Description: I have a a document that, when i open it with libreoffice (up to latest 7.x Version on windows and linux --headless) will force libreoffice to generate an endless amount of pages. I will attach the document to reproduce the bug every time it is opened. I tried to track the problem down to its core and i guess it is caused by the lines that are made using underscore "_" chars. When i remove all the underscore lines, the document opens. As we use libreoffice in headless mode on a web-service this is a major problem, because the service will hand without notice or error. I / we need help ... Steps to Reproduce: 1. Open the document Actual Results: Using Windows you see the page counter at bottom left increasing with no end. On Linux (headless) it just get stuck. Expected Results: application is stuck until OOM Reproducible: Always User Profile Reset: No Additional Info: open the file or error out
Created attachment 164253 [details] will crash libreoffice on open
Reproduced in Version: 7.1.0.0.alpha0+ Build ID: df37937018fe8e87dad5dd97689632044ba56de3 CPU threads: 4; OS: Linux 4.19; UI render: default; VCL: gtk3 Locale: en-US (en_US.UTF-8); UI: en-US Calc: threaded The document only has 1 page
also reproduced in Version: 5.2.0.0.alpha0+ Build ID: 3ca42d8d51174010d5e8a32b96e9b4c0b3730a53 Threads 4; Ver: 4.19; Render: default;
Also reproduced in Version 4.1.0.0.alpha0+ (Build ID: efca6f15609322f62a35619619a6d5fe5c9bd5a)
@Justin, I thought you might be interested in this issue
Created attachment 164267 [details] Example file (no-looping) 1. Open the attached file (doesn't loop by default) 2. Right click the frame at the right 3. Anchor to paragraph/to character and looping starts
Format -> Page Style - Header tab -> Uncheck Same content on first page and looping goes away
this bug is not about changing styles to make it loop but to prevent the loop / death of application that runs forever and never quit with an error (except OOM). Using LibreOffice with this flaw can lead any service in headless mode to get stuck while in a GUI driven environment you could possibly kill the task. If i would use the libreoffice binary directly for each call i could trap it with a timeout (bad idea) and send a kill to that pid. But it is running as a service with --headless --invisible --nocrashreport --nodefault --nologo --nofirststartwizard --norestore --accept=socket,host=127.0.0.1,port=2002,tcpNoDelay=1;urp;StarOffice.ComponentContext as soon as the service will process the attached file it is stuck and blocks itself.
(In reply to Telesto from comment #6) > Created attachment 164267 [details] > Example file (no-looping) > > 1. Open the attached file (doesn't loop by default) > 2. Right click the frame at the right > 3. Anchor to paragraph/to character and looping starts very well analysed i guess - so it is a combination of content and format that will force the crash / loop / dead-end.
Let's assume duplicate of existing major bug. *** This bug has been marked as a duplicate of bug 38575 ***
(In reply to Timur from comment #10) > Let's assume duplicate of existing major bug. > > *** This bug has been marked as a duplicate of bug 38575 *** you speak about the "libreoffice blocks" from 7 years ago? So - if this is a dupe ... will it mean the authors of this application are unable to fix that bug within 7 years? For me as a headless user, this is a very critical bug ... I for myself found a way to bypass it by watching the task (linux side) and kill it, if it is not finished within 30 seconds. This is a very brute way and will not work for jobs taking longer than 30 seconds but for what i need it (convert document to pdf) - it will do the job.
Timo, your comment is often seen butis also "no-value", it just servers to annoy people. Who do you complain to? To us, volunteers, who are givers? While you as a user are 'angry' taker? You want it to be fixed? Good, do it yourself, of find someone who will do it as a volunteer, or pay to programmer or company to have it done, or just wait, month or decade. I'l never understand people like you, who don't bother to understand how OSS project functions but complain instead, repelling others who contribute.
(In reply to Timur from comment #12) > Timo, your comment is often seen butis also "no-value", it just servers to > annoy people. > Who do you complain to? To us, volunteers, who are givers? While you as a > user are 'angry' taker? > You want it to be fixed? Good, do it yourself, of find someone who will do > it as a volunteer, or pay to programmer or company to have it done, or just > wait, month or decade. > I'l never understand people like you, who don't bother to understand how OSS > project functions but complain instead, repelling others who contribute. I know how open source works and at no point have I discredited the "givers". Maybe it is because i am not writing in my native language here and therefore use words or phrases that attack you or someone else. I'm sorry for that. As said, for me this is a critical bug, not just a cosmetics or the like. If a bug is known for like 7-9 years and will / can harm a normal user in a normal daily situation, would'nt it get some kind of "priority" to fix it? That is what i wanted to say with my short feedback. I did my best in reporting and am again sorry when someone feels attacked.