A little bit of off topic fast fastness for buildy goodness.

[Alex Zavatone]

So, after the glow of joy that I am feeling from the Xcode team mostly fixing 
my, "stop animating everything" bug and the fact that I've been pretty much 
working 7 days a week for the past n+1 months, I decided to look into some 
funness related to Xcode before I need to wake up at 3 AM for a lovely Saturday 
stretch.

For giggles, I wanted to check out how Ludicrous Throughout affected my app's 
build time on my latest 1 year old laptop.  

First of all, I created a RAM drive by conspiring great evil and entering this 
terminal command

diskutil erasevolume HFS+ \"RAM Disk\" `hdiutil attach -nomount 
ram://4194304`

Don't enter this at all until you have checked the diskutil man file and have 
proven that I'm not trying to erase your HD.

Now that you're proven that, this creates a roughly 2 GB RAM drive.

Running the AJA sped test app, this gave a R/W throughput of 3457.3 MB/s / 
3503.4 MB/s throughput.

Meh.  It's not AppleTalk throughput, but it will have to do.

Already underwhelmed, I copied my iOS project that builds it's own app, then 
links to source and builds about 3 linked libs, onto the RAM Drive.

I had to wait for a whole blink of my eyelids to complete before this 
completed.  Pained sigh.

Next, I set the build location in prefs to be set to legacy.  This makes the 
build folder(s) within the project's  folders, but doesn't touch the derived 
date location.

Command Shift K.

Command B.

A build that normally takes 2 mins and 35 seconds of our iOS project took 31 
seconds.

Sigh.  Pained sigh.  ONLY 5 times faster?  

This will never do.

By moving the system and Xcode and the user folder onto a large enough RAM 
drive and gotten stupid faster builds where it takes more time to launch the 
simulator than it does to compile from clean.


Why this matters to me - as an experiment, if nothing more, is that we are 
rapidly getting to the point where we have an expanding matrix of builds to 
fire off on an automated basis.  Each multiple of 5 in speed that we save is 5 
x more processing bandwidth we have.  That and the fact that Mac TrashCan Pros 
are religiously opposed to > 1 socket, this limits our access to the 24 virtual 
cores (12 real) that I used to enjoy.


Just a fun little diversion from "why doesn't this work?"


Here's my CrappleScript which wraps a shell script that I keep handy to make 
RAM drives if you feel so inclined for a taste of ludicrous speed.

on run {}

Start()

end run

on Start()

set myBlocks to 2097152 -- 1 GB in blocks
set myGigs to 6

set myResult to display dialog "Desired size in GB for your RAM Disk:" default 
answer myGigs
set myGigs to the text returned of myResult -- as string
set myBytes to myBlocks * myGigs
set myShellScript to "diskutil erasevolume HFS+ \"RAM Disk\" `hdiutil attach 
-nomount ram://" & myBytes & "`"

do shell script myShellScript
end Start


And the output:


tell application "Script Editor"
display dialog "Desired size in GB for your RAM Disk:" default answer 6
--> {button returned:"OK", text returned:"2"}
end tell
tell current application
do shell script "diskutil erasevolume HFS+ \"RAM Disk\" `hdiutil attach 
-nomount ram://4194304`"
--> "Started erase on disk5
Unmounting disk
Erasing
Initialized /dev/rdisk5 as a 2 GB case-insensitive HFS Plus volume
Mounting disk
Finished erase on disk5 RAM Disk"
end tell
Result:
"Started erase on disk5
Unmounting disk
Erasing
Initialized /dev/rdisk5 as a 2 GB case-insensitive HFS Plus volume
Mounting disk
Finished erase on disk5 RAM Disk"


 ___
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list  (Xcode-users@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/xcode-users/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: Xcode and compatible OS X versions

[Sean McBride]

On Fri, 9 Sep 2016 12:35:46 -0700, Carl Hoefs said:

>For various reasons, I'm having to develop on OS X 10.10.5, targeting
>iOS 9.x devices. Does Apple publish a matrix or table that tracks which
>versions of Xcode support developing for/running on which versions of
>macOS / iOS?  In the past I've had to download the 5GB+ Xcode package,
>only to find out that I would have to juggle SDKs and/or upgrade the OS
>to use it.  Do the forthcoming macOS 10.12 and Xcode 8 change this situation?

I dunno if Apple does, but Wikipedia does:



Cheers,

-- 

Sean McBride, B. Eng s...@rogue-research.com
Rogue Researchwww.rogue-research.com 
Mac Software Developer  Montréal, Québec, Canada



 ___
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list  (Xcode-users@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/xcode-users/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Xcode and compatible OS X versions

[Carl Hoefs]

For various reasons, I'm having to develop on OS X 10.10.5, targeting iOS 9.x 
devices. Does Apple publish a matrix or table that tracks which versions of 
Xcode support developing for/running on which versions of macOS / iOS?  In the 
past I've had to download the 5GB+ Xcode package, only to find out that I would 
have to juggle SDKs and/or upgrade the OS to use it.  Do the forthcoming macOS 
10.12 and Xcode 8 change this situation?

-Carl

 ___
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list  (Xcode-users@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/xcode-users/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Re: Missing functions???

[Howard Moon]

Never mind, I found the problem right after asking.

There was code like this:

#ifdef SOMETHINGSPECIAL
if (somethingspecial) {
#else
if (somethingnotsospecial) {
#endif
blahblahblah();
}

Apparently, the two opening braces confuse the Xcode parser too much, even 
though they are within separate #ifdef blocks, and only one can ever be 
compiled.

Now at least I can go back to debugging properly!

Thanks,
Howard


> On Sep 9, 2016, at 10:13 AM, Howard Moon  wrote:
> 
> I am trying to debug a problem in an audio plug-in (VST3), where none of my 
> controls (knobs, etc.) seem to be actually changing anything besides the GUI 
> itself.  There is a function called setParameter() that is called when a 
> control like this changes, and when I run it in the debugger, it steps into 
> that function, hits the switch statement that responds to each individual 
> specific control, but strangely it then simply steps over to the default 
> clause, which does nothing, and then exits the function.  I can see that the 
> variable used in the switch statement is a valid value that should cause it 
> to hit one of the case statements, but it does not go there.  Also, if I look 
> at the drop-down list of functions in that .cpp file, it says that I am 
> inside another function called timerChecks(), not in setParameter(), even 
> though I am clearly inside setParameter(). If I open that drop-down list of 
> functions, then the list abruptly ends at that timingChecks() function!
> 
> I’ve tried cleaning the project, quitting Xcode and deleting the entire build 
> folder, and rebuilding, but it’s still the same. It’s as if Xcode has no idea 
> I even have any functions after timingChecks(). If I enter invalid code in 
> any later function, the build fails on the line of garbage, but it still says 
> I am inside timingChecks().
> 
> How can I fix Xcode so that it sees all of my functions, and so that my code 
> executes the code I have actually written and compiled?
> 
> Thanks,
>   Howard
> 


 ___
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list  (Xcode-users@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/xcode-users/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Missing functions???

[Howard Moon]

I am trying to debug a problem in an audio plug-in (VST3), where none of my 
controls (knobs, etc.) seem to be actually changing anything besides the GUI 
itself.  There is a function called setParameter() that is called when a 
control like this changes, and when I run it in the debugger, it steps into 
that function, hits the switch statement that responds to each individual 
specific control, but strangely it then simply steps over to the default 
clause, which does nothing, and then exits the function.  I can see that the 
variable used in the switch statement is a valid value that should cause it to 
hit one of the case statements, but it does not go there.  Also, if I look at 
the drop-down list of functions in that .cpp file, it says that I am inside 
another function called timerChecks(), not in setParameter(), even though I am 
clearly inside setParameter(). If I open that drop-down list of functions, then 
the list abruptly ends at that timingChecks() function!

I’ve tried cleaning the project, quitting Xcode and deleting the entire build 
folder, and rebuilding, but it’s still the same. It’s as if Xcode has no idea I 
even have any functions after timingChecks(). If I enter invalid code in any 
later function, the build fails on the line of garbage, but it still says I am 
inside timingChecks().

How can I fix Xcode so that it sees all of my functions, and so that my code 
executes the code I have actually written and compiled?

Thanks,
Howard


 ___
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list  (Xcode-users@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/xcode-users/archive%40mail-archive.com

This email sent to arch...@mail-archive.com