Hacker News Reader

So I'm sitting here, just unlocked my Mac through putting my finger on the touchbar and reading your post, wondering who that "pro" is that I keep on hearing in conjunction with Apple during the last 6 months. I'm a dev, doing the Kool Aid for a 100bn$+ company, together with designers in my team. I always thought devs and designers would be that pro usergroup people are referring to. I've read all the complaints in the last time of "Apple is abandoning the pros" but funny thing, I only heard that on the internet and in hypothetical perspectives, nobody actually complained about the lack of Kaby Lake irl. I'm very fine with the current lineup and I'm wondering if that "pro" you're referring to is actually just an extremely narrow niche.

I guess the "pro" is anyone who wants more than 16GB RAM ?

those "pros" complaining about the 16gb limit, I wonder how they've been getting any work done up until now?

Because they bought their computers 3 years ago when you asked why anybody would ever need more than 8gb.

For comparison, the minimum amount of RAM required by the precursor of Mac OS, NeXT, - 32MB (that's megabytes) - seemed insane. But then again, 90% of the computer science - including experiments in AI - had been done (at least, up until mid-80s) on computers that had less than 1 megabyte of RAM. (IIRC, the first version of UNIX was designed to run on a computer with 24 kilobytes of RAM.)

It really depends on what you are doing; I spend a lot of my time doing dev on a 512mb system and my main system where I do 80% of my work on has 4gb (with ssd though!). For the work I do: JS dev, Android dev, most image recog, C/C++ dev, Python dev and just playing around with esoteric langs that works fine. I do not use heavy window managers (I prefer Ratpoison or i3; both are incredibly light) and when I need graphical software that is not being devved by me at the moment itself, I start it, bundle my tasks and close it again. It works well for both battery life (15-17 hours on 2011 tech) and for me it is perfect. I was an avid Mac user before that but I like Linux more for some reason; matter of taste.

It seems however though, and this is an assumption I can only base on personal experience of what I see, that people like all the 'stuff' during their work that eats battery and memory and cycles.

All the animations and things that go on in the popular wms and OSs and browsers that deliver no 'direct value' to me (like said; this is I personal thing I believe) seem a must for most people even though they are devving.

Why would one use the incredibly heavy and animated WMs over something fast and simple which even works faster (for me!). When I use Unity (latest) my laptop has 5 hrs battery; switch to i3 and plop 10+ hours more.

Do people like that so much or am I just completely wrong?

(one other reason I see with colleagues is very heavy IDEs but let us not go onto that)

I think you're confusing cause and effect. Heavy WMs are only possible because there is so much RAM to spare on the average machine. The RAM is originally needed for applications.

For example, I'm working on a Macbook, but inside a Linux VM. It was running at 3-4 GB RAM for a long time, but I had to bump it up to 6 GB when I was working on an interconnected set of Rails applications. (Ruby and Rails are notoriously memory-inefficient.)

Another example: In 2010, I had to upgrade my notebook from 1 GB to 2 GB of RAM, not because of KDE 4 (which was considered absurdly resource-intensive at the time), but because of Git. (Git is designed to use the filesystem cache extensively, and will become absurdly slow when the cache is too small.)

>I think you're confusing cause and effect

I was responding to many people who 'cannot work' because they have less than 16gb, the latest discrete gpu and cpu. Which is where all the mb pro whining now revolves around. You are playing to 'if you have more you spend more?' human condition right? If you have more money, memory, speed whatever, you will find some way to use it fully; that is true. But you do not /need it/. I was the same a few years ago until I dropped Mac OS in favor of Debian and bought a stack of old (2011) laptops. After tweaking everything but the stuff I do not need on it, it works as well /for te dev I need it for/ as the last (2015) MB pro I touched for a fraction of the price and i3 does make me more productive.

Anyway; I am not trying to convince or sway anyone, just saying that, imho, it is nonsense that most (by very far margins) really need that kind of metal. But sure, if you can get them, then get them. In my slightly older age I try to get back to simpler and more optimal. The cyclewaste/bitwaste is getting annoying to me (IF it is waste; I do deep learning and play/fiddle with games on heavy metal).

"640K ought to be enough for anybody."

The lack of capacities > 16GB has been covered to death ... if Apple put non low-power RAM in their machines they would be crucified for the drop in battery life. Until Intel actually makes a release date (increasingly unlikely) there's not a lot they can do before the next refresh.

Apple could have kept the extra millimeters in thickness for a larger battery. High-power >16GB RAM plus long battery life would be "pro". Nobody would be crucifying if MBP was same thickness as before.

It would be nice yeah but I doubt the niche is big enough to justify the R&D + logistical burden when those limitations are resolved by Intel in next year's chipset.

It sucks but IMO the memory line of argument is probably the weakest in this whole discussion.

Apple created its own dilemma of having to choose between RAM or battery life when it decided to reduce the battery capacity.

Even if they had a bigger battery, doing so would have required them to do develop and maintain another board/chipset which is nontrivial.

More than 16 GB RAM or CUDA or a multiplatform graphics API that wasn't 6 years out of date...

This is a misguided rebuttal. Fixated against the a sub-set of users that are clearly and unanimously voice a complaint, when the problems here is apple.

PROs can work with anything, not matter how bad, and get results done.

However, that not mean we are so STUPID to not know when a tool is made for us or not.

Is about a specific spec? NO!.

Is about the trend that Apple have in the whole ecosystem.

Just now, after years waiting for a refresh, what is given is a stronger confirmation that the trend not only is unstoppable but the trade-off is not good enough.

> PROs can work with anything, not matter how bad, and get results done.

I picture this being written on a Commodore 64.

My 2013 MBP is great, I can see it lasting a few more years even. It was that forward looking.

A 2016 MBP... I can't see it holding up. The cap on RAM I'm already butting up against it...

Photoshop, VMs, site crawls, loading large spreadsheets and trying to crunch the numbers... 16 GB taxes my current system.

I've sort of just given up, I have a gaming desktop I built myself a year or so back with 64 GB of RAM. I just use that when I want to do "work" now... everything loads faster. Freakin' hate that I can't just buy a new MBP for that stuff, but... it's not top of the line, isn't built to last like previous generations were.

I'll keep my 2013 MBP until it gives out on me I guess... running out of Apple Care on it and the battery cycles are high... also not wild about trusting work to a 3-year-old SSD... but cross my fingers I guess.

My sister still uses my old MBP from 2008 or so -- it holds up for most of her needs after a replacement battery and SSD, it's still a solid computer. I don't see these new MBPs holding up like that at all.

Not going to say the new MBPs are shit, but they aren't built to be light-years ahead like they once were. That's what I think everyone is upset about.

I think what's most upsetting to me is that not only are they not built to be light-years ahead, but the fact that they aren't is explicitly due to self-imposed constraints by Apple. And then because of this, they now sell a machine that's effectively the same as the one they sold 3 years ago, but now at a higher price.

This is virtually unprecedented in the history of computing, and the last time these kinds of shenanigans happened was at the tail end of the 8-bit era, and all of those companies, even ones who had a virtual monopoly on the market, don't exist anymore.

Think it's an intentional move to have a shorter product lifespan to sell more computers? I go to thinking about that more... seems like they benefit form making less-great computers than they used to.

I'm clinging for dear life to my beloved five year old MBP. I don't know if my next laptop will even be Apple at all.

Same here. New battery, new secondary SSD, reconditioning of the MagSafe power supply and maxed out RAM are underway. Performance-wise it should be about the same as the latest Macbook, which doesn't feel like an upgrade for software development.

Slow hardware helps one write good software after-all..

Gave my Macbook Pro 13 (2013) to the missus a few months ago as I was looking forward to the new one.

As it is, i've just spent £1400 on a XPS 13 fully loaded and instead of a iMac i'll be buying a Surface Tablet. The missus after having each iPhone since the first is going to be buying a Pixel.

Apple have lost the plot and are taking the piss out of their users. We are voting with our wallets.

Yeah Android really has iOS beat right now on UI/UX. It's pretty embarrassing.

A friend of mine has a Surface Pro, and I have a not-insignificant amount of envy. It's the first time in almost 20 years that I've looked at a Windows laptop and felt it was in any way ahead of Apple. Great design, and it doesn't have the inevitable pointlessly-cheap compromises that plague every other "high end" Windows laptop.

Of course, it's still Windows underneath. Sooner or later, I'd get tripped up by the pathetic reality of 8.3 case-insensitive filenames buried in a not-deep-enough grave.

I know multiple people who are buying 2014 and 2015 model years or even PCs instead of the new one.

Isn't this (form-over-function) a similar but not exactly identical phase that Microsoft went through, that resulted in the now publicized "Raymond Chen camp vs. MSDN camp" mantra?

https://blogs.msdn.microsoft.com/dareobasanjo/2004/08/25/the...

Is this ever avoidable once you gain a large swath of market share, but are still under pressure to grow even more?

>We lose good functionality for no good reason, other than products become "simpler" and "more pure".

You can disagree with the reasons, and certainly many do, but it's just silliness to allege that moving to USB-C was done for "no good reason".

It's the exact same silliness that surfaced in the griping about removing 9-pin serial ports, and optical drives, and Ethernet ports in laptops, and FireWire in favor of something more than an order of magnitude faster, and every other innovation by Apple.

It's always for a good reason.

Well, not always. Removing the iPhone headphone jack was dumb and pointless.

Losing the ethernet jack on Pro laptops is also pretty bad. The reason is thinness, I dunno if that's good or not. WiFi is sufficient for most people, but it sure as hell isn't better when you need high bandwidth and consistent low latency.

> Removing the iPhone headphone jack was dumb and pointless.

I suspect there was a point. This is part of Apple's standard MO.

It's a way to wean people onto wireless (BT). Apple's expensive BT headphones point the way to cheap third party BT headphones that aren't painful to connect (right now bluetooth infrastructure is pretty sucky).

Once people are comfortable without the headphone wires they will have inductive charging like the watch and have a water resistant device like the new watch.

Why would you want to move to USB-C other than not having to flip the cable when you inevitably put it in the wrong way.

If your answer is power delivery, remember that not all cables and not all devices support power delivery on USB-C so it's nothing that couldn't have been done with "USB4" utilising the same plug we've always had.

Few reasons for me:

* HDMI and DP alt modes make laptop docks a lot cheaper and more universal

* Don't need an OTG cable to use a Type-C device on my Type-C phone

* Potential Thunderbolt 3 port reuse

Actually, I never mentioned USB-C.

But let's take it as an example: adding USB-C was a great idea.

Removing every other kind of port, including the fantastic Magsafe, on the other hand, was a bad idea. Many people try to justify it by saying that it had to be done so that we can "make progress". Well I'm fine with progress, but I don't want to be held hostage. And it certainly feels like it: if I buy the new Macbook, I will be able to connect exactly ZERO of my devices to it.

But — what I actually meant was things like magsafe, charging light on magsafe, function keys, Esc key, little wire organizers on the charger, keyboard key travel, or the headphone jack on the iPhone. You could easily make the same case for battery life (I'd happily trade 2mm of thickness for longer battery life) and ports: I really don't care that much about thickness, so don't sacrifice ports to make my machine 2mm thinner. Again: we lose good functionality for no good reason.

Sorry, I don't find the MagSafe fantastic at all, at least not in the MagSafe 2 incarnation. The thing detaches far too easily for my liking, leaving me ever more frustrated. It's the only thing I dislike about my 15" rMBP

Agree. The first MagSafe from 2009 worked, but I went through multiple chargers. The MagSafe on my new MBP (previous gen) doesn't work nearly as well, and just disconnects every time.

> This in itself would not be a problem for me, if it weren't for the lack of competition. Mac OS is so far ahead in terms of "getting things done"

I wouldn't touch MacOS for it. It's nowhere comparable in productivity for me to something like current KDE Plasma 5 on Linux. Not sure if you ever used it, but Linux is nowhere limited to Ubuntu's Unity.

I didn't want to get into these kinds of discussions, but I'll venture a guess that you don't often connect/disconnect multiple monitors to your machine, and especially not one HiDPI and one HD. That experience alone is enough to ditch Linux altogether.

BTW, Windows 10 is slightly better, but not by much. The major difference being that one can predictably get a signal on both monitors, I guess, which is good, but then the HiDPI story is still pretty bad.

I do. Plasma nailed down multi-monitor setup in 5.8.x. Before that it was significantly worse.

For different DPI on each monitor you'd need to use Wayland though.

As a keyboard user, I feel Windows more polished than macOS when you work with multiple screens.

To name a simple example, pressing CMD + Shift + -> moves an application to the next monitor, with macOS you should pay for a window manager to get a similar functionality, or give up the keyboard and embrace a touchpad-centric gimmick like Expose, which I perceive as distracting with these animations.

It is certainly an example of YMMV.

Spectacle is free and does this and a lot more should you need it.

Ah, multi-monitor support. I always joke how I ditched hardware random-number generators because the window placement of macOS is a superior source of randomness.

Meanwhile, multi-monitor support in Plasma 5 has worked for me for years without any trouble whatsoever.

Well, it was buggy until not very long ago:

* https://bugs.kde.org/show_bug.cgi?id=365455

* https://blog.martin-graesslin.com/blog/2016/07/multi-screen-...

This also was rather annoying and took a long time for distros to catch up: https://bugreports.qt.io/browse/QTBUG-42985

Since Plasma 5.8.x, it's much much better.

Can it work with a high dpi laptop monitor? I want to put Linux on an old MacBook Pro Retina but it appears that large parts of the interface are still hard coded for certain resolutions. Also read that battery life will take a big hit.

Yes, it can. Plasma 4 wasn't really suitable for it, but Plasma 5 is OK. Can't say much about battery life - it can differ a lot depending on the laptop and how well various power features of the chipset are documented. A lot of time Intel for example are pretty sloppy and don't document things like that, so there isn't much Linux developers can do.

mjg59 wrote a number of comments on this subject.