FvwmPager: correct fonts and balloons initialization #1189

ONykyf · 2025-05-03T18:24:06Z

Prevent some weird FvwmPager crashes and balloons not showing up because fonts and boolean flags are not initialized an start. Also restore the functionallity of "Balloons" option without parameters.

Sometimes (and almost always when swallowed by FvwmButtons) FvwmPager silently crashes. Debugging has shown that this happens in do_label_window() function because Scr.winFfont is uninitialized. Also balloons are not always shown in FvwmPager because Balloon.show_in_pager and Balloon.show_in_icon boolean flags are undetermined, and
Balloons with no parameters has no effect. This has been corrected.

ThomasAdam · 2025-05-09T11:25:15Z

Hi @ONykyf

This might fix the immediate issue, but I don't think it is the way to go.

I would rather see the Balloon struct initialised in initialize_balloon_window() -- you might have to rejig the ordering of how that's called in init_fvwm_pager().

ONykyf · 2025-05-11T09:24:20Z

All fallback fonts initialization is in initialize_pager_window() and initialize_balloon_window() now.

ThomasAdam · 2025-05-11T09:55:32Z

Hi @ONykyf

Better. Thanks! But you need to squash those two commits together, please.

ThomasAdam · 2025-05-15T20:15:31Z

Hi @ONykyf

FvwmPager now segfaults with this change. I'm presuming something like the following fixes it:

diff --git a/modules/FvwmPager/x_pager.c b/modules/FvwmPager/x_pager.c
index 7dc6819d0..c2ca3f71b 100644
--- a/modules/FvwmPager/x_pager.c
+++ b/modules/FvwmPager/x_pager.c
@@ -336,7 +336,7 @@ rectangle set_vp_size_and_loc(struct fpmonitor *m, bool is_icon)
 
 void set_desk_size(bool update_label)
 {
-       if (update_label) {
+       if (update_label && Scr.Ffont != NULL) {
                label_h = 0;
                if (use_desk_label)
                        label_h += Scr.Ffont->height + 2;

But you need to be very careful with the testing you do...

For reference:

Core was generated by `/usr/local/libexec/fvwm3/1.1.3/FvwmPager 11 5 none 0 8 * *'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  set_desk_size (update_label=true) at ../modules/FvwmPager/x_pager.c:342
342                             label_h += Scr.Ffont->height + 2;
(gdb) bt
#0  set_desk_size (update_label=true) at ../modules/FvwmPager/x_pager.c:342
#1  0x000055870ce4b641 in initialize_pager_size () at ../modules/FvwmPager/init_pager.c:410
#2  0x000055870ce4a763 in init_fvwm_pager () at ../modules/FvwmPager/init_pager.c:68
#3  0x000055870ce49f56 in main (argc=8, argv=0x7ffe1247a518) at ../modules/FvwmPager/FvwmPager.c:292
(gdb) p 2
$1 = 2
(gdb) l
337     void set_desk_size(bool update_label)
338     {
339             if (update_label) {
340                     label_h = 0;
341                     if (use_desk_label)
342                             label_h += Scr.Ffont->height + 2;
343                     if (use_monitor_label)
344                             label_h += Scr.Ffont->height + 2;
345             }
346     
(gdb) f 2
#2  0x000055870ce4a763 in init_fvwm_pager () at ../modules/FvwmPager/init_pager.c:68
68              initialize_pager_size();
(gdb) f 1
#1  0x000055870ce4b641 in initialize_pager_size () at ../modules/FvwmPager/init_pager.c:410
410             set_desk_size(true);
(gdb) quit

ONykyf · 2025-05-16T06:50:30Z

Yes, this is because initialize_pager_size() comes before initialize_pager_window(), and font initialization is in the latter. I think it is better to move it to the first, then the proposed check Scr.Ffont != NULL is unnecessary, although can be retained for safety.

somiaj · 2025-05-16T12:31:06Z

Be careful moving the order around, it is fairly specific and moving the calls around may have other unattended consquences. Also realize not everyone uses fonts, and *FvwmPager: Font None is considered a valid option, it is always best to check that fonts exist before trying to use them.

somiaj · 2025-05-16T12:35:38Z

modules/FvwmPager/init_pager.c

+	if (use_window_label && !Scr.Ffont)
+		Scr.winFfont = FlocaleLoadFont(dpy, NULL, MyName);


Shouldn't this be !Scr.winFfont here to match what you are actually setting.

Thank You, it should be !Scr.winFfont indeed.

somiaj · 2025-05-16T12:36:30Z

modules/FvwmPager/x_update.c

-	if (t == NULL || w == None || Scr.winFfont == NULL ||
+	if (t == NULL || w == None || !use_window_label ||


I don't think you want to remove the Null check here.

Now the logic of the code is such that once initialize_pager_size() is completed, we have either use_window_label == false and Scr.winFfont == NULL, or use_window_label == true and Scr.winFfont == whatever non-zero, which is either default or non-default font. But inside parse_options(), yes, it is safer to leave Scr.winFfont == NULL check as is.

ThomasAdam · 2025-05-16T16:28:11Z

I can't say I like the churn on this change -- I'm not a fan.

We seem to have now decided on resetting Balloon.Ffont and Scr.Ffont back to NULL in several places along different code paths, which just makes this look a bit of a mess.

I would say "try again". I'm sure you can think of a better way, @ONykyf...

ONykyf · 2025-05-16T17:02:55Z

Maybe I have not understood the intentions behind this code well, then I'd like to ask for a clarification.

My idea was the following: first we declare what we are going and what we are not going to use:
use_desk_label = true; use_window_label = true; use_monitor_label = false;

All fonts are set to NULL == not set. Then parse_options() can consume "Font", "WindowFont", "SmallFont", or "BalloonFont" options arbitrarily many times. Each time, if the option value is "None", then the respective boolean flag is set to false (we have no intention to use this), and the font pointer to NULL. Otherwise, if the option value is not "None", we set the respective font. We can reconsider, then the previous choice is overwritten.

Finally, if we have not made a choice, but the intention to use some font has not been rejected, then the default font is used.

Is this logic acceptable?

somiaj · 2025-05-17T18:24:09Z

modules/FvwmPager/init_pager.c

+		} else if (StrEquals(resource, "Balloons")) {
+			Balloon.show_in_pager = true;
+			Balloon.show_in_icon = true;


This should not be here, this is already set on lines 1225, and correctly honors the options, this is overriding what is done below and making it so the option does not work according to the manual page.

Oh I see what this is doing, but it is not doing it correctly. It does appear that the parser will ignore *FvwmPager: Balloons, without any options. It does state that this should enable balloons in the manual page, though really just use *FvwmPager: Balloons All or anything except null and it will process correctly. I'm not sure the best thing to do here, but this is not it. This will make it so Balloons None or Balloons Pager will fail.

I might just update the manual page to state that you have to provide an option vs try to work around dealing with it being Null.

Note the default config is using *FvwmPager: Balloons All

Oh I see what this is doing, but it is not doing it correctly. It does appear that the parser will ignore *FvwmPager: Balloons, without any options. It does state that this should enable balloons in the manual page, though really just use *FvwmPager: Balloons All or anything except null and it will process correctly. I'm not sure the best thing to do here, but this is not it. This will make it so Balloons None or Balloons Pager will fail.

I might just update the manual page to state that you have to provide an option vs try to work around dealing with it being Null.

I just followed FvwmPager.adoc (and hence FvwmPager.1): "If type is anything else (or null) balloons are always shown", so no need to change the man page. I had *FvwmPager: Balloons and then *FvwmPager: BalloonFont "xft:Sans:size=8:medium:antialias=True" in my config, and this did not worked (not shown).

This will make it so Balloons None or Balloons Pager will fail.

You are right, flags = 0 should be added there to continue parsing.

This should not be here, this is already set on lines 1225, and correctly honors the options, this is overriding what is done below and making it so the option does not work according to the manual page.

I would not say so, on lines 1225-1238 Balloon.show_in_pager and Balloon.show_in_icon are set only if Balloons option with (and now without) parameters is used. Otherwise these flags are never initialized. Probably this should also be corrected?

When I updated the way the parser worked, I just overlooked this, and should have updated the manual page at that time, I think it is okay to require a parameter and not make it optional.

Even if you added flags = 0 to continue parsing, it would still fail because it won't continue if there is no argument. I am going to update the manual page, *FvwmPager: Balloons is now invalid, use *FvwmPager: Balloons All, the updates to the parser make it not worth adding a work around to accept null in this case any more.

Could You explain why it would still fail? If there is no argument to Balloons, it will just free resource and skip to the next cycle iteration, but Balloon.show_in_pager and Balloon.show_in_icon have been already set to true, so everything is ok.

Of course You can remove this option, but I see no problem with it now.

if (!arg1) {
/* No inputs, nothing to do */
free(resource);
continue;
}

This statement below guards the cycle body from unnecessary continuation, doesn't it?

Line 990 bellow,

if (!arg1) /* No inputs, nothing to do. */ continue;

If there is no arg, it will also not continue processing. All options below are expected to have an argument, which is why I said I didn't want to design a special case for this option when just not allowing null and removing the comment that it is allowed from the manual page is my preferred solution. Really this would be a single exception that really isn't needed. People can update their configuration.

People can update their configuration.

Yep. I'm more than happy for this -- wouldn't be the first time, and certainly will not be the last. ;)

Ok, removing (null) case.

somiaj · 2025-05-17T18:32:03Z

Can you share a config that causes the crash you are trying to fix? That way we can test what is actually crashing. The default-config and I use balloon windows and I have never seen FvwmPager crash due to this when swallowed inside of FvwmButtons. I also think your fix is trying to add functionality we may not need, but if we better knew what was crashing we can maybe suggest a better way to fix the initialization issue (really testing of a font is Null is a sufficient test, a new variable doesn't need to be added).

Also why are you adding a new configuration option "Balloons" without any documentation on what it does?

ONykyf · 2025-05-18T08:14:12Z

Can you share a config that causes the crash you are trying to fix? That way we can test what is actually crashing. The default-config and I use balloon windows and I have never seen FvwmPager crash due to this when swallowed inside of FvwmButtons.

In fact FvwmPager crashed always, but immediately when swallowed by FvwmButtons, or after Midnight Commander (or any other program like xterm that changes its title while running) is started. Nothing interesting in the config, just

*FvwmPager: UseSkipList
*FvwmPager: Back grey60
*FvwmPager: Fore MidnightBlue
*FvwmPager: SolidSeparators
*FvwmPager: Font "xft:Sans:size=8:medium:antialias=True"
*FvwmPager: Hilight #c3c3c3
*FvwmPager: Geometry 90x70-25-35
*FvwmPager: SmallFont "xft:Sans:size=7:medium:antialias=True"
*FvwmPager: Balloons
*FvwmPager: BalloonFont "xft:Sans:size=8:medium:antialias=True"

I also think your fix is trying to add functionality we may not need, but if we better knew what was crashing we can maybe suggest a better way to fix the initialization issue (really testing of a font is Null is a sufficient test, a new variable doesn't need to be added).

Also why are you adding a new configuration option "Balloons" without any documentation on what it does?

As You can see, Balloons option is already there and does not work as announced, so no new new functionality, just aligning the code with the manual page.

New boolean variable use_window_label is necessary to distinguish two situations when Scr.winFfont == NULL : when neither WindowFont nor SmallFont has not been set, but we accept the default font, from the case when WindowFont or SmallFont has been set to "None", hence no font should be loaded at all. Once the initialization is complete, this variable is not used anymore.

ONykyf · 2025-05-18T09:08:23Z

Corrected so that now Balloons option looks for an argument (thanks to @somiaj).

Set initial values for Balloon.show_in_pager and Balloon.show_in_icon to true to avoid a lottery if no options are set by a user (or different defaults are preferred?)

ONykyf · 2025-05-18T13:52:27Z

@somiaj I would rather leave editing manpage to You, I am not a native English speaker.

somiaj · 2025-05-18T15:41:25Z

I really want to track down the crash you are experiencing, because I want to ensure that the changes are needed. So far I cannot reproduce the crash, I have used your configuration and swallowed it in FvwmButtons as shown.

DestoryModuleConfig FP1:*
*FP1: UseSkipList
*FP1: Back grey60
*FP1: Fore MidnightBlue
*FP1: SolidSeparators
*FP1: Font "xft:Sans:size=8:medium:antialias=True"
*FP1: Hilight #c3c3c3
*FP1: Geometry 90x70-25-35
*FP1: SmallFont "xft:Sans:size=7:medium:antialias=True"
*FP1: Balloons All
*FP1: BalloonFont "xft:Sans:size=8:medium:antialias=True"

DestoryModuleConfig FPButtons:*
*FPButtons: Geometry 400x400
*FPButtons: Rows 1
*FPButtons: Columns 1
*FPButtons: (1x1, Swallow FP1 'Module FvwmPager FP1 *', Frame 0)

Module FvwmButtons FPButtons

I have changed the title of my xterm multiple times and it is not crashing for me.

ONykyf · 2025-05-18T17:04:56Z

I really want to track down the crash you are experiencing, because I want to ensure that the changes are needed. So far I cannot reproduce the crash, I have used your configuration and swallowed it in FvwmButtons as shown.
DestoryModuleConfig FP1:*
*FP1: UseSkipList
*FP1: Back grey60
*FP1: Fore MidnightBlue
*FP1: SolidSeparators
*FP1: Font "xft:Sans:size=8:medium:antialias=True"
*FP1: Hilight #c3c3c3
*FP1: Geometry 90x70-25-35
*FP1: SmallFont "xft:Sans:size=7:medium:antialias=True"
*FP1: Balloons All
*FP1: BalloonFont "xft:Sans:size=8:medium:antialias=True"

DestoryModuleConfig FPButtons:*
*FPButtons: Geometry 400x400
*FPButtons: Rows 1
*FPButtons: Columns 1
*FPButtons: (1x1, Swallow FP1 'Module FvwmPager FP1 *', Frame 0)

Module FvwmButtons FPButtons
I have changed the title of my xterm multiple times and it is not crashing for me.

Could You try the following with and without *FP1: SmallFont None line:

DestoryModuleConfig FP1:*
*FP1: UseSkipList
*FP1: Back grey60
*FP1: Fore MidnightBlue
*FP1: SolidSeparators
*FP1: Font "xft:Sans:size=8:medium:antialias=True"
*FP1: Font None
*FP1: Hilight #c3c3c3
*FP1: Geometry 90x70-25-35
*FP1: SmallFont "xft:Sans:size=7:medium:antialias=True"
*FP1: SmallFont None
*FP1: Balloons All
*FP1: BalloonFont "xft:Sans:size=8:medium:antialias=True"
*FP1: BalloonFont None

DestoryModuleConfig FPButtons:*
*FPButtons: Geometry 400x400+0+100
*FPButtons: Rows 1
*FPButtons: Columns 1
*FPButtons: (1x1, Swallow(UseOld,NoClose) FP1 'Module FvwmPager FP1 *', Frame 0)

Module FvwmButtons FPButtons

Please also check whether None lines take effect.

somiaj · 2025-05-18T17:17:12Z

Ahh I see now, setting the font to None after it has been set does cause a crash. This is due to bad configuration (no reason to set the font then not set it), but I will think about a good way to protect against that.

ONykyf · 2025-05-18T17:18:35Z

Ahh I see now, setting the font to None after it has been set does cause a crash. This is due to bad configuration (no reason to set the font then not set it), but I will think about a good way to protect against that.

That was the reason why I proposed to set NULL in this case.

In fact there is a reason: I use "layered" configurations when a derived one changes settings of a parent.

I believe that options parsing should be fool-proof or at least override-proof.

somiaj · 2025-05-18T17:25:12Z

I am able to fix that issue with a single line:

--- a/modules/FvwmPager/init_pager.c
+++ b/modules/FvwmPager/init_pager.c
@@ -1187,6 +1187,7 @@ void parse_options(void)
                           StrEquals(resource, "SmallFont"))
                {
                        FlocaleUnloadFont(dpy, Scr.winFfont);
+                       Scr.winFfont = NULL;
                        if (strncasecmp(next, "none", 4) != 0)
                                Scr.winFfont = FlocaleLoadFont(
                                        dpy, next, MyName);

somiaj · 2025-05-18T17:29:01Z

modules/FvwmPager/init_pager.c

+	/* balloons are shown by default */
+	Balloon.show_in_pager = true;
+	Balloon.show_in_icon = true;


Please don't change default behavior. Balloons are not shown by default.

Please don't change default behavior. Balloons are not shown by default.

I have not found the place where Balloon.show_in_pager and Balloon.show_in_icon are set to false. I asked what is the preferred initial behaviour, but it must be set.

Variables are false by default, so that is why they were not explicitly set to false. Current behavior is balloon windows are not shown. This has been the default for a while now, no need to change that.

Variables are false by default, so that is why they were not explicitly set to false. Current behavior is balloon windows are not shown. This has been the default for a while now, no need to change that.

For global variables defined in FvwmPager.c, yes. But I have been searching through the code where Balloon struct is zeroed, and have not succeeded. Can You point me to the place?

somiaj · 2025-05-18T17:30:12Z

modules/FvwmPager/init_pager.c

-	if (!Balloon.Ffont)
+	if ((Balloon.show_in_icon || Balloon.show_in_pager) && !Balloon.Ffont)


This is pointless, Balloon.show_in_pager or Balloon.show_in_icon is not set, this code is never reached in line 484 above.

I agree, You are right.

ONykyf · 2025-05-18T17:33:13Z

I am able to fix that issue with a single line:

--- a/modules/FvwmPager/init_pager.c
+++ b/modules/FvwmPager/init_pager.c
@@ -1187,6 +1187,7 @@ void parse_options(void)
                           StrEquals(resource, "SmallFont"))
                {
                        FlocaleUnloadFont(dpy, Scr.winFfont);
+                       Scr.winFfont = NULL;
                        if (strncasecmp(next, "none", 4) != 0)
                                Scr.winFfont = FlocaleLoadFont(
                                        dpy, next, MyName);

Then it is necessary to set fonts at the beginning to the default, not when pager window is initialized (whether these fonts are necessary or not), as @ThomasAdam proposed.

Have You observed that setting "None" after a real font set, is ignored in other two cases as well? This is not fixed.

Prevent some weird FvwmPager crashes and balloons not showing up because fonts and boolean flags are not initialized an start. Also the functionallity of "Balloons" option without parameters is restored. FvwmPager: Move default fonts initialization to init_pager_window() function Correct a misprint and move fonts initialization to initialize_pager_size() Scan further for an optional argument to Balloons Balloon (null) is no more allowed Remove superfluous conditions and restore defaults

somiaj · 2025-05-18T18:13:03Z

The only crash or issue that I saw was with setting SmallFont None, setting Font None and even turning on DeskLabels afterwards had no issue. BalloonFont None is not and hasn't ever been valid. If you want to disable balloon windows use Balloons None.

I'm just trying to identify what issues are being fixed. Anyways, as I see it a single line will stop the crash with setting SmallFont None after it has already been set, and the manual page should be updated to clarify that Balloons must have an input and that Balloons None turns off showing the balloons.

Everything else seems to be wanting to change default behavior / add features / try to make things more unfiorm, but I don't see an issue outside of that crash. I agree that things aren't uniform, BalloonFont is treated differently than other Fonts, but this is an historical blunder I don't think we need to fix.

ONykyf · 2025-05-18T18:34:24Z

The only crash or issue that I saw was with setting SmallFont None, setting Font None and even turning on DeskLabels afterwards had no issue. BalloonFont None is not and hasn't ever been valid. If you want to disable balloon windows use Balloons None.

I'm just trying to identify what issues are being fixed. Anyways, as I see it a single line will stop the crash with setting SmallFont None after it has already been set, and the manual page should be updated to clarify that Balloons must have an input and that Balloons None turns off showing the balloons.

Everything else seems to be wanting to change default behavior / add features / try to make things more unfiorm, but I don't see an issue outside of that crash. I agree that things aren't uniform, BalloonFont is treated differently than other Fonts, but this is an historical blunder I don't think we need to fix.

Probably, as a mathematician, I prefer things to be uniform:) And I would not like a font to be loaded just to be removed. So what with initial values for Balloon.show_in_pager or Balloon.show_in_icon? Where do they become false ?

somiaj · 2025-05-18T18:40:30Z

Sure it would be nice if the fvwm code base was more uniform, welcome to 30 years of people doing adding features / patches / etc, also back in the last 90s and early 2000s when most this code was written, standards were far looser with code than they in modern times. Anyways I don't think this is going to be accepted as is, you are changing things that don't need to be changed, and the changes will probably lead to segfaults for other setups as was commented above. A lot more is going to have to be done to make sure that these changes are correctly adapted though the full code, and they aren't really achieving anything.

Are you having any other issue with FvwmPager not working due to explicitly setting variables to false initially? If so we can address that, but I have a small patch that address the actual issues you have pointed out.

I agree it would be great if the code was more uniform, but that would require a much larger change than just trying to piece in the pieces you are here.

somiaj · 2025-05-18T18:43:13Z

modules/FvwmPager/init_pager.c

-			if (strncasecmp(next, "none", 4) == 0) {
+			FlocaleUnloadFont(dpy, Scr.Ffont);
+			if (strncasecmp(next, "none", 4) != 0)
+				Scr.Ffont = FlocaleLoadFont(
+					dpy, next, MyName);
+			else {
+				Scr.Ffont = NULL;
 				use_desk_label = false;
 				use_monitor_label = false;
-			} else {
-				FlocaleUnloadFont(dpy, Scr.Ffont);
-				Scr.Ffont = FlocaleLoadFont(dpy, next, MyName);


This is the reason @ThomasAdam got a segfault and will cause other segfaults. Scr.Ffont is expected to be a valid font in the code, so that is why it was explicitly not unloaded with the None setting. You are unloading the font when it shouldn't be.

No, this is not possible. All references to Scr.Ffont are either guarded by use_desk_label or use_monitor_label, or are inside draw_desk_label() all calls of which are guarded by use_desk_label or use_monitor_label, hence this cannot cause a crash. I wrote previously why @ThomasAdam got a crash.

somiaj · 2025-05-18T19:01:06Z

Anyways, unless you can point out other actual issues (vs just want to unify things), we will probably just go with a simple patch to fix the crash you are experiencing vs try to refactor the code (and deal with changes due to the refactoring). I have added that fix and updated the manual page to clarify actual behavior of balloon windows in the js/pager-fix-1189 branch.

Thanks for the discussion and if you have any actual issues that need fix with a way we can reproduce them, please share.

ThomasAdam added the relates:module Issue is in module code label May 9, 2025

ThomasAdam added this to FVWM3 May 9, 2025

github-project-automation bot moved this to PRs in FVWM3 May 9, 2025

ThomasAdam added this to the 1.1.3 milestone May 9, 2025

ONykyf force-pushed the on/init-pager branch from e96ab5a to deae0b2 Compare May 11, 2025 13:46

somiaj reviewed May 16, 2025

View reviewed changes

ONykyf force-pushed the on/init-pager branch from deae0b2 to 78d6a86 Compare May 16, 2025 15:45

somiaj reviewed May 17, 2025

View reviewed changes

ONykyf force-pushed the on/init-pager branch from 78d6a86 to f82bd0a Compare May 18, 2025 08:58

ONykyf force-pushed the on/init-pager branch from f82bd0a to 27dd492 Compare May 18, 2025 13:47

somiaj reviewed May 18, 2025

View reviewed changes

ONykyf force-pushed the on/init-pager branch from 27dd492 to ec5c3de Compare May 18, 2025 18:11

somiaj reviewed May 18, 2025

View reviewed changes

ThomasAdam added the skip:changelog Issue/PR should skip CHANGELOG label May 19, 2025

ThomasAdam removed this from the 1.1.3 milestone May 19, 2025

ThomasAdam closed this May 19, 2025

github-project-automation bot moved this from PRs to Done in FVWM3 May 19, 2025

somiaj mentioned this pull request May 19, 2025

Fix crash in FvwmPager and clarify Balloons option. #1197

Merged

		if (use_window_label && !Scr.Ffont)
		Scr.winFfont = FlocaleLoadFont(dpy, NULL, MyName);

		if (t == NULL \|\| w == None \|\| Scr.winFfont == NULL \|\|
		if (t == NULL \|\| w == None \|\| !use_window_label \|\|

		if (!Balloon.Ffont)
		if ((Balloon.show_in_icon \|\| Balloon.show_in_pager) && !Balloon.Ffont)

Uh oh!

FvwmPager: correct fonts and balloons initialization #1189

FvwmPager: correct fonts and balloons initialization #1189

Uh oh!

Conversation

ONykyf commented May 3, 2025

Uh oh!

ThomasAdam commented May 9, 2025

Uh oh!

ONykyf commented May 11, 2025

Uh oh!

ThomasAdam commented May 11, 2025

Uh oh!

ThomasAdam commented May 15, 2025

Uh oh!

ONykyf commented May 16, 2025

Uh oh!

somiaj commented May 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ONykyf May 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ThomasAdam commented May 16, 2025

Uh oh!

ONykyf commented May 16, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

somiaj May 17, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ONykyf May 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ONykyf May 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

somiaj commented May 17, 2025

Uh oh!

ONykyf commented May 18, 2025

Uh oh!

ONykyf commented May 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ONykyf commented May 18, 2025

Uh oh!

somiaj commented May 18, 2025

Uh oh!

ONykyf commented May 18, 2025

Uh oh!

somiaj commented May 18, 2025

Uh oh!

ONykyf commented May 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

somiaj commented May 18, 2025

somiaj commented May 16, 2025 •

edited

Loading

ONykyf May 16, 2025 •

edited

Loading

ONykyf commented May 16, 2025 •

edited

Loading

somiaj May 17, 2025 •

edited

Loading

ONykyf May 18, 2025 •

edited

Loading

ONykyf May 18, 2025 •

edited

Loading

ONykyf commented May 18, 2025 •

edited

Loading

ONykyf commented May 18, 2025 •

edited

Loading

ONykyf May 18, 2025 •

edited

Loading

ONykyf commented May 18, 2025 •

edited

Loading