Is there anyone from Adobe who can help me regarding ID Bug 424115 (ID hangs)?

[Gerben Wierda]

It has almost been a year (Sep-Oct 2023) since I ran into trouble updating my ID document. ID started to misbehave: it would go to 100% CPU with spinning disk cursor and it would never come out. Once I let it run for ages, and at some point it ran out of resources and crashed. From that moment on, I was trying to find out what was wrong, I tried a lot (older versions of ID, starting from the IDML, removing caches, different fonts, etc.), you name it, I tried it. Hanging happened at various moments, but at one point I found a reliable way to trigger it with a certain paragraph edit action at a specific point.

The document is a very large and complex document. It is about 240 Letter sized mostly two-column pages with 340 diagrams (sometimes inline, sometimes floating), index, references, etc., the works. I started the document in 2011-2012 and roughly once every three years I updated it (using the latest version of ID). You can find about it here: https://ea.rna.nl/mastering-archimate-edition-3-1/ (the last version produced in 2021, the first 50 pages are a free download)

I contacted Adobe. My document was checked by Adobe and judged to be 'not corrupted'. The issue was passed on to Adobe Engineering and they confirmed to me that it was an actual ID bug. It is apparently only triggered extremely rarely (only my document is a known trigger as far as I know), but it is a real bug (of course, going into some resource-wasting endless loop is a bug by definition, but you can allow for corrupted data to trigger such behaviour — my document is not corrupted though). It had the case number E-001071727.

Since then, I have been in the situation where I have been told various times that the bug would be fixed in an upcoming version, the first time on 1 December 2023 that "This issue will be planned to be fixed in version 19.2 of InDesign". That became 19.3 or 19.4. Every time I did not hear anything, except for instance when I asked, I was asked the same things by the same person (like: "do you have a small document that triggers it?" for which the answer is obviously no). The 'funniest' was on 6 Aug 2024 when I was told: "This issue will be resolved in the 19.5 version of InDesign, but its ETA is not fixed yet, so once that is informed, I will let you know.". I replied: "19.5 is out since 12 July 2024" (almost a month before). They replied: "This has been moved to be fixed in InDesign version 20.0, which will be released at the end of August or the first week of September" (that reply also contained the same thing from May: "Engineering has asked for the below detail- Can you help with a smaller document to look into this issue?"...). This made it clear that (a) engineering did not really look deep into it and thus (b) 'to be fixed in 20.0 which arrives late August/begin September' cannot be true.

I seem to be communicating with someone who is telling my what I like to hear, not necessarily what is true. It may also be that they are not correctly informed by the actual engineers.  The last message said "As per the engineering team this issue is now with the PDF team as the issue is also with the annotations in PDF on which they need sometime more to work. Please allow me another few days to come back to you with their findings.". That was 5 days ago. This last thing is so specific that it might mean they have actually looked into it, but by now, I have trouble believing anything that I am told.

In the meantime, I have let my CC subscription expire (no use paying for it if this is why I have it in the first place).

So, after 10 months, this message is a desperate attempt to get traction (and truth). There is a deep and nasty bug which can hit anyone (as it is a bug with unknown triggers) but might be so rare that it almost never does. But if it does, you are done for. End of your document, start fully over (and potentially run into it again, so a lot of work with a risk it won't do you any good).

[Robert at ID-Tasker]

If you are the only one who is experiencing this bug... well, I wouldn't hold my breath... 

Can you describe in detail this bug? What are the circumstances of it happening? Maybe we can find workaround to avoid it? 

[Gerben Wierda]

The bug is triggered in multiple ways. I've had it happen when I moved frames (with images and captions in them). The reliable way for me now is (was, I do not have a license anymore and I won't until this has been fixed) when I remove the 'start at next column/page' on a specific paragraph. The rearranging that happens after that seems to trigger an endless loop (1 core to 100%).

I've asked engineering that a workaround ("don't do this") is OK too. But no workaround has been forthcoming, and of course, a reliable workaround must come from knowing what the bug is or what triggers it and we do not know.

The last message (about this being the case in PDF as well) seems to suggest they know now what is going on, but I do not know if what I am being told is factual (so far it hasn't been).

My post says "you name it, I tried it" and really that is the case.

[Robert at ID-Tasker]

@Gerben Wierda

Any chance you can share your document - with links? 

It might have been something as simple as bad font - or corrupted image. 

Maybe, as you've suggested, they didn't really have time to properly investigate it... 

[Gerben Wierda]

No can't do without NDA. This is source material that is not open source.

Note: Adobe has already confirmed to me it is an ID bug and that my material is not corrupted. I shared everything, IDML, images, fonts, etc.

What I need is action from the people who can fix this. That is the only thing.

[Robert at ID-Tasker]

@Gerben Wierda

I don't mind signing NDA.

But if it's really somehow linked to the PDF export - and if you were not able to edit text - I would really suspect a bad font. Or corrupted Anchored / InLine graphic. 

[Gerben Wierda]

I don't know if it is linked to PDF export. Might be, but it hung when doing simple editing, not exporting. But my images are almost all vector PDF and it could be that my PDF vector images trigger this. So that it is a problem in the 'PDF' side (handling of PDF inside ID). This is corroborated by something I noticed when trying to import in Quark with a trial license of that tool. There were some strange things happening there too with images and especially coordinates (PDF being displayed outside the frame it was supposed to be in), which I (logically) ascribed to Quark's IDML import being iffy. So, a PDF issue (in which case I would suspect technically correct PDF but a form that is very rare) could be part of the problem. However, it is notable that this would then only appear now, when I have been using these images since 2020/2021. Possible (there are scenarios), but strange.

Thanks for your offer, it is very kind. I may take you up on that (or buy another license, just for testing) if the engineering people do not get back to me in the coming weeks.

[Gerben Wierda]

I would be happy with any help, so yes, please.

I've reactivated my CC account in case accidentally 19.5 would fix this. Alas. I sarted trying to slim down the document until the bug disappears. That turns out to be complicated. At one point I thought the problem disappeared when I removed the last section (spreads with content), but that was a false positive. It seems there is some background thread running that is doing things (updating numbering maybe?) that in a later stage returns the document to a state that the bug can be triggered again. E.g.

• Remove stuff, test bug immediately: no more bug
• Remove stuff, wait a few minutes (say 10): test bug: it hangs

I tend to save/open via IDML duing testing to minimise the chance of passing the bug on via the .indd file.

[Gerben Wierda]

Oh, by the way. I have been changing fonts, so that was my first suspicion as well, but Adobe has confirmed me that this wan't the problem. Note that when I use a professional licensed Linotype font which I purchased, the same thing happens. So, as far as I'm concerned, font trouble is an unlikely candidate.

[Peter Spier]

If engineering wants/needs a small document, have you tried triggering the bug with a file made from just part of the larger file?

[Gerben Wierda]

Hi Peter,

Thanks for the suggestion (actually you may remember that you are mentioned with thanks in the book...)

Engineering has everything. They have all my stuff. They can easily open the file and try to save it with lots removed. They do not need me for that.

And as it is now, I don't have a (currently useless) subscription anymore. So I can't anyway.

If they had looked into it and suggested this 9 months ago, I might have tried. 

Frankly, doing this with the existing document seems unnecessary. What keeps them from hooking up a debugger to see what is going wrong?  (assuming the compiler optimiser is not too blame).

Yours, 

Gerben

[Gerben Wierda]

I might add that the document was so unreliable that many things I did (even typing something) triggered the bug. So it is likely that I would not have been able to create a smaller version.

[Peter Spier]

Here's my take on this as of now...

Your bug is so rare nobody else has reported it so it is probably low priority, so if you want them to fix it YOU need to do as much for them as possible. If that means making a smaller file, even if they could do it themselves or debug the one you sent, you should try to do that. If that means now that you have to ask someone else to extract pages for you, I could do that or I suspect @Robert at ID-Tasker would also be willing (and maybe faster than me -- I'm currently travelling). You should also provide with the file a description of the steps you have confirmed will trigger the bug in the past so it's possible to verify it still exists in the smaller file.

[Gerben Wierda]

It is easy to overlook how long this has been going on and the information I received. When I still had an active CC license I repeatedly offered to do anything they wanted me to, indeed: as much as possible. I was wel aware that 'I need to do as much for. them as possible'. They were telling me they were looking into it, but that apparently wasn't really true. They never asked me to produce a slimmed down original document, I would have gladly tried (though most likely it would have hung as soon as I tried). And they repeatedly told me "it will be fixed in 19.x, 20.x".  It seems nobody cares not so much that they did not do anything, but that they apparently told me things that weren't factual. And do not forget, Adobe themselves have told me there is no corruption in my data (of course, given the experience so far, one cannot be sure that this is factual, so an extra check could be in order).

[Laubender]

Hi @Gerben Wierda ,

the one thing you could still try is signing up for InDesign Prerelease so you could test with the upcoming InDesign 20.0 version. This will be under NDA, of course. So do not comment here, but discuss the issue in the InDesign Prerelease forum.

https://blog.developer.adobe.com/how-to-join-creative-cloud-prerelease-programs-ccecc10ae7d3

Regards,
Uwe Laubender
( Adobe Community Expert )

[Gerben Wierda]

So, after a year of being lied to by my Adobe contact at a hunch, I though to check the document version for which I was convinced that Ophian font was gone. Turned out, it was still there (in a table of contents that was not updated, and only a few words had this character format, very weird). So, I totally removedOphian an ID stopped completely hanging. Yay!

But th ereplacement ftonts I tried were (variable) OTF and I still had trouble. For instance, setting a variable on one font made ID for no obvious reason decide to try to set it too on a completely different font, which did not have that variable (or any variable at all). I kept struggling. I tried commercial OTf/variables, nothing worked properly.

Finally, on a hunch, I removed all OTF fonts and only used TTF, and nothing variable. Suddenly, not only do I no longer have problems and not only that, suddenly my ID is very responsive when editing.

Conclusion for me: stay away from OTF, and certainly from variable fonts. 

And don't believe a word Adobe engineering is telling you.

[Peter Spier]

I would bet the problem is with the variable fonts rather than .otf as a class -- we've been using .otf for years without issues.

On the other hand variable fonts are fairly new and support for them in InDesign is even newer, and anecdotaly not ready for prime time.

[Gerben Wierda]

Except that Ophian isn't variable and my issues started when I had no variable font in use whatsoever. That only started when I was looking for replacements.

[rob day]

Conclusion for me: stay away from OTF

Hi @Gerben Wierda , The problem can’t be the OTF format—OTF is now a standard font format, and has replaced the Type1 Postscript font standard, which is no longer supported by Adobe or Apple. All of the fonts available via AdobeType with a CC subscription are OTF.

It’s more likely a problem with the specific font. Ophian looks like a rip-off off Linotype’s Optima, and its vendor FontSpring has a tryout version. Is your version purchased or is it the tryout version?

[Dave Creamer of IDEAS]

It looks like FontSpring adds protection to their fonts. I wonder if that is triggering ID somehow?

Also, is the bug platform specific? (If the OP mentioned it, I must have missed it...)

[Robert at ID-Tasker]

Also, is the bug platform specific? (If the OP mentioned it, I must have missed it...)

By @Dave Creamer of IDEAS

OP works on a Mac but sent me files - same problem on Windows - temporary file starts getting bigger and bigger and InDesign is in semi locked state - you can kind of "click around" but it's overloaded.

I never let it fill my drive but it's pretty much inevitable. 

[rob day]

Does it happen if you hide the fonts and force a font replacement on open?

[Robert at ID-Tasker]

Does it happen if you hide the fonts and force a font replacement on open?

By @rob day

The problem is only with this specific font. 

If I don't substitute - or substitute to Myriad / Minion - everything is perfectly fine. 

[Gerben Wierda]

Thanks all.

I puchased the font (like I purchased 4 typefaces of Monotype's OptimaNova before but came belatedly to the conclusion that its license doesn't allow the use I need, so I was looking for an alternative). I am willing to pay for a font, but the amounts that Monotype charges is more than I ever make on this hobby. This is legally true for almost all fonts that are shipped to people, including the standard fonts on computers (macOS, Windows), they come with severely restricted licenses which are generally ignored by most people, but I tend to pay attention, even if the license holders seem not to enforce their license for those widely seen uses.

I now use the static version of Commissioner, which has variable version on Google Fonts which gave me a lot of other problems in InDesign, even some crashes. On github there is a static version of  the max-flaired version "Commissioner Flair" in TTF. Since I moved to that my document has become very responsive, whereas when I was using non-Ophian OTF (variable or not) it wasn't (apart from the problems I encountered with variable changes). From my anecdotal experience it is wise for me to stay away from variable (especially if you actually change the variable axes), and from OTF.

[Peter Spier]

I'm curious what font restrictions you've encountered that prevent you from using fonts like Optima Nova.

I've rarely seen a license on a commerial font that prevented anything other than distribution of the font code itself by posting it on a server. Embedding the fonts in a PDF for distribution is generally not an issue.