend'.
ramenmeal 11y ago
Interesting. Snow Leopard was OS X's peak for me. It was Leopard with perf and stability improvements. Lion was when the iOS influence came to the mac.
mbrock 11y ago
I think we're heading for a synthesis of these opposing trends, the common core being a fatigue with complex bloated user interfaces, and a return to simpler design. Personally I enjoy command line interfaces, especially in full screen with huge fonts, and I've been using Ratpoison (one of the first tiling X window managers) like this since like 2004. But I also enjoy, at least in principle, tablet-style UIs. Neither is based on overlapping windows, for one thing, and they are both helpful for my quasi-ADHD/OCD tendencies. Mac OS now is in a weird limbo between the Unix heritage, the Mac heritage, and the iOS heritage. I think unifying these strands is an enormous challenge, but also really cool and inspiring. Maybe Apple are going to focus on making cars or whatever, and some weird little upstart is going to come along and make something totally new. You can see people starting to talk about Slack as an operating system, but that's still very primitive. Chat and command lines are interesting because they represent a huge paradigm that's been overshadowed by the "GUI" paradigm, namely the paradigm of queries and responses: you ask the computer something and get something back, perhaps asynchronously. It's such a useful, coherent model. Easy to program with. Portable across interfaces. Good for exploration. Cognitively appropriate. Etc etc. I think the complexity and bug-riddenness of most "modern" desktop operating systems come from an over-complicated and incoherent model of operation. Mobile represents a new start. It currently lacks in flexibility, but it's probably a good thing to start with rigid simplicity than to try for flexible complexity, and unless you can perfect a genius UI right away I do think those are the options, if you're aiming for the general public who don't want to learn vim. I also think the divide between application programmers and computer users is politically a Bad Thing and also a result of exaggerated complexity. Unix was always about user scripting, and it managed that because of textual data, worse-is-better, and RTFM. Apple might not be the company to get back to this, because they're making tons of money from making shiny coherent appliances for people who pay to think as little as possible. So just generally I'm not holding my breath for Apple to come up with a wildly new and interesting paradigm here. They're a dinosaur. A very pretty dinosaur. I'm more interested in startups with weird ideas, like maybe let's use e-ink displays, cheap ultralight computers, cloud servers, natural language processing, and the interactive fiction adventure exploration paradigm to provide a new style of terminal gadget that's even cooler than Linux. (If you steal this idea, please don't fuck it up.)

by nextos 11y ago

wwweston 11y ago

IIRC, Snow Leopard was actually the best release -- essentially Leopard but with a focus on stability and performance, plus the code signing stuff. It's the last OS X release that felt like an improvement to me, and one of the last OS X releases that I trusted. Every one since has actually had feature regressions (let's make a formerly visible folder invisible!) or stability issues. I still have a 2008 Macbook Pro running Snow Leopard and it's more stable than a 2012 MBP running Mavericks and performs about as well.

fuzzywalrus 11y ago
Mavericks was tough to give up, stable, fast. No discoveryd. IMHO, its the best release of OS X. Maybe everyone has amnesia but pre 10.3, OS X wasn't really usable full time. Each 10.2.x release was a whirlwind of changes. Early iOS had wonky reset issues too.
- VonGuard 11y ago
  Very true statements. However, those were growing pains. I feel like Apple's problem now is that it has completely given up on its overarching philosophy: to make software simple. Instead, they're just offering us their own versions of things just because they can, not because their's is any better. Maps being exhibit A, and Apple's mail client not changing in 10 years (aside from some minor adds and removes) are examples of this. Apple used to mean software for people who were not technology savvy. Now it means phone software for people who are not technologically savvy, and computer software that's just like everyone else's. Frankly, I can never figure things out on an iPhone. They're too confusing in their interface. Especially compared to my beloved Newton 2000.
  - pyre 11y ago
    > Apple's mail client not changing in 10 years (aside from some minor adds and removes) are examples of this How is Mail.app supposed to change though? It still completes the task that it was designed for. I'll agree that Mail's stability has been spotty across releases, but how are you expecting them to change such a core app?
  - pivo 11y ago
    Regarding Mail, I don't really want it to change. Other than the improvements in stability and speed that I've noticed over the years, Mail.app is pretty much exactly what I want in a mail reader on OSX. Sometimes Apple's changes seem purely for the sake of change and result in reduced usability (e.g. Photos), I don't want that to happen with mail.
- shampine 11y ago
  I'm so glad both my work and personal machines are still on 10.9. I've been contemplating updating, but everything I've ready continues to say I shouldn't.
- pyre 11y ago
  Don't worry. I remember. cifs.kext is hella more stable nowadays than back then. I used to have the entire OS come to a grinding halt if a mounted Samba (or NFS) share disappeared (e.g. mount local share at home, put laptop to sleep, wake laptop at school). I remember that in (I think 10.3) they "fixed" the issue by having a timeout dialog popup and ask if you wanted to disconnect... only it was too sensitive and would popup (and then go away before you could react) if there was any jitter in the network connection.
mergy 11y ago
Absolutely. Snow Leopard was the pinnacle. I could run for months without a reboot and did a tremendous amount of heavy work on that old Macbook Pro. Starting with 10.7 and seeing the decline coming, I was happy that Ubuntu ran very well on that laptop and that held me over until I left Apple hardware entirely.

tachyonbeam 11y ago

Ubuntu has been doing the same thing as Apple with their new UI, trying to jump in on the tablet bandwagon, with this belief that tablets are the future. This is fairly stupid, considering that Ubuntu's marketshare was pretty much 100% PC, 0% tablets. They gave their main demographic a slap in the face, and they never really did acquire any tablet market. Now, according to DistroWatch, Ubuntu is below Mint and Debian in terms of popularity. There's something to be said about reinventing the wheel just for the sake of reinventing it, change for the sake of change. Constantly redesigning UIs that were working perfectly well, chasing after fads. In implementing a new UI, Ubuntu acquired a lot of new bugs, broken features and reliability issues. That's normal. Newer code is buggier code. You might think this "old" code is crufty, it might not be designed in the ideal way you like, but by throwing it away, you throw away years of testing and fixes too. That's an argument in favor of incremental evolution and refactorings IMO. Maybe I'm biased. When I was 16, I used to find colorful desktop backgrounds and fancy UIs cool. Now I'm 30, and I just want the damn thing to f'ing work reliably. I don't need rounded corners, or transparency, or animations or even a desktop background. I'd be OK with a bland UI that looks like Windows 98, so long as the machine can do all I need it to reliably and fast.

agumonkey 11y ago
Immaturity leading to think that following market trends instead of believing in (and even 'still understanding') you're own quality, will be the only way to sustain your business. Jitter.
- slgeorge 11y ago
  Every business (and I'd argue project) needs to grow and change with how the world is developing. The Palm Pilot software was perfectly fine in 2004(?) but wouldn't stand a chance serving users-needs in 2016. The Linux distro's actually serve multiple 'customer' segments. On the client side there are at least two customers (at a minimum), one is end-users, the other is the OEM's who pay for Linux to be preloaded. The traditional OEM's need a solution to the PC market shrinking while the devices market has eaten their lunch. It's true you have to determine the difference between a short-term 'trend' and a long-term shift. I'm sure you don't think that the mobile market is a short-term 'trend'.

ramenmeal 11y ago

Interesting. Snow Leopard was OS X's peak for me. It was Leopard with perf and stability improvements. Lion was when the iOS influence came to the mac.

mbrock 11y ago

I think we're heading for a synthesis of these opposing trends, the common core being a fatigue with complex bloated user interfaces, and a return to simpler design. Personally I enjoy command line interfaces, especially in full screen with huge fonts, and I've been using Ratpoison (one of the first tiling X window managers) like this since like 2004. But I also enjoy, at least in principle, tablet-style UIs. Neither is based on overlapping windows, for one thing, and they are both helpful for my quasi-ADHD/OCD tendencies. Mac OS now is in a weird limbo between the Unix heritage, the Mac heritage, and the iOS heritage. I think unifying these strands is an enormous challenge, but also really cool and inspiring. Maybe Apple are going to focus on making cars or whatever, and some weird little upstart is going to come along and make something totally new. You can see people starting to talk about Slack as an operating system, but that's still very primitive. Chat and command lines are interesting because they represent a huge paradigm that's been overshadowed by the "GUI" paradigm, namely the paradigm of queries and responses: you ask the computer something and get something back, perhaps asynchronously. It's such a useful, coherent model. Easy to program with. Portable across interfaces. Good for exploration. Cognitively appropriate. Etc etc. I think the complexity and bug-riddenness of most "modern" desktop operating systems come from an over-complicated and incoherent model of operation. Mobile represents a new start. It currently lacks in flexibility, but it's probably a good thing to start with rigid simplicity than to try for flexible complexity, and unless you can perfect a genius UI right away I do think those are the options, if you're aiming for the general public who don't want to learn vim. I also think the divide between application programmers and computer users is politically a Bad Thing and also a result of exaggerated complexity. Unix was always about user scripting, and it managed that because of textual data, worse-is-better, and RTFM. Apple might not be the company to get back to this, because they're making tons of money from making shiny coherent appliances for people who pay to think as little as possible. So just generally I'm not holding my breath for Apple to come up with a wildly new and interesting paradigm here. They're a dinosaur. A very pretty dinosaur. I'm more interested in startups with weird ideas, like maybe let's use e-ink displays, cheap ultralight computers, cloud servers, natural language processing, and the interactive fiction adventure exploration paradigm to provide a new style of terminal gadget that's even cooler than Linux. (If you steal this idea, please don't fuck it up.)