Bug 103056 - Endless scrolling upwards after reaching top of document
Summary: Endless scrolling upwards after reaching top of document
Status: RESOLVED DUPLICATE of bug 82762
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Writer (show other bugs)
(earliest affected) release
Hardware: All All
: medium normal
Assignee: Not Assigned
Keywords: bibisected, regression
Depends on: 112082
Blocks: 49260
  Show dependency treegraph
Reported: 2016-10-08 10:24 UTC by Yousuf Philips (jay) (retired)
Modified: 2020-09-17 09:22 UTC (History)
2 users (show)

See Also:
Crash report or crash signature:


Note You need to log in before you can comment on or make changes to this bug.
Description Yousuf Philips (jay) (retired) 2016-10-08 10:24:59 UTC
1. Open attachment 69995 [details] and notice that it might automatically jump to page 2
2. Scroll to bottom of document
3. Scroll to top of document
4. Once reaching top of page 1, it jumps me back to page 2

Regression as this doesnt happen in 3.6. 4.0 and 4.1 are really buggy to use with this document. Not sure bibisect is needed for this.

Build ID: f309531cfe1d6a1b6ea1306d45ed3e121145bc5f
CPU Threads: 2; OS Version: Linux 3.19; UI Render: default; 
TinderBox: Linux-rpm_deb-x86_64@70-TDF, Branch:master, Time: 2016-10-04_03:52:18
Locale: en-US (en_US.UTF-8); Calc: group
Comment 1 Joel Madero 2016-10-09 04:53:19 UTC
Build ID: f32b300c9e071c2bdee2a7b925200feb93add702
CPU Threads: 2; OS Version: Linux 3.16; UI Render: default; 
Locale: en-US (en_US.UTF-8); Calc: group

Comment 2 jebsolutions 2016-10-21 21:21:34 UTC
I am seeing the same thing.  5.3.0 Alpha 1 on Windows 10 64bit and Lubuntu 16.04.

The layout/redraw of file 69995 is definitely weird.  It's acting like the anchor positions of all the images/pages are self-referential and it's wrongly recalculating everything every time you scroll/move?  Everything is jumping around when I scroll.
Comment 3 Timur 2017-08-29 09:51:48 UTC
This one could be a duplicate of original Bug 49260.
Comment 4 Buovjaga 2018-05-24 17:45:10 UTC
I am unable to repro the original thing, but with GTK3 backend LibreOffice freezes in such a way that the mouse stops working. Luckily I thought to try Alt-tabbing away into the terminal so I could kill it.

Arch Linux 64-bit
Build ID: eeaf6dee2d278eaa037d95a756ad0ffab3314bc2
CPU threads: 8; OS: Linux 4.16; UI render: default; VCL: kde4; 
Locale: fi-FI (fi_FI.UTF-8); Calc: group
Built on May 24th 2018
Comment 5 Buovjaga 2018-07-07 19:49:11 UTC
Bibisected on Linux with 42max. The targeted "bad" commit was such that if I scrolled to the bottom, my whole system locked up and I had to reset it from Virtualbox. Thus, I relied on the view on 1st page. I had to skip many builds that crashed on launch, but it did give me a single commit:

commit d48013eaa4a4d0aa2c702588af0cefe8fad307a5
Author: Matthew Francis <mjay.francis@gmail.com>
Date:   Sat Sep 5 17:59:25 2015 +0800

    commit 12a4200e8ff7f045efcc7e9d15a24b15b248c437
    Author:     Armin Le Grand <alg@apache.org>
    AuthorDate: Thu Aug 9 08:42:27 2012 +0000
    Commit:     Caolán McNamara <caolanm@redhat.com>
    CommitDate: Tue May 28 15:55:23 2013 +0100
        Related: #i120498# Enhanced Undo/Redo and user experience...
        when editing texts in graphic objects and/or tables
        (cherry picked from commit a096725779b210c7a2706e72596fd7c80e049fdd)
Comment 6 Buovjaga 2018-07-07 20:20:42 UTC
Probably my last comment is useless. I tried with 43max now and its oldest commit does not have the problem. The latest ones again had the terrible system-freezing effect. I did a bibisect with it, but had to skip a ton of builds that crashed on start and I got this huge list of potential commits (repo commits), which seem to be in random order:

There are only 'skip'ped commits left to test.
The first bad commit could be any of: 107ef960b6ebe35aed35af88d65627a1f1b6f9f9 41cb2fd3cdbc22c166e006445946959b62f216da c98b9a8c73751bf93370d3feae309757f40861f1 e63de08bd71059bc9620e3ef2614884c6b9abfc5 362ab4dc1ac92046567f742b0c668ec8b8dafdf8 62377c3a3f75dc71cc2915361f20da59077215b5 3639f96588f9e619fe9de64122eb8687275d5a75 a571aed326b970d74a80f07f3f4fdcfdb6f27095 a9fb029ede8a9b0b83697c017eb9288597dd81aa feaf6ed613ae8388d1aab35e7068b84eafd94e8d 5f77321d3b767eda718614d0a2647447cf4f6f2c af3738481b95e4747478021f93aac10e80a35c52 10675d424c1196987b0393c416846ca30c856eb7 a28f949237320d7571fd463e89050342ba17d6a5 cc2e0648242c4a8ce9c80e51745195704ad567e5 d89f6896b830e4a69980934f11ddb741d91713b0 6c34995165ba70136af5bc792bb7e54ca5252c81 801bd681d5a07350b610b10800b71d6939aa418e 6a9cbdb0047accc2182407454015345e47dd812d b88fdcb82bbfef449e7b39122232aa5c80fa74fa 63b9201c09523506ca25a68285821226d1572f02 7d55904fe749e104ffdbf7200346e53a4988f5af 12af9036241585d5100a046879d8e148f698f2de eeec746d9b9a090137faa72c3f93865364531e04 5963d2b0fe230b433f16a00f4d4b03cadf4bcb44 f9bbb93682b5028ef66fbf65a0e93aeefc1482ec aaeb37674e976284cafcddfa89aaa6a56ad5c9f6 fd83000b95a86a2b443cfacfb8b00b199e3d29d7 6ffdecbc51957bb177eeab3419745cb69f6515b3 aaf802edc84dfa85398fce41d822f4051e626a19 3196cf5110dea9c2838ad342ad8579d880eb13c2 9ae371b822d8765ff386b58c6d06b27803503eb8 0e23110e5607fc5a47dad7bab02fd32202dfe053 77f2dd09089e58cacd33e6c13f865f65999f7b98 5b38b5bf32ed27c814aaec578410e87300b2feaa ddbe8c28cd323f1d48cf58f272a0c6b15a2d18f4 1b61dacb77b68fd6dd15e473aa2b4426ee049c38 a7293616b080d83b9d90e494f5aae5c3e23158e0 3f845334fb0e08fb58b306831b4f43871dc17b9d fe37be5fddb1e252c89af3431b26d08a9da58531 bc6bc69d0ab1e28c307e136d72fff12685e5e6ce
We cannot bisect more!

I don't have time now to somehow script pulling the source hashes from that list.
Comment 7 QA Administrators 2019-07-08 02:48:18 UTC Comment hidden (obsolete)
Comment 8 Timur 2020-09-17 09:22:52 UTC
attachment 69995 [details] freezes LO 7.1+ on 1st scroll. 

But regardless,key here is to search "attachment 69995 [details]" to find many bugs on the same 3-pages 2007 DOCX which originated in bug 49260 . 
File is wrong itself, seen in resaved in MSO, it opens differently. 
So no point ih having bugs for it, except freeze which is bug 82762.
So I close this redundant one.

*** This bug has been marked as a duplicate of bug 82762 ***