Re: Aquamacs "not responding", "stuck", and no window on startup

[David Reitter]

David,
That would be

open -a Aquamacs —args -q

I am not sure what is causing this, given your description. You would want to delete the whole folders, not just the files you mention. Let me know if that helps.

Best,
David


--
http://aquamacs.org -- Aquamacs: Emacs on Mac OS X
http://aquamacs.org/donate -- Could we help you? Return the favor and support the Aquamacs Project!




> On Dec 12, 2014, at 12:58 PM, David R Forrest <d...@vims.edu> wrote:
>
> I've re-installed version 3.2 of Aquamacs from http://aquamacs.org/download-release.shtml on my Mac Yosemite, and cleaned out all of these files:
>
> ;;~/.emacs (deprecated -- meaning 'should not be used for new installations, but will continue to be supported' -- in Aquamacs on OS X)
> ;; /Library/Preferences/Emacs/Preferences.el
> ;; /Library/Preferences/Aquamacs Emacs/Preferences.el
> ;; ~/Library/Preferences/Emacs/Preferences.el
> ;; ~/Library/Preferences/Aquamacs Emacs/Preferences.el
>
> But when I attempt to start Aquamacs, either with 'open -a Aquamacs', select it from Applications in Finder, or from Spotlight, I get nothing but a "Aquamacs (not responding)" task in "Force Quit Applications" and "Activity Monitor", 51% CPU and sleepting/running in 'top'
>
> Is there a way to debug its startup more than on http://aquamacs.org/reporting-bugs.shtml ? I can't even get to the starting point of the point of the Help menu listed there.
>
> Maybe 'open -a Aquamacs --args --debug' or something?
>
> The top bit of the "Force Quit"-generated report is included below.
>
> Thanks,
> Dave
>
> ##########################
> Date/Time: 2014-12-12 11:44:03 -0500
> OS Version: 10.10.1 (Build 14B25)
> Architecture: x86_64
> Report Version: 21
>
> Command: Aquamacs
> Path: /Applications/Aquamacs.app/Contents/MacOS/Aquamacs
> Version: 3.2 (3.2.1415369726)
> Parent: launchd [1]
> PID: 5903
>
> Event: hang
> Duration: 6.30s (process was unresponsive for 982 seconds before sampling)
> Steps: 64 (100ms sampling interval)
>
> Hardware model: MacBookPro8,2
> Active cpus: 8
>
> Fan speed: 6196 rpm
>
> --------------------------------------------------
> Timeline format: stacks are sorted chronologically
> Use -i and -heavy to re-report with count sorting
> --------------------------------------------------
>
> Heaviest stack for the main thread of the target process:
> 64 start + 52 (Aquamacs + 12020) [0x100002ef4]
> 64 main + 5851 (Aquamacs + 792043) [0x1000c15eb]
> 64 Frecursive_edit + 423 (Aquamacs + 797271) [0x1000c2a57]
> 64 recursive_edit_1 + 207 (Aquamacs + 796639) [0x1000c27df]
> 64 internal_catch + 243 (Aquamacs + 1318131) [0x100141cf3]
> 64 top_level_1 + 79 (Aquamacs + 874095) [0x1000d566f]
> 64 internal_condition_case + 251 (Aquamacs + 1320011) [0x10014244b]
> 64 Feval + 106 (Aquamacs + 1323738) [0x1001432da]
> 64 eval_sub + 1092 (Aquamacs + 1310548) [0x10013ff54]
> 64 apply_lambda + 320 (Aquamacs + 1324160) [0x100143480]
> 64 exec_byte_code + 2331 (Aquamacs + 1563595) [0x10017dbcb]
> 64 Ffuncall + 649 (Aquamacs + 1325865) [0x100143b29]
> 64 exec_byte_code + 6053 (Aquamacs + 1567317) [0x10017ea55]
> 64 internal_lisp_condition_case + 554 (Aquamacs + 1319370) [0x1001421ca]
> 64 eval_sub + 1573 (Aquamacs + 1311029) [0x100140135]
> 64 Ffuncall + 649 (Aquamacs + 1325865) [0x100143b29]
> 64 exec_byte_code + 2331 (Aquamacs + 1563595) [0x10017dbcb]
>
>
>
>
>
>
>
>
>
>

[David R Forrest]

I still had some intermittent hanging problems, then with the below command-line invocation I found and deleted another unexpected file.


open -a Aquamacs --args --debug-init # discover startup parsing


The file I deleted was:

~/Library/Application\ Support/Aquamacs\ Emacs/site-start.el

...which seemed left over from an very old install of ECB. Maybe the intermittency was whether or not I visited certain sorts of files during a session?

I think it's all good now. It also seems to initialize lots quicker now.

Dave


> On Dec 13, 2014, at 11:36 AM, David R Forrest <d...@vims.edu> wrote:

>
>
>> On Dec 12, 2014, at 5:20 PM, David Reitter <david....@gmail.com> wrote:
>>
>> David,
>> That would be
>>
>> open -a Aquamacs —args -q
>>
>> I am not sure what is causing this, given your description. You would want to delete the whole folders, not just the files you mention. Let me know if that helps.
>>

>>> ...
>>>
>>
>

[benjamin....@gmail.com]

I've been dealing with this problem for a long time, but seem to have made some progress today. I had disposed of my .emacs.d already and cleaned up the Preferences directories. Upon starting Aquamacs the CPU would go red hot (hear the fan, see it on Activity Monitor), I noticed that after force quit, restarting would put me back into exactly the same state. SOLUTION: In ~/Library/Saved\ Application\ State/ I removed the directories (folders) org.gnu.Aquamacs.savedState/ and org.gnu.Emacs.savedState/ (the latter probably did not have to be removed, but I did ayway). Aquamacs started normally and promptly after that.

[Perry Wagle]

I'm running MacOS 10.11.  Aquamacs on fresh system (nothing in ~/Library) works once, but when Quit and restarted can't start any windows.  If I do the below (--args -q), then it kinda works, but won't remember the window size.  Is aquamacs dead?  Nothing since 2014...

On Friday, December 12, 2014 at 2:21:01 PM UTC-8, David Reitter wrote:

[David Reitter]

Perry,

If it works with -q, then there are some settings that are being retained.
Check what’s in Library. This issue (and the workaround) might apply:

https://github.com/davidswelt/aquamacs-emacs/issues/86

I’m running Aquamacs on 10.11 without problems, and I haven’t had reports like yours.

A new version of Aquamacs based on Emacs 25 is in the works.

- DR

--
http://aquamacs.org -- Aquamacs: Emacs on Mac OS X
http://aquamacs.org/donate -- Could we help you? Return the favor and support the Aquamacs Project!

> --
> You received this message because you are subscribed to the Google Groups "aquamacs-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to aquamacs-deve...@googlegroups.com.

[Perry Wagle]

If I rm "/Users/wagle/Library/Preferences/Aquamacs Emacs/.id",  then it starts "normally", but when I quit, it leaves am .id file which locks things up.

All the debugging seems tied to putting up windows to display things.  Is there a way to write to stderr or to a file without having a window?

Thanks! 

-- Perry

PS.  nightly has the same problem as of 12 hours ago.

[David Reitter]

Perry,

You may have some sort of networking issue. Could that be the case?

Try this:

(setq aquamacs-version-check-url nil)

in a file ~/.emacs.

- David

--
http://aquamacs.org -- Aquamacs: Emacs on Mac OS X
http://aquamacs.org/donate -- Could we help you? Return the favor and support the Aquamacs Project!

[Perry Wagle]

Hi David --

if the .id file exists when i start up aquamacs, then it gets stuck.  If I delete it first, it doesn't.  Repeatably.

If I --eval "(toggle-debug-on-error)" on the command line when the .id file exists, debugging is turned on in the menu bar (which is present, but can't do anything).  If I delete it first, debugging doesn't get turned on.  I was unable to view or dump the *messages* buffer to see why it turned on debugging.

I tried the (setq aquamacs-version-check-url nil) with an .id file present, and it got stuck.

I really think its creating bogus .id files.  Here's the current one:

6
735881
-733996611438218460
0
307

Thanks!

-- Perry

[David Reitter]

What does it show when it’s stuck?

Does Ctrl-g help in any way?

[Perry Wagle]

I decided to falsify my claim that it was a bogus .id file.  I succeeded:

If you delete the .id file, then if there is a crash report, it offers to report it.  This path leads to a working aquamacs.  If there isn't a crash report and there is no .id file, then it gets stuck (with my current preferences/etc in ~/Library/).

I figure you are in the UTC time zone, so are getting ready to sleep, o I'm giving progress reports in case you see something along the way.

Thanks!

-- Perry

[Perry Wagle]

It shows nothing but the menu bar, and most of it can't do anything without being able to put up windows.

There is no window to press control-g into.  I tried anyway, and nothing.

[Perry Wagle]

Okay, it can't handle my external monitor on my old MacBookPro when it has a frame-positions el file  If I unplug my external monitor, then it works fine.
Aquamacs worked fine in MacOS 10.10, but apparently not in 10.11.

$ cat "/Users/wagle/Library/Preferences/Aquamacs Emacs/frame-positions.el"
;; Saved Frame Positions


(if (fboundp (quote smart-frame-set-initial-frame-alist)) (smart-frame-set-initial-frame-alist (quote ((left . 0) (top . 23) (width . 129) (height . 35)))))



(setq smart-frame-prior-positions (quote nil))

I hope this is an easy fix...  8/

-- Perry

[Benjamin Grinstein]

Have you tried deleting the org.gnu.Aquamacs.savedState folder under 

"~/Library/Saved Application State”

This fixed the problem for me every time.

Ben

Too short? Here’s why: http://emailcharter.org

You received this message because you are subscribed to a topic in the Google Groups "aquamacs-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/aquamacs-devel/VwRkKvGEcsQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to aquamacs-deve...@googlegroups.com.

[Perry Wagle]

Didn't work for me. 8/  Maybe there's more than one problem.

[David Reitter]

OK, so this is it, right? No relation to the .id file.

I’ll hook up an external screen to my MBP to test.
- D

[Perry Wagle]

Well, having no .id file AND having a crash dump sends it on a code path where it works.

This seems to have started with El Capitan (10.11).

-- Perry

[Perry Wagle]

Any "luck"?

Thanks!

-- Perry

[David Reitter]

Perry,

I did reproduce some problem where after a configuration change the new frame would open off-screen.

The next version of Aquamacs could have a fix for this problem. Would you like to look at the code and develop a solution? I think that when frames are created, it should check whether the coordinates are available at that time.

- David


--
http://aquamacs.org -- Aquamacs: Emacs on Mac OS X
http://aquamacs.org/donate -- Could we help you? Return the favor and support the Aquamacs Project!

[Perry Wagle]

Do you know whereabouts that needle is in the haystack?  I can probably fix the needle, but not find it.

-- Perry

[David Reitter]

Start by looking at frame-creation-function-alist, which points you to a function in the smart-frame-positioning package, which might be where one can take care of this.

[Perry Wagle]

Oh interesting: with system preferences, if you arrange the auxiliary monitor ABOVE the builtin monitor, like mine is, it fails.  If you arrange them alongside each other, it works.

-- Perry

[Perry Wagle]

Which version of the source code should I use?

[David Reitter]

You can get the git repository, aquamacs3 branch. I’m not sure what else you would choose.

Later this week or next, I will be pushing a completely new repository for Aquamacs based on Emacs 25, but the problem you’re looking at should not interact.

[Perry Wagle]

I'm sorta helplessly spinning my wheels, getting not much of anywhere.

I can make modifications to "/Applications/Aquamacs.app/Contents/Resources/lisp/aquamacs/smart-frame-positioning.el", but all of the symbols so far have turned out to be unbound.  And I search your github repo, and find no definitions.

Can you come up with a better hint for what to look for?  The smart-frame-positioning.el file does indicate something about needing to switch from carbon to cocoa, but I haven't even seen where the elisp is calling out to carbon (initial-window-system is 'ns).

If I cause an error in smart-frame-positioning.el, then it aborts and gives me a window.  If I don't cause an error, then it can't put up windows, and no error messages are in the *Messages* buffer (I wrote an exit hook to write it to a file).

-- Perry

[cris....@gmail.com]

I was having similar symptoms today and read through this thread.  If you have not tried just renaming ~/Library/Preferences/Aquamacs Emacs/, it might be worth a try, and easily reversible. Anyway, it worked for me. Aquamacs recreated it automatically.

[David Reitter]

You’re going to want to work in a system configuration (same external monitor all the time) that doesn’t cause the problem.

You know what causes it - it’s trying to open a frame on a display that has since been removed. (If that’s not correct, then we need to re-assess this problem - but that’s how I reproduced it.)
So, find the function that determines the location of a new frame, and modify it so it knows more about the current display geometry.

--
http://aquamacs.org -- Aquamacs: Emacs on Mac OS X
http://aquamacs.org/donate -- Could we help you? Return the favor and support the Aquamacs Project!

[Perry Wagle]

I reproduce the problem it by configuring 10.11 (El Capitan) "System Preferences" -> "Displays" -> "Placement" to have the external monitor right above the notebook.  Then it leaves the apparently fatal "frame-positions.el" file in my preferences when it exits.

I *avoid* the problem by configuring the screen placement to put the external monitor to the right of the notebook monitor.  Then it repeatably works.

My notebook is in the shop (trackpad broken) for 3-5 days, so I won't get to test more until next week sometime.

The external monitor is not the same pixel size as the notebook (MBP), something that the comments in the "smart-frame-positioning.el" library file warns about.

-- Perry

[Perry Wagle]

PS.  This worked fine in 10.10 (Yosemite), so it's something about 10.11 (El Capitan).

[David Reitter]

That would explain why I haven’t seen a report about this before.

I think I tried a different display placement, so perhaps that’s why.

What the “main screen” is is probably important, as it determines where the zero point is. Perhaps El Capitan deals with coordinates on multiple displays differently. It would be worth checking Apple’s release notes.

> On Oct 16, 2015, at 8:02 PM, Perry Wagle <pwa...@gmail.com> wrote:
>
> PS. This worked fine in 10.10 (Yosemite), so it's something about 10.11 (El Capitan).
>
> On Friday, October 16, 2015 at 5:00:46 PM UTC-7, Perry Wagle wrote:
> I reproduce the problem it by configuring 10.11 (El Capitan) "System Preferences" -> "Displays" -> "Placement" to have the external monitor right above the notebook. Then it leaves the apparently fatal "frame-positions.el" file in my preferences when it exits.
>
> I *avoid* the problem by configuring the screen placement to put the external monitor to the right of the notebook monitor. Then it repeatably works.
>
> My notebook is in the shop (trackpad broken) for 3-5 days, so I won't get to test more until next week sometime.
>
> The external monitor is not the same pixel size as the notebook (MBP), something that the comments in the "smart-frame-positioning.el" library file warns about.
>

> — Perry

[Perry Wagle]

I don't know the apple api for either 10.10 or 10.11, and aquamacs worked in the former.  I don't know the aquamacs source, and the elisp file I'm looking at has nothing I find obvious in terms of how it works.  Apparently, all the magic is off someplace else in the source.  That's a huge haystack you are expecting me to understand.  Would take me weeks.

 But as I think I said before, point me at the needle in the haystack, and I can probably fix it.

Aquamacs is the only app of mine confused by the two monitors, one above the other.

-- Perry

[Perry Wagle]

I saw something that indicated that things needed to be compiled under the latest Xcode for MacOS 10.11, so I tried to do that.

Either I'm not doing it right (found 2 contradictory instructions), or it won't build.

-- Perry

[Perry Wagle]

Never mind.  Apparently installing the command line tools from the GUI that said it'd do this isn't enough.  You have to "xcode-select --install" to get /usr/include.

-- Perry



[Perry Wagle]

I Compiling under XCode 7 in 10.11 doesn't help.

[David Reitter]

As I said, my best guess would be that this code is in smart-frame-positioning (or perhaps one-buffer-one-frame).

Emacs Lisp comes with many levels of indirection, which makes it difficult to penetrate. But, over time, it gets easier.

D

[Perry Wagle]

I'm wondering what to make of the repeated idea in the comments that multi-monitor support is simplified solution pending The Right Way To Do It, for example:

;; A better implementation could be done once the Cocoa port is
;; available, so we're not going to invest much time in the Carbon
;; portion here.

in smart-frame-positioning.el, and similar language in other places.

[David Reitter]

Perry,

according to the git log, this comment dates back to 2006, and changes have happened since then.
Of course we _are_ on the Cocoa port, and have been for a long time.

What is true is that this library implements a lot of lower-level functionality about window positioning that probably could be achieved by simply using the higher-level Cocoa API (which Emacs does not expose to Lisp). Generally, the same is true for a lot of Emacs code.

- David

[Perry Wagle]

The objC function ns_make_monitor_attribute_list in nsfns.m seems confused by the fact that screen 0 is the external monitor, and screen one doesn't have the emacs-frame on it, so it all returns no frames in the frame list.


PS  As for the below, its like going to France, not knowing French, taking a taxi, taxi breaks down, and driver (with a prominent "master mechanic" tattoo) says "well find the toolbox, and fix it!"..  So you find the toolbox, and a manual.  You laboriously translate the manual, and it says "car design is incomplete and needs to be modified".  You point this out to the driver, and he points to the copyright date for that page.  You still figure something is fundamentally misdesigned, but you ontinue because you have to get to where you are going.  And, you are already taking time off from fixing the Mozilla train-wreck that you are similarly being expected to fix.

[Perry Wagle]

In the below, "NSScreen *screen = [[view window] screen];" returns null for the second (internal) screen, this makes searching for it impossible.  This is documented to happen "The value of this property is the screen where most of the window is on; it is nil when the window is offscreen.", but that was true for 10.10 apparently, so something else is wrong.

I overrode that search with the below "i = 1;", and it worked.  Not a fix, but an indication that we're close.

I'm not sure where to go from here.


static Lisp_Object
ns_make_monitor_attribute_list (struct MonitorInfo *monitors,
                                int n_monitors,
                                int primary_monitor,
                                const char *source)
{
  Lisp_Object monitor_frames = Fmake_vector (make_number (n_monitors), Qnil);
  Lisp_Object frame, rest;
  //NSArray *screens = [NSScreen screens];
  NSArray<NSScreen *> *screens = [NSScreen screens];
  int i;

  FOR_EACH_FRAME (rest, frame)
    {
      struct frame *f = XFRAME (frame);

      if (FRAME_NS_P (f))
        {
          NSView *view = FRAME_NS_VIEW (f);
          NSScreen *screen = [[view window] screen];
          NSUInteger k;
          i = -1;
          for (k = 0; i == -1 && k < [screens count]; ++k)
            {
              //if ([screens objectAtIndex: k] == screen) {
              //  i = (int)k;
                i = 1;
              //}
            }

          if (i > -1)
            ASET (monitor_frames, i, Fcons (frame, AREF (monitor_frames, i)));
        }
    }
  return make_monitor_attribute_list (monitors, n_monitors, primary_monitor,
                                      monitor_frames, source);
}

[Perry Wagle]

Actually, setting i to 1 (the second index) shoves both (?) frames into the second element of what "(ns-display-monitor-attributes-list)" returns in the elisp.  This is probably wrong:

(((name . DELL E2015HV) (geometry 0 0 1680 1050) (workarea 0 23 1655 1027) (mm-size 330 206) (frames) (source . NS)) ((geometry 80 -900 1600 900) (workarea 80 -900 1600 877) (mm-size 432 235) (frames #<frame  Aquamacs 0x1048b5650> #<frame Aquamacs 0x101a9bb90>) (source . NS)))

[David Reitter]

> On Oct 24, 2015, at 12:48 AM, Perry Wagle <pwa...@gmail.com> wrote:
>
> The objC function ns_make_monitor_attribute_list in nsfns.m seems confused by the fact that screen 0 is the external monitor, and screen one doesn't have the emacs-frame on it, so it all returns no frames in the frame list.

Looking at it, I don’t quite understand - why would it matter if the monitor is external or not?

I thought all of this was more an issue with smart-frame-positioning, but it seems I may be wrong.

> PS As for the below, its like going to France, not knowing French, taking a taxi, taxi

You are spot on about this. And if I may add, your taxi driver drives 20 different taxis, each a different model. In C.S., the cars have gotten more complex over the past 40 years, but also easier to work on. Emacs, on the other hand, is 40 years old, and complex.


D

[Perry Wagle]

Do you know why there are two frames and not just one?

On Saturday, October 24, 2015 at 3:25:31 AM UTC-7, David Reitter wrote:

> On Oct 24, 2015, at 12:48 AM, Perry Wagle <pwa...@gmail.com> wrote:
>
> The objC function ns_make_monitor_attribute_list in nsfns.m seems confused by the fact that screen 0 is the external monitor, and screen one doesn't have the emacs-frame on it, so it all returns no frames in the frame list.

Looking at it, I don’t quite understand - why would it matter if the monitor is external or not?

In the apple documentation, it says screen 0 is the main screen.  I thought maybe this assumption was built into aquamacs somewhere (else).
 

I thought all of this was more an issue with smart-frame-positioning, but it seems I may be wrong.

There are two emacs-frames, apparently.  Why two and not one?  Tracking the Vframe_list global is necessary, and its all over the place.

What direction for debugging now?

> PS  As for the below, its like going to France, not knowing French, taking a taxi, taxi

You are spot on about this.  And if I may add, your taxi driver drives 20 different taxis, each a different model. In C.S., the cars have gotten more complex over the past 40 years, but also easier to work on.   Emacs, on the other hand, is 40 years old, and complex.

 

Was worried that you wouldn't get the joke.  But, my taxi driver designed the taxi.  Oh well.
 

D

[David Reitter]

On Oct 24, 2015, at 8:07 AM, Perry Wagle <pwa...@gmail.com> wrote:
>
> Do you know why there are two frames and not just one?

Maybe.

Aquamacs has to create a hidden frame some times in order to accommodate a situation where the application has no visible frames but isn’t quit. GNU Emacs cannot handle a state in which no frame is selected, while standard applications on the Mac have to be able to do that. Hence we’re creating a hidden frame which serves as the recipient of keyboard events when all other frames are closed.

This could be it - or you’re seeing something else.

- David

[Perry Wagle]

Well, cool.  The very latest clone from aquamacs3 works.  It even remembers which screen the emacs-frame was last on, which is better than my hack.

Currently seeing if I can figure out what changed.

-- Perry

[David Reitter]

Good. I didn’t fix it. Must’ve come from Emacs.

[Perry Wagle]

Well, thanks for merging!  8)

[Iliya Lefterov]

you are terrific! 

Aquamacs… is even better…:)

-I.

--

Iliya Lefterov, MD, PhD
University of Pittsburgh

ili...@pitt.edu
ilef...@icloud.com

[mac.das...@gmail.com]

I this something for which there is (eyes turn to stars) an installer somewhere?  I have Aquamacs 3.2 (Nov. 2014), and I notice that Check for Updates says, "3.2a is the most recent Aquamacs version available," but I can't find a way to download 3.2a.

If I must build starting from git... okay, not sure what support stuff other than XCode 7.1 and the command line tools I wd need.

I do indeed have my big monitor positioned above the laptop screen-- and I connect and disconnect it from the computer frequently.

 --Steve

[mac.das...@gmail.com]

The current http://jazz.aquamacs.org/~dreitter/Aquamacs/24/Aquamacs-nightly.tar.bz2 is empty; the end of http://jazz.aquamacs.org/~dreitter/Aquamacs/24/aquamacs-build.log looks like some errors to me.

So I am trying to build from

git://github.com/davidswelt/aquamacs-emacs.git

using XCode 7.1 on El Capitan.

"24" and the current branch "origin/aquamacs3" disagree, but I'm not against building something new.

Unless I'm going down a bad path here, please ignore my previous call for help...

 --Steve

[David Reitter]

On Oct 26, 2015, at 6:26 PM, mac.das...@gmail.com wrote:
>
> The current http://jazz.aquamacs.org/~dreitter/Aquamacs/24/Aquamacs-nightly.tar.bz2 is empty; the end of http://jazz.aquamacs.org/~dreitter/Aquamacs/24/aquamacs-build.log looks like some errors to me.

Oh. OK, I’ll look into it. It shouldn’t be empty.

> So I am trying to build from
> git://github.com/davidswelt/aquamacs-emacs.git
> using XCode 7.1 on El Capitan.
>
> "24" and the current branch "origin/aquamacs3" disagree, but I'm not against building something new.

Yes, see my message to the list. git pull -f.

D

[mac.das...@gmail.com]

On Monday, October 26, 2015 at 6:33:52 PM UTC-4, David Reitter wrote:

On Oct 26, 2015, at 6:26 PM, mac.das...@gmail.com wrote:
>
> The current http://jazz.aquamacs.org/~dreitter/Aquamacs/24/Aquamacs-nightly.tar.bz2 is empty; the end of http://jazz.aquamacs.org/~dreitter/Aquamacs/24/aquamacs-build.log looks like some errors to me.

Oh.  OK, I’ll look into it.  It shouldn’t be empty.

Thanks. 

 

> So I am trying to build from
> git://github.com/davidswelt/aquamacs-emacs.git
> using XCode 7.1 on El Capitan.
>
> "24" and the current branch "origin/aquamacs3" disagree, but I'm not against building something new.

Yes, see my message to the list.

 

Sorry, which message to which list?  I am not subscribed by email and the archives seem to end in early Oct.  

 git pull -f.

 

$ git pull -f

From git://github.com/davidswelt/aquamacs-emacs

   47e9556..9d6ec23  emacs      -> origin/emacs

Already up-to-date.

$ 

 

This built and launches, but right away I noticed it won't paste.   Here's from its about screen:

Aquamacs 3.3dev  GNU Emacs 25.0.50.1 (x86_64-apple-darwin15.0.0, NS appkit-1404.13 Version 10.11.1 (Build 15B42))

 of 2015-10-26 on 1dc84f86e5711e9015e84bf14c388359b5157cc2

The error message is: "Symbol's function definition is void: ns-get-pasteboard".

From a message called "New Repository" I get the feeling I shouldn't be trying to build 3.3 at all since it's bleeding edge, yes?

 --Steve

[David Reitter]

On Oct 26, 2015, at 8:09 PM, mac.das...@gmail.com wrote:

> Sorry, which message to which list? I am not subscribed by email and the archives seem to end in early Oct.

I meant the developer list. The archives are here:

https://groups.google.com/forum/#!forum/aquamacs-devel

See message from Oct 24.

> This built and launches, but right away I noticed it won't paste. Here's from its about screen:
> Aquamacs 3.3dev GNU Emacs 25.0.50.1 (x86_64-apple-darwin15.0.0, NS appkit-1404.13 Version 10.11.1 (Build 15B42))
> of 2015-10-26 on 1dc84f86e5711e9015e84bf14c388359b5157cc2
>
> The error message is: "Symbol's function definition is void: ns-get-pasteboard”.

I fixed this today. I just committed the change, so try again.

> From a message called "New Repository" I get the feeling I shouldn't be trying to build 3.3 at all since it's bleeding edge, yes?

Ha, well, do build from it. We need to sort out the bugs somehow!

I have started using it for my editing. It’s not that terrible now that copy/paste works. :-)

[mac.das...@gmail.com]

On Monday, October 26, 2015 at 9:41:22 PM UTC-4, David Reitter wrote:

On Oct 26, 2015, at 8:09 PM, mac.das...@gmail.com wrote:

> [With the previous 33dev] The error message is: "Symbol's function definition is void: ns-get-pasteboard”.

I fixed this today.  I just committed the change, so try again.

Yay!  Thanks!  It has built here and I confirmed that the paste is working.

 

> From a message called "New Repository" I get the feeling I shouldn't be trying to build 3.3 at all since it's bleeding edge, yes?

Ha, well, do build from it.  We need to sort out the bugs somehow!

I have started using it for my editing.  It’s not that terrible now that copy/paste works. :-)

Okay!  I have put it at /Applications/Aquamacs and will use it.

 --Steve