Gear maker

Great job, well done.  Well, I guess not all credit goes to you.  Congratulations to you and your wife.

The documentation team have agreed this is a good idea and will hopefully be adding some sort of tutorial at some point which will cover safely backing up/exporting your settings etc. 
I'm not sure if it will be written or as a video or even both. I'll try and dig this thread out and let you know once it has been done. Cheers :)

First full day free to myself for a while and Ive been using this style in a lot of client work over that last couple of years, so thought Id see what happens if I'm not watching the clock and pack it with shading and detail, really quite please with the outcome. I'm going to add some characters to this next week if I get another free day.

Reseting an Affinity app to factory defaults is often suggested as a cure for various problems. While that may work, it is a very destructive step that wipes out all user customizations for keyboard shortcuts, the layout of the Tools palette & the main menu bar, fills, styles, brushes, etc. & returns just about everything to the new user defaults values.
 
Some of these things, for example keyboard shortcuts & Styles categories, can be exported/saved prior to the reset & imported/loaded back into the app afterwards, which can reduce the considerable amount of time & effort sometimes needed to return the app to the previous highly customized state.
 
However, there does not seem to be a comprehensive list of all the things that could be exported or saved for this purpose, or any mention of which (if any) should not be imported or loaded back into the app to reintroduce some problem the reset was meant to cure. For many of them, exporting/importing each category individually would be required since there is no global 'export/import all categories' feature in their menus, so even with a comprehensive checklist this can be a tedious & error prone process.
 
Any guidance on this would be appreciated, as well as if or when less destructive alternatives would suffice.
 
One example of this is I have noticed than on a Mac, some problems can be resolved by trashing the ~/Library/Containers/com.seriflabs.affinitydesigner/Data/Library/Caches files or the equivalent path for the other Affinity apps (including the betas & possibly for the Photos extensions as well). Note: for anyone wishing to try this, it should be done only while the app is not running.
 
In particular, this seems to resolve unusually long launch times & some other sluggish behaviors, but only when the contents of the Caches subfolders have become large, at least more than 1 or 2 MB.

Ha! Good old auto-correct... *sigh*

Alternatively just select the top node which u want to keep; then nudge it over using the arrow keys a set amount (2 clicks is usually enough). Then select and delete the lower node. Select the upper node u nudged, and nudge it back the same number of clicks as before but in the opposite direction.

Matt, I see what you mean.  I had a group selected that was collapsed.  Then I just changed the visibility of a layer 10 layers down from that selected and the selected group expanded.  That shoved the layer I had changed the visibility on way down from my cursor.  So when I clicked it again to toggle the visibility back it changed the visibility on another layer instead.
 
Mike

I think we're going to modify it so that it doesn't auto-expand if you select the top level of a group (because it interferes with selection in the Layers panel) but apart from that, it's actually quite a nice little tweak... :)

Hi team, just wondering if there has been any headway on the boolean tidying up issue? See attached.
 
It's weird because it seems inconsistent. It appears that the closer two paths are to each other the more nodes gets placed all along each path and sometimes even a third path is generated from the two intersecting/overlapping paths.
 
What makes this problematic is that after the boolean if you want to further manipulate the resulting shape, you are dealing with sometimes hundreds of nodes. For example here in the attached windshield/windscreen image I would have duplicated the darker "window seal" element to build a shadow area under the upper ridge. Because of all of the nodes now I'm forced to redraw a new element for that and try to get it into the same shape. Not to mention cleaning up and removing unwanted extra pieces...
 
Sorry to keep harping on this but to me this is very important stuff and I hope it's being addressed.  :)

Don’t let us get greedy … ;)
 
I am still trying my best to fathom the depths of symbols and constraints. These are immensely powerful tools that opened a world of new possibilities and new flexibility for quickly adjusting designs. And when you consider the parameters that have to be taken into account in order to make these tools work as they already do, I believe that was not quite a small achievement …  ;)

Ich versuch’ es ab und zu …  ;)

I have had more duplicate objects since this was changed than I have had in the 2 years before.  I'd rather not have the measurement tool than to have to press cmd after the beginning of the move.  Oh I sound all negative!  That's not the case, I'm still loving AD.

It isn't possible in AP and in AD to copy only layer effects, like a shadow effect, from a vector object to other vector forms, it overwrites also the color of the object. Would be nice to see a feature like in PS, where we can copy only effects to other layers. That's a really important function.

R C-R's instructions worked for me.  I get his exact results also.  First the open Recovery File and then the file locked.

We're trying to make the 1.5 features work reliably first, before adding yet more :)

Matt,
 
Thank you for addressing my problem with silly zooms.  So far it's been working even in to 1.3 million %.  I haven't had a need to go further, so far :D .  Thank you, thank you, thank you!
 
Mike

This may sound like an odd question but...
 
In AD the Gaussian Blur Radius what does the amount of pixels mean?  I'm looking for something a little more detailed than what the help file says "the extent of the effect."
 
On a pure white document I made 2 squares (500 pixels each).  One pure white and the other pure black.  Putting a 50 pixel Gaussian Blur on the black square I then moved the white square away until the edge was no longer discernible.  The distance between the edges of the squares was 125 pixels.
 
Gaussian Blur                    Distance of effect
50 pixels                            125 pixels
100 pixels                          250 pixels
200 pixels                          500 pixels
 
I then figured that maybe it was a point that had a 50% gray.  But...
 
At a 50 pixel blur setting and a point 50 pixels away from the black square the gray measured to be a 14 or 15% gray (218, 218, 218).  The same for 100 pixels from a 100 pixel blurred box and 200 pixels from a 200 pixel blurred box.
 
So what does this setting mean?  Sorry if this is a stupid question, but I sometimes want to know how far the effect reaches and now I know that it reaches 250% of the setting.
 

I hear you MEB, and I totally understand that the team is small and trying to put out fires while advancing the product. :)  but I'm inclined to somewhat agree with JimmyJack.
 
I guess what I'd like to express here is the critical importance of addressing some of the current users on-going core issues before racing to get version X out. I would have thought fixing this would get classified as a bug fix/improvement... I don't know how many would agree with me on this but I would much rather a solid, ironclad app that delivers everything as promised before moving over to adding features that may in itself potentially introduce more bugs or issues.  :mellow:  
 
I know a big part of creating and getting award winning products like these to market and keeping them on the consumers radar is the promise of new features and maintaining a certain level of excitement. This is fine and expected as long as some of the core features aren't short changed along the way in favour of introducing "whiz bang feature X".  :P
 
The last thing I want to do is dis the dev team, I know you are all working hard on 1.5  Hey, I'm your biggest fan and I want you to take over the world! I do!!  :D  I'm just also pining for those beautiful booleans and enchanting expand stroke features I read about on the box...  :rolleyes:

I'm really dismayed to hear this.
Personally I think all (okay, I'll say MOST) time and effort spent on "advancements" should halt until basic functionality such as this is rock solid.
 
(on a related note..... how's the pen tool coming along  :rolleyes: ?... It's a drawing program after all. Expand stroke?)

I would want to see 1.5 beta at last before demanding other stuff ;)

Hi andonbray,
Welcome to Affinity Forums :)
The best/correct way to solve this would be to use a lighter weight of the font. If that's not possible a quick n'dirty way to do it is to subtract an expanded stroke from the letters.
Here's a small clip showing how to do it (no sound).

A Transform each would be excellent. Similar to AI function, giving options in scale, rotation, skew, etc.

Glad you all like it.
 
A-B-C: Thar she blows!!  :D
 
R C-R: re: all of the above....
You aren't doing anything wrong, and yes I have seen this behavior. I believe a couple things are going on... neither of which are good. 
 
Extra nodes: Baking, I believe, is part and parcel with Expand Stroke.... which, sadly, we all know is very, um, troubled... (big problem #1). This is opposed to Convert to Curves command, which seems to be fine. Go figure. I think it might have something to do with the angle of the square when using the corner tool and or baking.... but I can't recreate reliably.
 
But it gets worse.....
 
Try making a perfect/level square. Do the corner tool maneuver. You end up with a perfect circle with four nodes right? Wrong. You get eight nodes.... pairs on top of each other. This actually kind of makes sense if you think about what the tool is doing. But at the point where the moving corners have reached their max shouldn't they become one? Maybe , maybe not.
 
But this brings us to big problem #2. Coincident points in AD. Okay, so what do you do? There is nothing.
Try deleting one.... the curve distorts. They're already on the same curve so neither of the "join" options do anything. So you might say... pull them apart add a node, break that node shift click select those, delete... grab an end point and drag snap to the other and that will join the curves (or any variation of all that). First of all that is the most insane workflow ever, and the result still distorts the curve.
This desperately needs to be addressed. 
I would prefer an automatic join if two points are dragged on top of each other... (but I can understand that this may not always be desired either. And that special circumstances could make that action a messy prospect..... a "T" junction for example). At the very least there needs to be an option to merge coincident points... by selection or globally.
 
Sorry to say, these two issues alone really make it hard for me to take Designer all that seriously (there's a big rant just under the surface here  :o, but for now I'll just stuff it back down....  ^_^).
 
Another (semi-related) TIP:
A GREAT way to get an individual circle with multiple (more than 4) equally spaced nodes is to use the polygon tool. Use "X" number of sides and push the curve setting out to 100%, expand. 
The same sort of thinking can be used to get perfect midpoints on other shapes too.... Use the star tool with no indent etc....
Experiment!
 
Cheers.

Hey all,
 
So sort of two different topics here right. 
1) place an (odd) number of circles evenly around in a circle.
2) evenly distribute objects along a path that's not so friendly. Could be a spiral, free-form curve.... whatever.
 
A more robust distribute toolset, which we obviously need (and I believe is in the works?), could of course address both issues in one tool.
 
But for now....
 
1) I have a different option other than CMD-J worth mentioning (imho). CMD J is very powerful, but for a large number of objects it can get a bit.... tedious. Even if you're CMD-Jing just a few elements, grouping, doing another transform, CMD-Jing, grouping etc....
This method takes about a minute (if I'm not talking trying to describe what I'm doing  ;)) .
And that's a minute if you're doing 9 objects or 999 objects.
Video (sorry about the clickety-clack... I come down hard on my keys).
 
2) That's a tougher one. Could be done.... but very labour intensive, and relying a little too much on eyeballing etc.
I would use the tape measure concept discussed in an earlier thread (can't find it at the moment) to place nodes and divide the path in order to snap objects to end points. Like I said, PITA.
EDIT (smacks forehead): Of course ;)  you can just use a copy of the curve as a Text path and use a bullet character (circle, square, or something more ornate) tabbed at even distances (or just use letter spacing). Then convert to curves. 
If you know how to make your own font it could be anything  :) .
 
 

Thank you Matt and team.  It seems to be working great.