www/drupal11: update to 11.4.7
This is security update.
11.4.5 (2026-08-06)
This is a patch (bugfix) release of Drupal 11 and is ready for use on
production sites. Learn more about the latest version of Drupal.
https://www.drupal.org/project/drupal/releases/11.4.5
11.4.6 (2026-09-03)
This is a patch (bugfix) release of Drupal 11 and is ready for use on
production sites. Learn more about the latest version of Drupal.
https://www.drupal.org/project/drupal/releases/11.4.6
11.4.7 (2026-09-16)
[12 lines not shown]
yices2: update to 2.7.0.
To fix build on -current.
New Features
Finite Fields support in the MCSat Solver
Yices2 Portfolio Python script
New cache mechanism in the MCSat Solver: Target cache and Best cache
New decision heuristics in the MCSat Solver
Partial restarts option in the MCSat solver
Print models in the SMT2 format by default
And many bug fixes
libextractor: update to 1.15.
Security fix release.
Fix pkglint.
Add test arget.
Mon Jun 29 07:58:39 PM CEST 2026
Fix potential 4 MB on-stack memory allocation that could
result in a stack-based buffer overflow in the OLE2 extractor.
Thanks to Haitam Lazaar for reporting.
Releasing GNU libextractor 1.15. -CG
Mon May 25 12:43:20 PM CEST 2026
Revive REAL plugin (fixes #2518).
Revive VLC plugin (fixes #2075).
Fix issues in ELF plugin (fixes #8514).
Fix issues in HTML plugin (fixes #8620).
Fix issues in EXIV2 plugin (fixes #8513, #9223).
Revive Applefile plugin (fixes #2517).
[3 lines not shown]
Update to version 9.2.1116.
Changes:
- patch 9.2.1116: 'tagfunc' not used for tag completion
- patch 9.2.1115: ml_updatechunk() rescans all chunks on every line update
- patch 9.2.1114: Visual selection is not highlighted at the cursor position
- runtime(nasm): move options to separate ftplugin
- patch 9.2.1113: use-after-free and memory leak in serverSendToVim()
- runtime(c,cpp): Update syntax, fix number patterns
- runtime(sh): touch does not end at curly block
- patch 9.2.1112: set_indent() adjusts text properties without any present
- patch 9.2.1111: Crash in a zero-width window with a status line
- patch 9.2.1110: regexp NFA compilation reallocates scratch buffers each time
- patch 9.2.1109: Popup image is drawn over a popup with a higher zindex
- runtime(doc): fix wrong notation of :qall in quickfix.txt
- patch 9.2.1108: Scrolling recomputes the height of every visible line
- runtime(nginx): syntax highlighting is slow on large configs
- patch 9.2.1107: getregionpos() is slow when only the region bounds are needed
- patch 9.2.1106: GTK4: cell edges are blurry with fractional scaling
[119 lines not shown]
lang/ruby40: update to 4.0.7
This release contains fix for CVE-2026-80212 and CVE-2026-80213 in resolv
gem.
4.0.7 (2026-09-15)
* Bug #22188: addr2line doesn't find symbols when compiled with GCC LTO
* Bug #22210: Check if issue #19969 is still reproducible / re-opened in 4.0.x
* Bug #22217: Segmentation fault when resuming execution with Coverage.start
and ruby/debug
* Fix formatting in ObjectSpace._id2ref error path by eregon · Pull Request
#18206
* Bug #22200: ObjectSpace._id2ref can return a different object than the
id's owner on Ruby 4.0 (stale id2ref_tbl entry for objects with generic
[59 lines not shown]
textproc/treemd: update to 0.9.1
(leaf pkg)
[0.9.1] - 2026-09-14
Fixed
Images inside a list item are drawn. - item  rendered the text and nothing else, and the image was not selectable. The list renderer drops images from its inline run on the grounds that they are drawn separately, which was true while the parser hoisted a list item's image out to a top-level block; since 2.0.0 it correctly keeps the image in the item, so nothing picked it up. The item now reserves the same placeholder rows a paragraph does, registers the image in the element index, and the existing image pass draws into them (#87)
A nested paragraph or list is no longer measured as if it reserved image rows. Only a top-level block reserves them, so counting a nested one that way overstated its height by thirteen rows and pushed everything below it out of position
Internal
The renderer is checked against the line count that production actually uses. The check previously compared against count_block_lines, which the live path does not call for a top-level list, paragraph or image, so those three were being validated against a function that never measures them. index_elements now records the rows it walked and the test compares that
[0.9.0] - 2026-09-14
turbovault-parser 2.1.0, which closes every parser defect reported from here since 0.7.0.
Fixed
[16 lines not shown]
