is the font selection bug fixed

[bananacakes]

Just wondering if the bug that crashed version one of the applications is fixed. The one where it you picked a new font or typed one in the search box ( critical fix especially for a publishing program one would have thought ) ?

[GarryP]

Are you asking if the problem has been fixed for V1 or V2?

If V1, that’s something which Microsoft need to fix and make an OS update for.

If V2, it looks like it’s working fine for me and I've not read any further complaints about it.

[Stepaan]

in V2 I didn't encounter a single crash while selecting or searching fonts in the dropdown menu. On V1 it was an instant crash here  on my Windows computer. I don't own Mac, so I don't know.

[PaulEC]

It seems OK for me in v2. Walt also posted elsewhere that he didn't get the crash in v2, so that's four of us that now seem OK. Hopefully the problem is solved! (Until Windows screws things up again! 🙄)

[bananacakes]

I did mean if it was fixed or didn't apply to versions 2, and I have done the workaround for V1 with no ill effects that I can detect.

I understand that it was likely a .Net upgrade that caused this problem, but I'm sure Serif could have fixed it especially as V2 seems okay, along with all other programs that use the font search feature on my computer.

I've actually dug out an old hard drive with the last versions of the X line of Serif products, I'm going to install those see how they get on.

Thanks for your replies everyone  

[GarryP]

6 minutes ago, bananacakes said:

but I'm sure Serif could have fixed it especially as V2 seems okay, along with all other programs that use the font search feature on my computer.

Apparently, from little bits and bobs I’ve read here and there in the forums (if I'm understanding things properly), the UI (at least) for Affinity 2 was a complete redesign which, by its nature, bypassed the WPF problems that arose with V1.

Any software that doesn’t use WPF, such as WinForm applications, UWP, etc., were/are probably ‘immune’ to the WPF issues that appeared.

[lacerto]

3 hours ago, GarryP said:

Apparently, from little bits and bobs I’ve read here and there in the forums (if I'm understanding things properly), the UI (at least) for Affinity 2 was a complete redesign which, by its nature, bypassed the WPF problems that arose with V1.

Based on xaml files included in the distribution v2 apps continue to use WPF. The issue that is still unfixed in 1.x versions (and might remain unfixed indeterminately) was fixed in version 2 either by applying a workaround avoiding the data virtualization feature that triggers the error, or by creating a proper fix by making changes in the code. That means that Serif knows how to fix the error but they have stated that there will be no updates to v1 apps, and that fixing is responsibility of Microsoft who should create an update that restores the former functionality of these modules. Whether the problem will be fixed by itself after a future Microsoft update remains to be seen, but we do not even know if Microsoft has officially acknowledged the issue (that is, that there is a defect in the .NET modules) and promised a fix: logging a reported issue does not mean that an issue is acknowledged or that there will ever be a fix; we all know that from Affinity apps. This condition has now lasted about 8 weeks. 

What we know is that so far multiple updated versions to .NET 4.8 modules have been released both for Windows 10 and 11 after the September update that initially triggered the error, but none of the release notes of these updates mention fixes made to the modules because of problems they have caused to apps that depend on them. All release notes that I have seen have mentioned that there are "no known issues" that installation of these modules cause, and they continue to be recommended and included in by default automatically applied updates. The users choosing to block the modules should understand that these updates contain also security fixes.

[walt.farrell]

14 minutes ago, lacerto said:

or by creating a proper fix by making changes in the code.

 

15 minutes ago, lacerto said:

Whether the problem will be fixed by itself after a future Microsoft update remains to be seen, but we do not even know if Microsoft has officially acknowledged the issue

Serif have said that they reported it, and that Microsoft has acnowledged the bug.. We will not know whether Microsoft will fix it, of course, until they do.

Serif has clearly done something in V2 that at least works around the problem, but we don't know if there are any negative consequences (e.g., to performance) from what they had to do. I still believe it's a Microsoft issue, which Microsoft should fix.

[bananacakes]

3 hours ago, lacerto said:

Based on xaml files included in the distribution v2 apps continue to use WPF. The issue that is still unfixed in 1.x versions (and might remain unfixed indeterminately) was fixed in version 2 either by applying a workaround avoiding the data virtualization feature that triggers the error, or by creating a proper fix by making changes in the code. That means that Serif knows how to fix the error but they have stated that there will be no updates to v1 apps, and that fixing is responsibility of Microsoft who should create an update that restores the former functionality of these modules. Whether the problem will be fixed by itself after a future Microsoft update remains to be seen, but we do not even know if Microsoft has officially acknowledged the issue (that is, that there is a defect in the .NET modules) and promised a fix: logging a reported issue does not mean that an issue is acknowledged or that there will ever be a fix; we all know that from Affinity apps. This condition has now lasted about 8 weeks. 

What we know is that so far multiple updated versions to .NET 4.8 modules have been released both for Windows 10 and 11 after the September update that initially triggered the error, but none of the release notes of these updates mention fixes made to the modules because of problems they have caused to apps that depend on them. All release notes that I have seen have mentioned that there are "no known issues" that installation of these modules cause, and they continue to be recommended and included in by default automatically applied updates. The users choosing to block the modules should understand that these updates contain also security fixes.

Thanks for the workaround by the way, it does seem to work, and I've not noticed any ill effects. I re-installed Drawplus X8, PagePlus X9 I forgot just how clean they looked and seem rock solid.

 

[Jamesx1]

Well, I'm stumped. I tried really hard but can't think of a more basic function than font selection. Coming back to Designer to do a slew of font work and V1 is unusable. Buck passed? Guess I'll need to upgrade to V2 then.

Is V2 better than V1, overall, anyone? At least I can type a font name without a complete crash...?

[N.P.M.]

11 minutes ago, Jamesx1 said:

Well, I'm stumped. I tried really hard but can't think of a more basic function than font selection. Coming back to Designer to do a slew of font work and V1 is unusable. Buck passed? Guess I'll need to upgrade to V2 then.

Is V2 better than V1, overall, anyone? At least I can type a font name without a complete crash...?

You can still tryout the options in the faq found here
Or wait till Microsoft brings out the update that tackles the issue.
As for V2 it doesn't have this issue and has some new features you might want consider.

[Stepaan]

25 minutes ago, Jamesx1 said:

Well, I'm stumped. I tried really hard but can't think of a more basic function than font selection. Coming back to Designer to do a slew of font work and V1 is unusable. Buck passed? Guess I'll need to upgrade to V2 then.

Is V2 better than V1, overall, anyone? At least I can type a font name without a complete crash...?

IMHO V2 is a significant leap forward. And yes, as I said, you can type a font name without the instant crash.

[bananacakes]

1 hour ago, Stepaan said:

IMHO V2 is a significant leap forward. And yes, as I said, you can type a font name without the instant crash.

In what way a significant leap forward? 

[nwhit]

On the Serif website, there is a massive amount of information regarding the changes and improvements, including short videos that depict those new features/changes. MUCH easier than someone trying to type out that much info.

[Jamesx1]

OK, thanks for the info. Glad they are able to work around it (V2).

Just deciding whether to uninstall the W10 updates or upgrade Affinity. I too am interested to know how it is a significant leap forward. Publisher needs a better table system, and backwards compatibility is not there either. Not wanting to complain, but I just have no idea at which point in time certain things will be improved.

Anyone know when the 40% off ends?

[kenmcd]

Regarding v1 updates, and 40%-Off offer period - see:

 

[lacerto]

9 hours ago, walt.farrell said:

I still believe it's a Microsoft issue, which Microsoft should fix.

The problem might have been solved by a workaround, a code change or even (last-minute) change in control technology, but the problem did bug also v2 apps and it was fixed, one way or another. We also know how it can be "fixed" (or rather worked around) in version 1 apps, and that Serif could do this work around as a favor to their 1.x Windows users, but as it is, this has now totally been left to users who are kinds of captives in a situation where the only crucial question seems to be, "whose fault"...

Personally I would choose to pay more to get this fixed (the deal of course gets you much more than just getting rid of this nuisance), but if version 2 is as "redesigned" as has been advertized, there might also be other issues lurking so in a way users of a fresh upgrade also become kinds of testers. And if they make the jump and continue editing and saving their jobs with version 2 apps without making backup copies, they should know that there is also no going back to version 1 and continue editing documents again with version 1 apps. For some reason it does not seem that version 2 release apps offer users a choice of opening a copy of the original document to avoid this problem (or warn whenever a version overwrite is attempted), or perhaps I have inadvertently chosen to turn off this kind of a warning when version 1 documents are opened (it does not however seem to be an option of e.g. inbuilt "Assistant" feature). 

[lacerto]

3 hours ago, kenmcd said:

Regarding v1 updates, and 40%-Off offer period - see:

Thanks! There was contradictory or misleading official information earlier and having this further information was good and important to know...

[bananacakes]

Anyone know is there is any news on a proper fix yet ?  There seems to have been a fix for V2 for a while and i can only imagine it's the same for V1

[GarryP]

8 hours ago, bananacakes said:

There seems to have been a fix for V2 for a while

As mentioned above, the UI (at least) for V2 was re-written differently and thus it ‘bypassed’ the problem caused by the Microsoft-initiated error which is seen in V1.

It wasn’t ‘fixed’ for V2, it’s just that V2 didn’t use the things that were broken by the Microsoft update.

As mentioned in the FAQ linked to in a previous post, Serif are waiting for Microsoft to issue a fix before Serif can do anything about it, if they need to do anything about it.

Even if Serif did ‘fix’ the problem with V1 and provide an update before Microsoft fixes the problem at their end, who’s to say that any fix provided by Microsoft wouldn’t undo the ‘fix’ done by Serif? In that case, Serif would then need to figure out how to fix it again and provide yet another update.

If there is any news in this area I feel sure that Serif will tell us when they have it as they are very much aware of the problem.

[lacerto]

What is stated above is misinformation [EDIT: does not accurately describe what is known about v2 app development history] and is based on a strange and naive concept on how framework updates (containing also security fixes), which often are also made to older versions of framework modules running on older operating systems, and which can and often do result in apps depending on them getting broken, should be handled. These kinds of "side effects" are not rare, but instead of raising hands and doing nothing, and arguing on who's to blame and waiting for some future operating system update to restore former framework functionality, responsible software companies do what they can to make their software still within the support lifecycle fixed and updated so that users can continue their work without needing to sacrifice anything of the system security, or needing to invent miscellaneous self-made workarounds.

This situation has now lasted over 9 weeks, during which time the modules in question have already been updated multiple times with different KB numbers, and there may also be other automatic system updates that are dependent on these updated modules, so what exactly needs to be uninstalled is not clear.

In situations an app is no longer supported, the version that is still supported will be updated, if needed, and users are advised to update or upgrade their apps (whether free or at some cost), and in case of subscription software, an update will normally become available automatically. In this case it is possible that an older version will never be updated. This is what it initially seemed would be the situation also with Affinity v1 apps, but later it was officially announced that version 1 apps continue to be supported ("certainly for the foreseeable future") and they will get patches that make them compatible and useable e.g. with macOS Ventura and any system updates made to the Windows operating system.

They have also officially stated that they will soon release v1 version apps where this specific problem is fixed. We do not know how it is fixed, and this is not important, but Serif knows how to fix it and such fix should be imminent.

[EDIT: The prolonged release of the implied update may also be explained by inside information about possible forthcoming update to the .NET modules; that remains to be seen, but meanwhile, the app setting workaround is available for those who wish to restore v1 app functionality without sacrificing system security and feature and quality improvements made to Windows operating system. UPDATE: Nov 29 Tuesday patch preview did not resolve the problem.]

UPDATE: There is now official 1.10.6 update for Windows that properly fixes this issue. Please use that instead of the workaround. The updated version will overwrite any manual changes made to the app setting file so you do not need to restore anything to get the apps behave as planned.