4 ms·

I started a lengthy reply before realizing you're just going to continue to pick sentence by sentence, ignore the spirit of the discussion, and then even bizarr

I started a lengthy reply before realizing you're just going to continue to pick sentence by sentence, ignore the spirit of the discussion, and then even bizarrely ignore the same points you previously attacked. I'm confused as to what you're trying to win though. The benchmarks you provide are useless because they're synthentic. No one runs one task at a time (well, unless you use the new fullscreen mode :-P). None of the benchmarks have anything to with the task of programming, which is what the original topic was about. I'd imagine if you stay under the upper bound on load, the performance is roughly on par. But, I can make an 8 GB MBP start swapping like crazy in about 5 minutes. I even added an SSD just to deal with slow swapping. Performing the exact same tasks, I didn't run into that with Snow Leopard. I probably need 16 GB RAM with Lion, but dropping another $1,500 to make that happen was a bit much. We clearly have different work styles. It's really okay if we have differing world views. I prefer to work with two maximized apps, one on each display. It works best for me. As more apps adopt the new form of full screen, which auto-kills the other display, my life gets harder though. It's really not fullscreen apps, it's single-focus apps. Great. It still doesn't work for me. If you haven't seen anyone complain about Lion, I really can't help you there. I honestly haven't seen anyone use Lion and not complain about it (well, I guess with you now being the notable exception). A good friend of mine just tweeted a few days ago that it took 15 min. to reboot Lion. I doubt he benchmarked it against Snow Leopard, but I doubt he was lyin' either (pun fully intended). At the end of the day, Apple is all about the customer experience, or so I've been told. The hardware doesn't work for me. Prolonged typing on the chicklet keyboards introduced with the unibody MBPs has destroyed my hands. The OS doesn't work for me. You don't have to agree with me. Maybe you're right and it's not the direction that's a problem but the current manifestation. It really doesn't matter because I need to get work done today and Lion isn't the best tool for the job. I suspect Mountain Lion is going to exacerbate the problem. Maybe it won't, but you don't know either. The nice part for me is I no longer have to care. I'm much happier with my switch.

batista 14y ago
>I started a lengthy reply before realizing you're just going to continue to pick sentence by sentence, ignore the spirit of the discussion Huh? Was there some overall spirit I missed, besides a general frustration with Apple's direction overall and the latest OS X? To that frustration one cannot reply anything, because it's vague and subjective. But one can reply to its constituents, or "pick sentence by sentence". I take a little offense to the accusation, because technical arguments are responded to "sentence by sentence" (i.e measurable issue by measurable issue), not with generic wand having. What could have I said about the overall theme of "Apple's direction"? One doesn't even know what it means until it is broken down into specific actions and problems. >I'm confused as to what you're trying to win though. How about nothing? I read some statement that Lion is resource hungry and slow, and I responded to the contrary. It's a technical question, not a competition. I also tried to make heads or tails of what this dissatisfaction with Apple's direction could entail --it seemed to revolve around this Lion is slow thing too. >* The benchmarks you provide are useless because they're synthentic. No one runs one task at a time (well, unless you use the new fullscreen mode :-P). None of the benchmarks have anything to with the task of programming, which is what the original topic was about.* I wouldn't call those benchmarks useless. How one app runs should be indicative of how many of them run, on successive versions of the OS. Except if Apple changed the Lion's scheduler so that despite one task at a time running the same or faster, when running many tasks the performance get worse. Well, with Darwin's code being available, I guess we can check this also. In any case, anecdotal evidence with no numbers is worse than a synthetic benchmark. As for the "task of programing", when you program you normally run "one task at a time". Mainly XCode, or Vim/TextMate. And then, after you wrote some stuff, a compiler or interpreter. Other apps running, like a terminal, are of negligible impact. Isn't programming, 99% of the time, one of the LESS resource hungry operations? I'm a programmer too --the heaviest thing I use is a VM, in which I test my code. Seems to go along just fine with 4GB (1GB allocated to the Centos VM). >I'd imagine if you stay under the upper bound on load, the performance is roughly on par. But, I can make an 8 GB MBP start swapping like crazy in about 5 minutes. Alright, can you share the process? That would be easy to reproduce, or not, on SL.
- nirvdrum 14y ago
  We'll just have to agree to disagree. I'm still trying to decide if I'm just being trolled. But no, when I program, I have a lot more going on. I have Postgres, Cassandra, and redis doing some work, another JVM instance running RubyMine or IDEA, possibly another JVM instance for the test runner, and maybe a VM. Nor do I shut down every other app on the system, so my email, chat clients, and browser windows are all open in the background. Usually a browser is open and available on the second screen so I can verify the results of my work (Web product) or reference API documentation. That workload in no way is represented by a Photoshop rendering test where that's the only task being run. In any event, I no longer own a Mac. It would be an assault on the scientific process for me to provide steps for an experiment from memory and possibly use imprecise language or miss a step. Moreover, I fail to see what it would even gain in this case, since neither of us could do anything about the results. FWIW, I did switch to Linux natively on my MBP before outright replacing it and found I hardly ever swapped, so I exhausted my options to rule out bad hardware. I'm quite certain it was the OS, but even if it wasn't and I'm just a fool, at least I'm a happy fool with my switch and that's good enough for me.

batista 14y ago

>I started a lengthy reply before realizing you're just going to continue to pick sentence by sentence, ignore the spirit of the discussion Huh? Was there some overall spirit I missed, besides a general frustration with Apple's direction overall and the latest OS X? To that frustration one cannot reply anything, because it's vague and subjective. But one can reply to its constituents, or "pick sentence by sentence". I take a little offense to the accusation, because technical arguments are responded to "sentence by sentence" (i.e measurable issue by measurable issue), not with generic wand having. What could have I said about the overall theme of "Apple's direction"? One doesn't even know what it means until it is broken down into specific actions and problems. >I'm confused as to what you're trying to win though. How about nothing? I read some statement that Lion is resource hungry and slow, and I responded to the contrary. It's a technical question, not a competition. I also tried to make heads or tails of what this dissatisfaction with Apple's direction could entail --it seemed to revolve around this Lion is slow thing too. >* The benchmarks you provide are useless because they're synthentic. No one runs one task at a time (well, unless you use the new fullscreen mode :-P). None of the benchmarks have anything to with the task of programming, which is what the original topic was about.* I wouldn't call those benchmarks useless. How one app runs should be indicative of how many of them run, on successive versions of the OS. Except if Apple changed the Lion's scheduler so that despite one task at a time running the same or faster, when running many tasks the performance get worse. Well, with Darwin's code being available, I guess we can check this also. In any case, anecdotal evidence with no numbers is worse than a synthetic benchmark. As for the "task of programing", when you program you normally run "one task at a time". Mainly XCode, or Vim/TextMate. And then, after you wrote some stuff, a compiler or interpreter. Other apps running, like a terminal, are of negligible impact. Isn't programming, 99% of the time, one of the LESS resource hungry operations? I'm a programmer too --the heaviest thing I use is a VM, in which I test my code. Seems to go along just fine with 4GB (1GB allocated to the Centos VM). >I'd imagine if you stay under the upper bound on load, the performance is roughly on par. But, I can make an 8 GB MBP start swapping like crazy in about 5 minutes. Alright, can you share the process? That would be easy to reproduce, or not, on SL.

nirvdrum 14y ago
We'll just have to agree to disagree. I'm still trying to decide if I'm just being trolled. But no, when I program, I have a lot more going on. I have Postgres, Cassandra, and redis doing some work, another JVM instance running RubyMine or IDEA, possibly another JVM instance for the test runner, and maybe a VM. Nor do I shut down every other app on the system, so my email, chat clients, and browser windows are all open in the background. Usually a browser is open and available on the second screen so I can verify the results of my work (Web product) or reference API documentation. That workload in no way is represented by a Photoshop rendering test where that's the only task being run. In any event, I no longer own a Mac. It would be an assault on the scientific process for me to provide steps for an experiment from memory and possibly use imprecise language or miss a step. Moreover, I fail to see what it would even gain in this case, since neither of us could do anything about the results. FWIW, I did switch to Linux natively on my MBP before outright replacing it and found I hardly ever swapped, so I exhausted my options to rule out bad hardware. I'm quite certain it was the OS, but even if it wasn't and I'm just a fool, at least I'm a happy fool with my switch and that's good enough for me.