<html> <head> <base href="https://bugs.documentfoundation.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEEDINFO " title="NEEDINFO - Text is invisible, both in UI and in editor on update to 5.3.2, older AMD and Intel GPUs" href="https://bugs.documentfoundation.org/show_bug.cgi?id=107166#c122">Comment # 122</a> on <a class="bz_bug_link bz_status_NEEDINFO " title="NEEDINFO - Text is invisible, both in UI and in editor on update to 5.3.2, older AMD and Intel GPUs" href="https://bugs.documentfoundation.org/show_bug.cgi?id=107166">bug 107166</a> from <a class="email" href="mailto:vstuart.foote@utsa.edu" title="V Stuart Foote <vstuart.foote@utsa.edu>"> V Stuart Foote</a> <pre>(In reply to Tom from <a href="show_bug.cgi?id=107166#c118">comment #118</a>) > ... > Now, from a systems analyst's point of view let's examine the issue..., > > 1. Just exactly WHAT changed between 5.2.7 and 5.3.4? Are there > change logs, work requests or other documents to be examined? It is all there well documented, up to YOU to dig it out--BZ [1], cgit [2], gerrit [3], opengrock [4] > 2. Is the particular change item that is causing this whole discussion > really necessary? Or is it just 'nice to have'? Very necessary. > 3. I keep seeing graphics stack mentioned here -- what graphics changed > and why? For code maintenance and consistent implementation we moved the font rendering to HarfBuzz cross platform. That, while implementing OpenGL rendering cross platform. The combination revealed weakness in our legacy GDI/GDI+ based graphics stack on Windows. It necessitated a lot of refactoring to move most of the font rendering to MS DirectWrite Direct2D calls with some residual cross platform GDI, and fall-back GDI when OS/GPU driver fails. Issue here has been buggy DirectX drivers provided by GPU vendors, the fix has been to provide better error handling and fall-back. > > Slow down and consider you customer base, and then make changes, please. Not going to happen. Users are welcome to stay on EOL releases. Rather the project expects input from all users and competent "systems analysts" during the development and QC cycle. The project is on a timed-release development cycle [5]--changes go ready or not, the key is early testing and dev/QC feedback. The more testers/testing the fewer the hidden regressions. The issue here with marginal driver support was simply not detected until more users shifted to a 5.3 release build. =-refs-= Tools for following Project development [1] <a href="https://bugs.documentfoundation.org/query.cgi?format=advanced">https://bugs.documentfoundation.org/query.cgi?format=advanced</a> [2] <a href="https://cgit.freedesktop.org/libreoffice/core/">https://cgit.freedesktop.org/libreoffice/core/</a> [3] <a href="https://gerrit.libreoffice.org/gitweb?p=core.git;a=summary">https://gerrit.libreoffice.org/gitweb?p=core.git;a=summary</a> [4] <a href="https://opengrok.libreoffice.org/">https://opengrok.libreoffice.org/</a> Our timed-release cycle [5] <a href="https://wiki.documentfoundation.org/ReleasePlan">https://wiki.documentfoundation.org/ReleasePlan</a></pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the assignee for the bug.</li> </ul> </body> </html>