fde101

For v1: Throw it in the trash. Empty the trash. For v2: also need to track down and remove the stuff that is installed from your account, if anything, unless shared with the other Affinity apps and those are being kept - not sure offhand where that is stored...

Many people generate PDFs to distribute over the internet. The larger the PDF is, the more bandwidth is consumed by each user who downloads it, each time they download it, and the longer it will take to download - making users wait unnecessarily long does not help to keep them happy. Many companies hosting files for people to download also pay for bandwidth, meaning that smaller files are cheaper to offer.

Just make sure we keep the v2 requests for this in one thread. Things get more confusing when they wind up scrambled all over the forum.

Serif had actually suggested that we should for requests which are still relevant. That was way back when v2 had just been released. Thankfully they don't make us "shell" out any additional money for posting our requests...

Because Serif made the same questionable choice in the opposite direction for that tool. Instead of ignoring Cycle Selection Box and forcing page orientation as Transform Objects Separately does, the Make Same commands instead ignore Cycle Selection Box and act on base box orientation instead. This does not appear to be a bug, and is not related to the beta. Both behaviors are valid, and are existing behaviors in 1.4, but the opposite behaviors are questionably missing from each of the two tools.

It does not align at all. When multiple objects are selected they are scaled based on their orientation to the page, regardless of the orientation of their base box. In other words, it scales their actual width on the page, not their width relative to their rotated coordinate system. Even in 1.4 it works this way, not related to using the "=" feature. It would be nice to be able to toggle this behavior as well when using Transform Objects Separately (we already can when working with single selected objects using Cycle Bounding Box so not sure why it should be different with multiple objects selected), but I believe that is a separate request.

They are not simplifying it artificially. They are ignoring nodes that are already flagged to be ignored by the file they are reading the shape from. To simplify an arbitrary existing shape they would need an algorithm to identify which nodes to remove and/or adjust, which is not part of what they are doing here.

These two goals are basically opposites of each other. When you draw with the pencil tool you are providing a continuous set of points with no real stops to indicate which ones are "important" to you, leaving the software to guess. How would it know which ones to straighten into a line segment and where it should make a turn to start a new one?

@GarryP that is not what @A_B_C is asking for - I understood the request quite clearly, and it is an obvious thing to have in place, which I definitely believe should have been there from the beginning. The request is not to open or close the entire panel, but the individual sections within the panel, all at once. In other words, clicking the triangle to collapse or expand "Ligatures", "Figure Position", "Capitals", etc. - but instead of just one of them, when a modifier is held down, collapse or expand all of them at the same time. The idea is that if they are all expanded, you could collapse all of them at once, then expand individually the one you are interested in, without having to go hunt through the expanded sections that might have the one you are looking for scrolled out of view.

No need for algorithms for a simple 1D barcode. Just use a font.

This is almost as bad as RED getting a patent on lossy-compressed RAW encodings that happen to be above a certain resolution. Making it specific to high resolutions suddenly gets past prior art of doing exactly the same thing at lower resolutions, which had been around since before RED existed? Patenting a specific algorithm for doing that - maybe, but stretching it. Patenting it as a concept? Inexcusable. There is plenty of software which moved (or duplicated) sliders and numeric fields onto the canvas in the form of handles, so creating on-canvas handles for yet another set of sliders is suddenly inventive enough to be patentable? The patent offices may be stupid enough to let something like that slide, but they certainly shouldn't be. Direct manipulation of just about anything should be an obvious end goal by now, hardly inventive enough to be considered for a patent, even if no one else has tried it yet.

https://apps.apple.com/us/app/iqr-codes-qr-code-art-studio/id491390884?mt=12

What would be really interesting would be to also find a way to represent at least some of the common ones as handles on the letters (when zoomed in far enough!!!) the way that handles are used for rounding rectangles and shaping gears and the like... The Typography panel would be a logical place for these otherwise, in my way of thinking.

Accordingly, shouldn't that be in the Feedback forum as it is not for this beta? In any case, since these fonts are constructed by center-line strokes, when this is done, please consider leveraging that fact by allowing brushes to be applied to them (for those not doing CNC work, anyway...).

I think what is being suggested is that the popup box where the URI is entered have a few tabs to build some of the common schemes for the user so that the user doesn't need to work out the encoding for all of them. For example, iQR (on the App Store for the Mac), in addition to the URL option (which they label "WWW"), has tabs for Phone (enter a phone number), SMS (phone number and message), eMail (to address, subject and body), Map (enter an address and it looks it up using Google Maps or some other map service), Contact (first/last name, phone number(s), email address, web page, street address), Calendar (event name, when, where, etc.), PayPal/Bitcoin/SEPA payments (account, description, price, etc.), WiFi (network name/SSID, password, encryption type, hidden?), and Text. It also provides an interesting range of formatting options to spice up the appearance of the generated codes: Granted that this range of formatting options is probably a bit out of scope for a program that is not specialized in producing QR codes, but still thought I would point it out. The more important thing to consider is to throw in the forms for a few of the common URI schemes.

An important step, but don't discount that color font support is also important. If not in parallel to variable fonts, please make sure these are not neglected either!

This seems to be in the wrong section of the forum?

In addition to these excellent points, I would also point out that if anyone is working in an area with poor or no internet connectivity, there is no reason to rob them of access to online help simply to add rich video content that is probably not as important as having a quick reference to some less familiar feature they are trying to use. Some users may also work in an environment where they need to be disconnected from the internet for security reasons and losing access to online help would similarly mean losing access to any kind of help. I agree that a change to online-only help is a TERRIBLE idea. PLEASE reconsider.

I would imagine some users would also find it nice if more of the alignment functionality were available broken out into individual icons so that they could be made more accessible without the extra click to open up the alignment panel. A fair subset of them are admittedly available already on the context toolbar, but only with specific tools selected. Different users place different priorities on different tools and functions and have differing amounts of screen space available, so making groups of commands available in popups and individual commands available to reduce clicks would allow users to better tailor the toolbar to their own requirements.

I think you've stumbled onto the difference between a preset and a style. A preset gets copied into the thing that uses it so the link between them is not maintained - if you apply a preset to a cross-reference, that cross-reference takes on a copy of the preset when it is created; it is not the same as a genuine style which maintains a link so that multiple instances can be updated in one place. Sadly, the Affinity products blur the line a bit in that they have object "Styles" (via the Style panel) which are not styles at all, but are in fact presets, exhibiting the same sort of behavior - they are copied to the objects and no link back to them is kept. In my opinion, presets in general are inappropriate for most of these things (including both of these), and I would have much preferred that Serif had implemented proper styles instead (particularly object styles, as QuarkXPress offers for example), but sadly that is not what they have done here.

This is 100% extraneous and would be foolish to waste toolbar space on. Both supported platforms have integrated notification systems at the OS level. It would be particularly dumb to reinvent the wheel on this. If Serif did want to switch to using a notification-based update mechanism, they should use the notification system already provided by the underlying OS, which can then behave like every other app which uses the OS-provided notification system. This is already available on both platforms. On the Mac it is called the App Store. On Windows it is called the Microsoft Store. It even provides consistent behavior for software across multiple vendors - including Serif.

If you start with a rectangle or rounded rectangle and convert it to a text frame instead of creating it as a text frame to begin with, you can access the corner options using the node tool (or the corner tool if you also have Designer). I still don't understand why they don't just create the text frames that way when you create them directly, as it seems kind of pointless to treat "normal" rectangular text frames as being so much more limited than converted ones. It is technically possible to get different stroke widths by adjusting the pressure curve, but it is very annoying to do so that way. Better tools for this have been requested numerous times.

Then it would be another related feature request... ideally with its own thread, like each feature request should have. The notion of a drafting persona kind of makes sense if you are trying to use Affinity Designer as a poor man's CAD application (in spite of there being *real* CAD applications which are even free and open source). When working with a CAD program there is a difference in how you interact with tools to accomplish the same things as you would do with the tools we already have. For example, in a typical drawing program, you create a circle either by dragging out from the center, or by dragging from one corner of its bounding box to another. In a CAD application, you might also create a circle by specifying two or three arbitrary points on the circle, or create a line by telling the application to run parallel or perpendicular to an existing line or shape and give it a point along the line. Adding stuff like this would not be unreasonable in and of itself, but it may create some expectation that other aspects of a CAD application will be covered, and I'm not sure Serif really wants to get into that market segment on top of everything else they are doing. I would not be opposed to it, but I don't know that I can really support it either unless Serif comes out and indicates this is something they want to involve themselves with and that they are going to take precision more seriously (among other requirements for proper CAD). In my opinion, if you want software that works like a CAD application, you should use a CAD application, which Affinity Designer is not.

There is technically a way to get this working now, but it is a tedious process made more tedious by a few seemingly arbitrary limitations (some of which are probably bugs). Make sure the Links panel is open. Within the group that you want to use as the mask, select each object (not the group itself, sadly) and choose Duplicate Linked from the Layer menu. Select one of the objects (the original or the clone) and drag the other one from the Layers panel onto the Transform box in the Links panel to link the transform (as the transform is not linked by default when using Duplicate Linked - everything else is). Drag the duplicate layers out of the original group and group them separately. Select one of the groups (either the original or the one containing the clones) and drag the other group from the Layers panel onto the Group box of the Links panel (yes, you can link groups separately, so why can't you just use Duplicate Linked in the first place? Silly arbitrary limitation or a bug?) Drag the duplicate group's icon on the Layers panel onto the thumbnail of the Layer you want to use it as a mask for. Now for both the grouped objects and the groups themselves the clones used as masks will follow the originals due to the links - the one major catch is that if you add an object to the group, you will need to separately duplicate it, link it, and add the linked copy into the other group. One obvious bug that I discovered is that if you have two layers selected and choose Duplicate Linked, both of them are duplicated, but only one of them is linked instead of both. It seems that in order to make this practical, two hopefully simple changes would largely cover it: Allow Duplicate Linked to be used on groups, and duplicate the group along with its content, linking ALL of their properties for both the group and its content (including Transform). This can be done manually now using the process I described above but it is annoying, time-consuming and tedious. When using Insert Inside and creating or pasting an object inside of a linked group, paste a linked copy of the same object (including its transform if the transform is linked for the group) inside the group(s) that are linked to it as well.

I don't think anyone is going to argue that this is a sorely missing feature, but you might want to take a look at the new Layer States feature that was added to version 2.4 in the meantime. When combined with use of master pages this may give a reasonably usable work-around to this missing feature for some use cases. By no means is it ideal for all cases, but it should at least make some of them reasonable to deal with, and may even be better for a few. That said, I too look forward to an eventual implementation of the real thing.