Live data from Hacker News

MacOS High Sierra

apple.com

561–570 of 630 posts

Re: MacOS High Sierra

#561

Earlier quoted context omitted.

Like the other commenters here, I can confirm that the FileVault -> APFS transition went smoothly.

Anyone know how the APFS conversion goes if you have hardware RAID setup? I’m sporting a Mac Pro mid-2010 and have a RAID-5 array that’ll need to upgrade.

System Information 10.13 reports a medium type "SSD" for my stock MacPro6,1 flash and medium types "SSD" or "Rotational", as appropriate, for SATA SSDs and hard drives connected via non-RAID Thunderbolt AHCI controllers.

For my two Thunderbolt SAS RAID arrays, however, System Information shows no "Medium Type" field at all. And Disk Utility reports "Solid state: No" for both, even though one is, in fact, all SSD.

Therefore, on this one particular Areca controller at least, I'd be willing to bet that SSD boot volumes would look just like rotational hard drives to APFS and the installer, and therefore would not have been auto-converted at installation time.

With that said, on the GM candidate, converting the volumes on these RAID arrays to APFS after the fact was as simple as right-clicking each in Disk Utility and selecting "Convert to APFS...". And it the week or so since I did so, I've had no problems or regrets.

All volumes on this system are FileVault-encrypted, and APFS conversion did nothing to change this.

Re: MacOS High Sierra

#562

"Conclusion In this blog post, we briefly discussed High Sierra's "Secure Kernel Extension Loading" (SKEL) and demonstrated a new 0day vulnerability that can be exploited to fully bypass this new `security' feature. Unfortunately when such `security' features are introduced - even if done so with the noblest of intentions - they often just complicate the lives of 3rd-party developers and users without affecting the b…

It's unfortunate that there is an alleged 0day in this functionality. I don't know why the author is trying to accuse Apple of having the real goal of "wrestl[ing] control of the system away from it users". SKEL looks like a very good idea to me, and the presence of a bug does not make it any less of a good idea. It's also unclear in what way SKEL is supposed to be taking control away from Apple's users anyway. If an…

It's part of an overall trend. For example the gatekeeper security settings ux has slowly been removing options to run unsigned binaries bit by bit.

In 10.12 there is no obvious way on how to run unsigned binaries now. When you double click on an app multiple times, it refuses to run. If you go to security settings you cannot run the unsigned binary from there. You have to right click on the app and press open to get the old dialog that would allow you to run an unsigned binary like it's some sort of easter egg. In older versions you didn't have these restrictions.

Re: MacOS High Sierra

#563
post #543

Earlier quoted context omitted.

They don't even need to do this. Users will literally just click "Yes" to anything obstructing their path forward. http://i.imgur.com/H0uVqFe.jpg

I'm going to ask for a citation on that. Where is proof that average users will ignore reasonably worded security messages? Because as far as I can remember (Windows back to 95, MacOS back to the 6100), we (ie computer science professionals) have never done what I would even call an "acceptable" job at attempting this. And the argument, for 20+ years now, has always been "Har har, users are too dumb to read. We'll ju…

How can a normal user make a judgement whether something is OK or not?

Re: MacOS High Sierra

#564
post #425

Earlier quoted context omitted.

> And yet, microsoft started incorporating ads inside of windows 10. Using daily, only thing I've seen - notification about Edge (which came with the OS). what?

Here are examples of the new, damned Edge ads: https://www.theverge.com/2017/3/17/14956540/microsoft-window...

OMG. I see you completely got flooded with those ads

Re: MacOS High Sierra

#565

Earlier quoted context omitted.

It's unfortunate that there is an alleged 0day in this functionality. I don't know why the author is trying to accuse Apple of having the real goal of "wrestl[ing] control of the system away from it users". SKEL looks like a very good idea to me, and the presence of a bug does not make it any less of a good idea. It's also unclear in what way SKEL is supposed to be taking control away from Apple's users anyway. If an…

It's part of an overall trend. For example the gatekeeper security settings ux has slowly been removing options to run unsigned binaries bit by bit. In 10.12 there is no obvious way on how to run unsigned binaries now. When you double click on an app multiple times, it refuses to run. If you go to security settings you cannot run the unsigned binary from there. You have to right click on the app and press open to get…

Right-click an app and select Open is always how you've been able to bypass Gatekeeper to run unsigned binaries. That hasn't changed.

AFAIK the only thing that's changed since Gatekeeper was introduced was the Security preferences lost the option to allow unsigned apps.

Re: MacOS High Sierra

#566
post #158

Earlier quoted context omitted.

Details? Unless you’re developing malware or low-level system extensions SIP should have no impact on day to day work.

It breaks all sorts of things, even the version of py2app that Apple themselves ship - but are too lazy to test or read bug reports about for year after year.

“All sorts of things” is a broad statement to toss around without at least some links. I primarily work in python and had never even heard of the py2app issue since it’s not an incredibly common tool and most python developers I know use newer versions of python.

That’s why I questioned the original broad claim: I know there are edge cases but most of the developers I know work on Macs and SIP just isn’t mentioned often enough for it to be anywhere near as bad in general as a few random commenters claim, not to mention that anyone I know who’s at all security savvy appreciates that it’s a trade off rather than a unilateral bad move.

Re: MacOS High Sierra

#567

Earlier quoted context omitted.

> I don't think there's any replacement for Excel That would be great. People over use Excel to no end and it causes problems. They need to use programming. Get people with R or Python and Pandas, or some other statistical program. (88% of Excel Spreadsheets contain human errors) These are human error. Use a program not an Excel sheet. https://www.forbes.com/sites/salesforce/2014/09/13/sorry-spr... One error cost $6…

Excel is programming. It's just not the kind you do or like. The number of programming things people have done in Excel and Access is astounding, as is the number of people who've learned to program without realizing it as a result of using those tools. Could they be better? Sure. But don't knock it as not "programming" on that basis -- PHP is also bad.

Even more: Excel is purely functional programming for the vast majority of functions that people use in Excel.

Re: MacOS High Sierra

#568

Earlier quoted context omitted.

It's part of an overall trend. For example the gatekeeper security settings ux has slowly been removing options to run unsigned binaries bit by bit. In 10.12 there is no obvious way on how to run unsigned binaries now. When you double click on an app multiple times, it refuses to run. If you go to security settings you cannot run the unsigned binary from there. You have to right click on the app and press open to get…

Right-click an app and select Open is always how you've been able to bypass Gatekeeper to run unsigned binaries. That hasn't changed. AFAIK the only thing that's changed since Gatekeeper was introduced was the Security preferences lost the option to allow unsigned apps.

I've run several unsigned apps on Sierra and previous versions with Gatekeeper, and not once have I right-clicked an app to do it.

Re: MacOS High Sierra

#569
post #384

Earlier quoted context omitted.

Unfortunately they don't (and probably can't) do it for everyone. What is needed is bug level compatibility with msoffice (including visual basic) and seemless interoperability (including add ons). Not only is this an insanely difficult target - it is also a moving and potentially hostile one to interoperate with. I've massive respect for the libreoffice developers but I don't envy them the task of msoffice interoper…

Office 2003 and Office 2007 upwards don't have bug level compat. it's even worse if you used special things in your .doc files the chances are/were high to render them different between the older and the newer versions. basically nobody cared as soon as a lot of people moved to ooxml. Also Office 2003 and LibreOffice4/5 have way more in common than Office 2003 has with 2010/2013/2016.

Bare in mind I've been often on the other side of this discussion....

None of this really matters - if you tell someone word ate a word document then they are sympathetic whereas if you tell them libreoffice ate a word document the reaction is much less favourable. I do not like this.

My solution - I just refuse to use any office software.

Re: MacOS High Sierra

#570
post #566

Earlier quoted context omitted.

It breaks all sorts of things, even the version of py2app that Apple themselves ship - but are too lazy to test or read bug reports about for year after year.

“All sorts of things” is a broad statement to toss around without at least some links. I primarily work in python and had never even heard of the py2app issue since it’s not an incredibly common tool and most python developers I know use newer versions of python. That’s why I questioned the original broad claim: I know there are edge cases but most of the developers I know work on Macs and SIP just isn’t mentioned of…

I develop against the deployment environment run in virtual machines for two reasons.

1) Too many packages/servers/etc. I've tried to install under OS X over the years just didn't quite work right. That's probably not the case so much any more, but I have experienced it recently.

2) Developing against a macOS localhost can mask problems associated with my code running in the deployment environment. So to avoid those surprises, I develop against the deployment environment.

If I need root for package install or other server deployments, I log into the VM and do it there. I rarely need to install stuff on my macOS workstation.

Post reply on HN