Menu
▾
▴
[Pcb-src-changes] pcb.git: branch: master updated (262ec1ab712879aa27a4e1b318b8128a8bf45fd4)
|
From: <gi...@gp...> - 2010-09-29 20:36:47
|
The branch, master has been updated
via 262ec1ab712879aa27a4e1b318b8128a8bf45fd4 (commit)
via 72f61208d539641f80004ed4608cbfddc3291f25 (commit)
via fdd585acc4f8f3c414af48e78d12b426aaa2b49e (commit)
from 5be5cf3f57c61c2e0bcb5da241514012350ccdbd (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
=========
Summary
=========
ChangeLog | 1743 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
NEWS | 60 ++
po/POTFILES.in | 5 +-
3 files changed, 1806 insertions(+), 2 deletions(-)
=================
Commit Messages
=================
commit 262ec1ab712879aa27a4e1b318b8128a8bf45fd4
Author: DJ Delorie <dj...@de...>
Commit: DJ Delorie <dj...@de...>
Update NEWS from ChangeLog
:100644 100644 d4713e6... d3435ec... M NEWS
commit 72f61208d539641f80004ed4608cbfddc3291f25
Author: DJ Delorie <dj...@de...>
Commit: DJ Delorie <dj...@de...>
Update ChangeLog from git logs.
:100644 100644 b26c431... 672dae6... M ChangeLog
commit fdd585acc4f8f3c414af48e78d12b426aaa2b49e
Author: DJ Delorie <dj...@de...>
Commit: DJ Delorie <dj...@de...>
Fix for distcheck errors.
:100644 100644 9ef4db5... 3bfb3ed... M po/POTFILES.in
=========
Changes
=========
commit 262ec1ab712879aa27a4e1b318b8128a8bf45fd4
Author: DJ Delorie <dj...@de...>
Commit: DJ Delorie <dj...@de...>
Update NEWS from ChangeLog
diff --git a/NEWS b/NEWS
index d4713e6..d3435ec 100644
--- a/NEWS
+++ b/NEWS
@@ -1,4 +1,64 @@
========================================================================
+Release Notes for PCB snapshot 20100929
+========================================================================
+This release represents over 240 commits and as such this summary
+clearly is not complete. See the ChangeLog file for the complete list
+of changes.
+
+Features
+--------
+
+- PCB can directly import (forward annotate) schematics while running.
+- Many places where measurements are entered now accept units, like
+ "5mm", and "cm" and "in" are allowed too.
+- Free Rotate Buffer added to menu
+- Polygon Hole tool.
+- F12 invokes the Lock tool.
+- Russian translation added.
+- DBUS is enabled by default when possible.
+- Command-line exporters can run action scripts too.
+- GTK tool tips on elements, pins, and nets.
+- Command-line actions in GUIs do not require (,,) syntax.
+- PCB can import footprints both in subdirectories and the named directories.
+- New GCode exporter.
+- Footprint files use ".fp" as the suffix; this is automatically
+ searched for if needed.
+- GTK dialogs do not show at startup unless requested.
+- Stackup can be specified on the command line for command-line exporting.
+- Reference card updated.
+- The snap-point in pads is the center.
+- The 's' key toggles polygons to clear/notclear all pads and pins
+
+Bug Fixes
+---------
+
+- Various PNG export fixes.
+- lesstif's magic toggle-layers-on-flip fixed.
+- Arc bounding box math fixed.
+- Fixed: sometimes, disabling silk also disabled the solder layer.
+- "pcb --help" prints all options again
+- Add mnemonic for "close without saving"
+- Don't select silk on the back side when it's invisible.
+- Various polygon rendering bugs fixed.
+- PNG drill bloat export fixed.
+- Attempts to change polygon clearance *on* polygons results in a hint.
+- PNG doesn't segfault if not all graphic formats are supported.
+- Various toporouter fixes and improvements
+- HID_Mixed is supported in attribute dialogs
+- Flipped rectangles in PS exports are drawn correctly.
+- Optimizer removes micro-traces under pads.
+- Vias, pins, and pads are now drawn under the mask instead of over.
+- Windows supported better.
+
+Other
+-----
+
+- Sources converted to C99
+- Doc extraction gives more control over categories.
+- Thermal documentation updated.
+- Actions can see the context they were invoked in
+
+========================================================================
Release Notes for PCB snapshot 20091103
========================================================================
This release represents almost 200 commits and as such this summary
commit 72f61208d539641f80004ed4608cbfddc3291f25
Author: DJ Delorie <dj...@de...>
Commit: DJ Delorie <dj...@de...>
Update ChangeLog from git logs.
diff --git a/ChangeLog b/ChangeLog
index b26c431..672dae6 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,1746 @@
+2010-09-29 DJ Delorie * dj AT delorie dot com *
+
+ * src/action.c: Make new text clear polygons
+
+2010-09-20 DJ Delorie * dj AT delorie dot com *
+
+ * src/parse_l.l: Don't print "can't open ./default_font for reading"
+ all the time
+
+2010-09-19 DJ Delorie * dj AT delorie dot com *
+
+ * src/file.c, src/parse_l.l: Wrap more messages in debug Messages about where fonts were found, or how many milliseconds it
+ takes to load a file, are not normally needed. Wrap them in DEBUG
+ conditionals.
+
+2010-09-14 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/png/png.c: If we don't use the outline to clip, don't draw
+ it either.
+
+2010-09-14 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/png/png.c: Fix bug tracker # 3064887 PNG exported all black in photo mode if the "outline" was exactly
+ the board size, because the left/bottom edges were clipped leaving a
+ big hole. With this patch, we do two things: 1. Detect a non-trivial outline, and only do outline processing if
+ found. Thus, a rectangular outline exactly the size of the board is
+ just ignored - no black border on the output image. 2. Any outline line that's drawn exactly on the bottom/right edge is
+ brought in by half a pixel to ensure it's drawn in the resulting
+ image.
+
+2010-09-13 DJ Delorie * dj AT delorie dot com *
+
+ * src/action.c: Avoid divide-by-zero Check for a zero dispersion before computing it.
+
+2010-09-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/thermal.c: thermal.c: Initialise flags of the LineType object
+ in square_therm() Keeps valgrind quiet about conditionals based upon uninitialised
+ variables.. must be a good thing.
+
+2010-09-10 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/action.c, src/autoplace.c, src/autoroute.c, src/buffer.c,
+ src/copy.c, src/create.c, src/crosshair.c, src/data.h, src/djopt.c,
+ src/draw.c, src/file.c, src/find.c, src/flags.c,
+ src/hid/batch/batch.c, src/hid/common/extents.c,
+ src/hid/common/flags.c, src/hid/gcode/gcode.c,
+ src/hid/gerber/gerber.c, src/hid/gtk/gtkhid-main.c,
+ src/hid/gtk/gui-config.c, src/hid/gtk/gui-output-events.c,
+ src/hid/gtk/gui-top-window.c, src/hid/lesstif/dialogs.c,
+ src/hid/lesstif/main.c, src/hid/lesstif/menu.c,
+ src/hid/nelma/nelma.c, src/hid/png/png.c, src/hid/ps/eps.c,
+ src/hid/ps/ps.c, src/line.c, src/macro.h, src/misc.c, src/move.c,
+ src/polygon.c, src/print.c, src/puller.c, src/rats.c, src/rtree.c,
+ src/rubberband.c, src/search.c, src/select.c, src/toporouter.c:
+ Split usage of "max_layer" into max_copper_layer and max_group PCB has an equal number of layer groups as copper layers, so the
+ "max_layer" #define'd variable has been used throught the codebase
+ to delimit iteration over both layer groups, and layers. This makes it hard for people reading the codebase to identify
+ whether a given loop is over layers or layer groups (since the
+ counter variable is often fairly terse or indescript). Confusingly,
+ in some of the code, layer groups indexes are stored in a variable
+ called "layer". To aid understanding of the code, this commit removes the max_layer
+ variable (#define'd to (PCB->Data->LayerN)), and replaces it with
+ two similarly #define'd variables, "max_copper_layer" and
+ "max_group". This commit also introduces convenience macros: #define solder_silk_layer (max_copper_layer + SOLDER_LAYER) #define component_silk_layer (max_copper_layer + COMPONENT_LAYER) Which make the code intention much clearer for a common case of the
+ old max_layer variable usage.
+
+2010-09-10 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/lesstif/main.c: Fix lesstif SwapSides magic layer option. If the first layer in the layer group is silk, the swap sides key
+ won't auto-toggle the solder and component groups, nor properly
+ activate them if it does. Add code to find a visible *copper* layer
+ in the layer group and use that.
+
+2010-09-09 Ineiev * ineiev AT gmail dot com *
+
+ * src/find.c: find.c: Add RATLINE_TYPE to switch statement in
+ BuildObjectList Part of Ineiev's patch from:
+ http://archives.seul.org/geda/user/Nov-2009/msg00456.html Modified-by: Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+2010-09-09 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/find.c: Fix crash due to uninitalised variable in DRC check
+ Bug #3033772 Andrew Armenia noted that the second pointer returned from
+ BuildObjectList(), "object_type_list" would be left uninitialised if
+ the object type was not encountered. A subsequent call to free() in
+ DRCFind() would then cause memory corruption. The issue was originally reported by Ineiev:
+ http://archives.seul.org/geda/user/Nov-2009/msg00456.html That patch addresses the lack of a RATLINE_TYPE in the
+ BuildObjectType function, but does not prevent a segfault for other
+ unknown types.
+
+2010-09-08 DJ Delorie * dj AT delorie dot com *
+
+ * src/misc.c: Fix arc bounding box math. Calculations for Arc bounding boxes weren't taking into account the
+ wide range of possible starting arcs. This patch uses modulus to
+ force angles into canonical forms and iterates through quadrants to
+ capture the full extents of each arc.
+
+2010-09-08 DJ Delorie * dj AT delorie dot com *
+
+ * src/parse_y.y: Fix typo in parse_y.y A typo in parse_y.y was causing the documentation to be corrupted.
+
+2010-09-07 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/gtk/gtkhid-main.c, src/hid/lesstif/main.c: Fix layer
+ visibility logic for some boards. If the Groups() line listed the c/s layer before the copper layer,
+ disabling silk would also disable the copper. This change searches
+ the entire layer group looking for any layer that might need to be
+ drawn, and uses that as the exemplar instead of blindly choosing the
+ first layer in the list.
+
+2010-09-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/gpcb-menu.res, src/pcb-menu.res: Add import settings to menus At the bottom of the Settings menu, add a submenu for import
+ settings for placing new elements.
+
+2010-09-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/action.c: Store units for import::disperse If the user specified the dispersement as "123mm" for example, store
+ that string as-is for convenience.
+
+2010-09-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/misc.c: Enhance GetValue Allow cm and in as units. Allow the units to be part of the value
+ string or separate.
+
+2010-09-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/action.c: Add new-part-location to import Adds import::newX, import::newY, and import::disperse attributes.
+ These determine where new parts are located when they're added to
+ the layout, including some random variation to keep them from being
+ all on exactly the same spot. Parameters were added to Import() to
+ set these: In @code{Import()} is called with @code{setnewpoint} then the
+ location of new components can be specified. This is where parts
+ show up when they're added to the board. The default is the center
+ of the board. @table @code @item Import(setnewpoint) Prompts the user to click on the board somewhere, uses that point.
+ If called by a hotkey, uses the current location of the crosshair. @item Import(setnewpoint,mark) Uses the location of the mark. If no mark is present, the point is
+ not changed. @item Import(setnewpoint,center) Resets the point to the center of the board. @item Import(setnewpoint,X,Y,units) Sets the point to the specific coordinates given. Example:
+ @code{Import(setnewpoint,50,25,mm)} @end table Note that the X and Y locations are stored in attributes named
+ @code{import::newX} and @code{import::newY} so you could change
+ them manually if you wished. Calling @code{Import(setdisperse,D,units)} sets how much the newly
+ placed elements are dispersed relative to the set point. For
+ example, @code{Import(setdisperse,10,mm)} will offset each part
+ randomly up to 10mm away from the point. The default dispersion is
+ 1/10th of the smallest board dimension. Dispersion is saved in the
+ @code{import::disperse} attribute.
+
+2010-09-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/misc.c, src/misc.h: Add AttributeRemove functions Add AttributeRemoveFromList() and the wrapper AttributeRemove()
+
+2010-08-28 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-top-window.c: hid/gtk: Remove unused colour
+ allocation for "BlackPixel"
+
+2010-08-28 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-top-window.c: hid/gtk: Remove unused extern HID
+ ghid_hid from gui-top-window.c
+
+2010-08-25 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/draw.c: draw.c: Remove the unused "Erasing" flag / counter This variable tracks whether the object being drawn is to be erased.
+ It is never actually tested, so remove it.
+
+2010-08-25 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/draw.c: draw.c: Initialise integer counters with 0, not
+ "false". For the "doing_assy" flag, fix its type to "bool" rather than "int".
+
+2010-08-25 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/draw.c: draw.c: Remove "HaveGathered" parameter from
+ DrawLineLowLevel() This is superflous, as it is only passed as TRUE when the Gathering
+ variable it overrides is known to be FALSE anyway.
+
+2010-08-25 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/data.c, src/data.h, src/draw.c: Remove unused global variable
+ "render", and "extern bool RedrawOnEnter" The "render" variable was set and reset, but was not used anywhere.
+
+2010-08-24 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-output-events.c: hid/gtk: Track previous state of
+ crosshair attachements when undrawing For correctness - this doesn't necessarily fix any rendering bugs.
+
+2010-08-24 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-utils.c: hid/gtk: Simplify ghid_draw_area_update() If the rect passed is NULL, we can simply pass that on to
+ gdk_window_invalidate_rect() which will invalidate the whole area.
+
+2010-08-24 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-utils.c, src/hid/gtk/gui.h: hid/gtk: Remove unused
+ functions relating to string display Removes ghid_string_markup_extents() and ghid_string_markup()
+
+2010-08-24 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-utils.c: hid/gtk: Remove unused
+ ghid_draw_area_clear() function
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/gtk/gui-top-window.c: Merge GTK attributes. Since the new help system only allows one attribute (i.e. command
+ line) list per HID, combine the two GTK lists into one.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/lesstif/main.c, src/hid/lesstif/menu.c: Merge lesstif
+ attributes. Since the new help system only allows one attribute (i.e. command
+ line) list per HID, combine the two Lesstif lists into one.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/main.c: Print all command line options with help. Keep track of which attributes we've printed as part of the HID
+ help. After that, print any remaining attributes we know about that
+ we haven't already printed.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/gtk/gtkhid-main.c, src/hid/gtk/gui-top-window.c: Export
+ GTK attributes. The new help system uses export attribute lists for help, so the
+ Lesstif HID needs to export them.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/lesstif/main.c: Export lesstif attributes. The new help system uses export attribute lists for help, so the
+ Lesstif HID needs to export them.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/action.c: Check for NULL element name when adding attributes. When setting element attributes, if the element is as yet unnamed,
+ use an "unnamed element" title instead of segfaulting.
+
+2010-08-15 Robert Spanton * rspanton AT zepler dot net *
+
+ * po/ru.po, src/hid/gtk/gui-dialog.c: Add a keyboard shortcut to
+ pcb's "close without saving" option This patch adds a keyboard mnemonic to the "close without saving"
+ option in the close confirmation dialog box.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/action.c: Update Import() docs Fix capitalization. Mention that elements that should be removed
+ are left, but selected, after an Import.
+
+2010-08-15 DJ Delorie * dj AT delorie dot com *
+
+ * src/buffer.c, src/gpcb-menu.res, src/pcb-menu.res: Add
+ FreeRotateBuffer to the Menu Modify FreeRotateBuffer() to prompt the user if no angle is passed.
+ Add an entry to the menus to call it that way.
+
+2010-08-10 kai-martin * kmk AT bibo dot iqo dot uni-hannover dot de *
+
+ * src/action.c: modified: action.c (fix element name position on
+ import schematic)
+
+2010-08-14 DJ Delorie * dj AT delorie dot com *
+
+ * tools/gnet-pcbfwd.scm: Quote strings Quote all strings that might contains commas or close parens, to
+ avoid syntax errors when those names are passed to actions.
+
+2010-08-03 DJ Delorie * dj AT delorie dot com *
+
+ * src/select.c: Don't select silk on the far side if the far side is
+ not shown The logic for determining if a layer was "on" didn't take into
+ account that the two silkscreen layers are handled differently.
+ Thus, if you tried to select shown silk, you would select hidden
+ silk on the other side too, unknowingly moving that around. With this patch, silk on the far side is only selected if the far
+ side ("invisible" side) is shown.
+
+2010-06-06 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/action.c, src/const.h, src/crosshair.c, src/flags.c,
+ src/gpcb-menu.res, src/hid/gtk/gui-icons-mode-buttons.data,
+ src/hid/gtk/gui-misc.c, src/hid/gtk/gui-output-events.c,
+ src/hid/gtk/gui-top-window.c, src/hid/lesstif/main.c,
+ src/pcb-menu.res, src/set.c: Introduce POLYGONHOLE_MODE for creating
+ holes in polygons Having selected polygon hole mode, the first click selects which
+ polygon to cut a hole in. A second click defines the start point of
+ the hole contour. The tool then behaves in a similar way to the
+ polygon drawing tool, with the hole ending when the start point is
+ re-clicked. To avoid creating illegal polygons, the hole drawn is subtracted
+ from a representation of the original polygon with the
+ poly_Boolean_free(). This consolidates any contours it intersects
+ with and prevents the user defining contours which intersect each
+ other. (Although we don't currently prevent the the user drawing
+ self-intersecting contours). The resulting POLYAREA is re-processed into PolygonType objects,
+ potentially more than one - if the hole drawn bisects the original
+ polygon. To keep undo operations simple, these are added as
+ completely new objects and the original polygon is deleted - along
+ with its ID.
+
+2010-06-06 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon.c, src/polygon.h: Expose APIs for creating POLYAREA
+ from PolygonType objects and back The PolygonPoly() API wraps polygon.c's original_poly() function,
+ whilst PolyToPolygonsOnLayer() converts the passed POLYAREA and all
+ those linked to it into discrete PolygonType objects on the board.
+
+2010-06-06 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/action.c, src/autoroute.c, src/buffer.c, src/copy.c,
+ src/create.c, src/create.h, src/crosshair.c, src/file.c,
+ src/file.h, src/global.h, src/insert.c, src/insert.h, src/mymem.c,
+ src/mymem.h, src/parse_l.l, src/parse_y.y, src/polygon.c,
+ src/polygon.h, src/remove.c, src/report.c, src/undo.c, src/undo.h:
+ Support holes in input polygons (incl. file-format addition) PCB file-format date is now 20100606, and files saved with this or
+ greater PCB version will not load with older versions of PCB. If a
+ particular board doesn't make use of the polygon hole feature, the
+ PCB revision date in the file can be manually reset to 20070407. The file-format addition is as follows. Previously, a polygon would
+ be specified as a series of coordinates, such as: Layer(1 "component") ( Polygon("clearpoly") ( [6000 6000] [81000 6000] [81000 59000] [6000 59000] ) ) This commit introduces the ability to specify negative contours
+ which form holes in the polygon shape, e.g.: Layer(1 "component") ( Polygon("") ( [6000 6000] [81000 6000] [81000 59000] [6000 59000] Hole ( [76000 55000] [76000 38000] [58000 38000] [58000 55000] ) Hole ( [10000 10000] [10000 28000] [27000 28000] [27000 10000] ) ) ) The winding order of the contours specified in the file does not
+ matter, since PCB will automatically invert the order of the points
+ as necessary (as it always did with the outer contour). Hole contours should not intersect or self-intersect (although this
+ isn't checked for at load time). Hole contours must not intersect
+ the polygon's outer contour. Technical details: The PolygonType structure has a number of new fields, the critical
+ ones being an array of indices defining the start of hole contours. (PolygonType *)->HoleIndex[n] The number of hole contours is stored in (PolygonType
+ *)->HoleIndexN, and the maximum allocated memory for indices in
+ (...)->HoleIndexMax. The first hole contour starts at the point
+ given by (...)->Points[(...)->HoleIndex[0]], and continues until the start of
+ the next contour, or the last point defined. By storing all polygon points (including holes) in the existing
+ array (...)->Points[], existing code which operates on the polygon as a
+ whole, e.g. translation and rotation, can operate without change. For other operations, determining wrap-around to operate within the
+ same contour requires more computation. Some helper functions have
+ been introduced in polygon.c to aid this, next_contour_point() and
+ prev_contour_point(). Where applicable, these have been used to
+ simplify existing code which used ad-hoc wrap-around code. polygon_point_idx() computes the array index of a point in a polygon
+ from its PointTypePtr address. This is used to replace a search
+ idiom used in a number of places. polygon_point_contour() returns
+ the number of the contour a given point index belongs in, 0 for the
+ outer contour, 1 for the first hole etc.. Undo: Undo with holes has become a little more complex. The undo for a
+ point removal must now record which contour the point came from.
+ This is determined by the index of the removed point, and a new
+ boolean flag "last_in_contour", indicating if the point was at the
+ end of its contour. This flag is passed to InsertPointIntoObject(),
+ which uses it to disambiguate inserting a point at an index on the
+ boundary of two contours. Undo operations for removing hole contours "cheat" by saving a copy
+ of the whole polygon into the undo buffer rather than attempting to
+ describe the operation as a delta change to an existing polygon.
+ When undoing, the object IDs are swapped to keep them consistent.
+
+2010-06-05 Jared Casper * jaredcasper AT gmail dot com *
+
+ * src/hid/gtk/gui-log-window.c: Add missing return statement.
+
+2010-06-05 Jared Casper * jaredcasper AT gmail dot com *
+
+ * src/gpcb-menu.res, src/pcb-menu.res: Fix Bug #2717258, Lock mode
+ with F12. Binds F12 to lock mode in the default gcpb-menu.res and pcb-menu.res
+ files, bringing them inline with existing documentation.
+
+2010-06-05 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/crosshair.c, src/hid/lesstif/main.c: Revert some "True" ->
+ "true" string changes These were inadvertently changed by the mechanised Boolean -> bool
+ replacement in commit ad5eb5a165442be38624aa622beeb57dc1d2afbd
+
+2010-06-05 Robert Spanton * rspanton AT zepler dot net *
+
+ * src/action.c, src/action.h, src/autoplace.c, src/autoplace.h,
+ src/autoroute.c, src/autoroute.h, src/box.h, src/buffer.c,
+ src/buffer.h, src/change.c, src/change.h, src/clip.c, src/clip.h,
+ src/command.c, src/copy.c, src/copy.h, src/create.c, src/create.h,
+ src/crosshair.c, src/crosshair.h, src/data.c, src/data.h,
+ src/djopt.c, src/draw.c, src/draw.h, src/drill.c, src/file.c,
+ src/file.h, src/find.c, src/find.h, src/flags.c, src/global.h,
+ src/gts/gts.h, src/hid/gerber/gerber.c,
+ src/hid/gtk/gui-command-window.c, src/hid/gtk/gui-dialog-size.c,
+ src/hid/gtk/gui-drc-window.c, src/hid/gtk/gui-library-window.c,
+ src/hid/gtk/gui-misc.c, src/hid/gtk/gui-netlist-window.c,
+ src/hid/gtk/gui-output-events.c, src/hid/gtk/gui-top-window.c,
+ src/hid/lesstif/dialogs.c, src/hid/lesstif/library.c,
+ src/hid/lesstif/main.c, src/hid/lesstif/menu.c,
+ src/hid/lesstif/netlist.c, src/hid/lesstif/styles.c,
+ src/hid/nelma/nelma.c, src/insert.c, src/insert.h, src/line.c,
+ src/main.c, src/misc.c, src/misc.h, src/move.c, src/move.h,
+ src/mtspace.c, src/mtspace.h, src/mymem.c, src/netlist.c,
+ src/parse_l.l, src/parse_y.y, src/polygon.c, src/polygon.h,
+ src/polygon1.c, src/rats.c, src/rats.h, src/remove.c, src/remove.h,
+ src/report.c, src/rotate.c, src/rtree.c, src/rtree.h,
+ src/rubberband.c, src/search.c, src/search.h, src/select.c,
+ src/select.h, src/set.c, src/set.h, src/thermal.c,
+ src/toporouter.c, src/undo.c, src/undo.h, src/vendor.c,
+ src/vendor.h: Change all Booleans to bool. c99 provides bool. Declaring Boolean adds some obscurity to the
+ code, and could also miss out on some machine-specific
+ optimisations. This patch removes the definition of Boolean, and changes all
+ instances of it to bool. If you've come across this commit because it's causing you issues
+ when rebasing on mainline pcb, then you should find the following
+ script useful. You can use it to rewrite all of your local commits
+ to use bool instead of Boolean like so: git filter-branch --tree-filter rename-bool.sh HEAD...${HASH} Replacing ${HASH} with the latest commit that's in mainline. if [ ! -f src/autoplace.c ] then echo "This doesn't look like the PCB source directory." echo "Cowardly exiting to avoid causing misery!" exit fi chg_bool () { FNAME=$1 ADDR=$2 sed -i -e "${ADDR}s/\bBoolean\b/bool/g" "$FNAME" sed -i -e "${ADDR}s/\bTrue\b/true/g" "$FNAME" sed -i -e "${ADDR}s/\bFalse\b/false/g" "$FNAME" } find -name '*.[ch]' | while read f do chg_bool "$f" done find -name '*.[yl]' | while read f do # Replace first '%%' line with __TMP_MARKER_ sed -i -e '0,/^%%/s/^%%/__TMP_MARKER_/' "$f" # Replace in section before the first '%%' chg_bool "$f" "0,/^__TMP_MARKER_/" # Replace in section after the last '%%' chg_bool "$f" "/^%%/,$" # Remove the temporary marker: sed -i -e 's/^__TMP_MARKER_/%%/' "$f" done
+
+2010-06-03 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/file.c, src/parse_y.y: Refer to git, not cvs when reading a
+ file from newer PCB.
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * po/LINGUAS, po/ru.po: Add Russian translation
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * po/POTFILES.in: Add source files with localized strings to .po
+ generation
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * src/hid/gtk/gui-keyref-window.c, src/rats.c: Adjust punctuation
+ and capitalisation of some messages
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * src/gpcb-menu.res: Add top level menu to localization
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * src/gpcb-menu.res, src/pcb-menu.res: String unification and white
+ space cleanup String Rip up selected auto-routed tracks unified, so it can be
+ translated as one string. Also some white spaces at end of line
+ removed.
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * src/hid/common/actions.c: Localize need coord message before
+ output
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * src/action.c, src/change.c, src/create.c,
+ src/hid/gtk/gtkhid-main.c, src/hid/gtk/gui-command-window.c,
+ src/hid/gtk/gui-config.c, src/hid/gtk/gui-top-window.c,
+ src/hid/ps/ps.c: Add strings to localization
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * src/global.h, src/hid/gtk/gui.h: Move internationalization macros
+ to one header
+
+2010-06-02 Sergey Alyoshin * alyoshin dot s AT gmail dot com *
+
+ * autogen.sh: Add rule to extract localized stings from .res files
+
+2010-06-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gtkhid-gdk.c: hid/gtk: Don't call gdk_draw_points()
+ when we have no grid points to draw Fixes this warning seen when zooming close in on a course grid
+ setting: Gdk-CRITICAL **: gdk_draw_points: assertion `(points != NULL) &&
+ (n_points > 0)' failed
+
+2010-06-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Fix node_label() function to work with
+ self-intersection Rather than just giving up if we encounter our own edges in the CVC
+ list at first, skip them until we either run out of edges, or find
+ one belonging to the other polygon. I'm not 100% sure this is the correct fix, but it "seems to work". Test-case: Layer(1 "component") ( Line[60000 70000 60000 90000 4000 2000 "clearline"] Line[80000 60000 80000 90000 4000 2000 "clearline"] Line[90000 90000 90000 50000 4000 6000 "clearline"] Line[60000 40000 80000 60000 4000 6000 "clearline"] Polygon("clearpoly") ( [10000 10000] [140000 10000] [140000 140000] [10000 140000] ) )
+
+2010-06-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Fix the polygon touching contour test in
+ poly_ChkContour The following test-cases were used to help verify the changes: This polygon forms a self-touching shape like this: \| However, the right-hand edge does NOT have a node at the
+ junction. /| This previously caused it to fail the
+ self-intersection test. It should be reported as good. Polygon("") ( [85000 50000] [85000 90000] [83000 90000] [83536 63535] [85000 59999] [83535 56464] ) This polygon forms a self-intersecting shape like this: |/ (The vertical section is a straight line with no node in the
+ middle) /| It must be reported as bad. Polygon("") ( [85000 50000] [85000 90000] [83000 90000] [83536 63535] [85000 59999] [89535 56464] ) This polygon self-intersects, and must be reported as bad: Polygon("") ( [160000 50000] [160000 90000] [170000 100000] [180000 120000] [180000 150000] [160000 150000] [160000 120000] [170000 100000] [180000 90000] [180000 50000] ) This polygon self-touches, and should be reported as good: Polygon("clearpoly") ( [120000 50000] [120000 90000] [130000 100000] [120000 120000] [120000 150000] [140000 150000] [140000 120000] [130000 100000] [140000 90000] [140000 50000] )
+
+2010-06-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Fix poly_ComputeInteriorPoint() to work correctly
+ for holes The step where the algorithm finds a convex node to start from must
+ take into account whether the polygon vertices are ordered as a hole
+ or an outer contour. We now correctly compute a point inside the
+ hole, rather than possibly outside it. This fixes an assertion on the following test-case. Prior to this
+ commit, the incorrect "interior" point tested for the concave hole
+ happens to lie inside the polygon's other hole, causing it to fail
+ an assert during processing. Layer(2 "solder") ( Line[340000 160000 183700 108000 1500 3000 "clearline"] Line[92000 121000 120000 90000 1500 3000 "clearline"] Line[270000 90000 120000 90000 1500 3000 "clearline"] Polygon("clearpoly") ( [40000 40000] [320000 40000] [320000 200000] [40000 200000] ) ) The bug was created in my attempt to speed up poly_ContourInContour:
+ commit 3d0a8bd1dae0816d364a774bf9b958faf2983ec7
+
+2010-05-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Speed up poly_ContourInContour() test by computing
+ interior point NB: This introduces a behaviour change in the boundary case, that
+ two identical contours will now be considered to be inside each
+ other. First perform a test on an arbitrary boundary node (proving that the
+ contour being testing for "insideness" is not outside the other
+ contour. (This cannot not conclusively prove the contour is inside). In many cases, this simple node test gives enough evidence to return
+ 0 for the ContourInContour test computing and testing an interior
+ point. Rather than checking each exterior point, compute a strictly
+ interior point (not on the boundary), and test that against the
+ second contour. Benchmarked to improve performance over other fixes for the buggy
+ test. Example board load (CPU) times for a complex board: 21.50 (buggy contour_in_contour - single node point test) 24.43 (brute-force node point tests) 21.79 (single node test, then internal point test)
+
+2010-05-10 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Fix poly_ContourInContour() test not to return
+ TRUE for touching contours This test could previously return true for touching contours, such
+ as: __________.... |_________ | : :........ || : :: /\ : || :
+ Note that the bounding box of A is inside that of B, :: / \ :/ \ :
+ such that initial bounding box checks won't reject the ::/ A \/ B
+ \: possibility of A being inside B. ::\ /\ /: :: \ / :\ /
+ : ::..\/..:.\/..: When testing for insideness, the first point on A's contour is
+ picked. In this case, unfortunately being the touching X point
+ between the two contours. This point (correctly) returns as being
+ inside B - and the false presumption is that the whole A contour is
+ inside B. This commit introduces an unfortunately slow, but more robust test,
+ where we check each node in A for whether it is inside B. We return
+ as soon as we find an A node outside B, however this means the test
+ is VERY much slower for the case where A _is_ inside B.
+
+2010-05-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Add comment explaining assumptions for
+ poly_ContourInContour function Also, document its buggy boundary condition where the arbitrary
+ point chosen to test happens to be a common node shared between two
+ separate contours (which the test should return FALSE for).
+
+2010-05-10 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * configure.ac: Make --enable-dbus default for the GTK and Lesstif
+ HIDs
+
+2010-05-10 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/main.c: Allow running of action scripts when running PCB as an
+ exporter
+
+2010-05-10 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/polygon1.c: Improve dump_poly() output Make it print each contour of a POLYAREA, and move the NEXT PLINE
+ and NEXT POLY messages inside the loops interating over these
+ structures.
+
+2010-05-10 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gtk/gui-output-events.c: hid/gtk: Add tool-tip to identify
+ element, pin and net when hovering
+
+2010-05-09 DJ Delorie * dj AT delorie dot com *
+
+ * src/gpcb-menu.res, src/pcb-menu.res: Clarify the meaning of keys
+ bound to SetValue() Clarifies that keys which are bound to SetValue change the styles,
+ not the existing objects, so they really affect *new* objects.
+
+2010-05-04 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/png/png.c: Fix PNG bloat for drills. Keep track of when the GC is doing an erase or drill, and invert the
+ bloat for holes.
+
+2010-04-23 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/gtk/gtkhid-main.c: Allow for only one confirm button. The GTK confirm_dialog() hook now supports the user passing only one
+ button. Before, if you passed just "ok" it would create two OK
+ buttons.
+
+2010-04-23 DJ Delorie * dj AT delorie dot com *
+
+ * src/change.c: Add hint about :MinClearGap(Selected,=10,mil) Add a hint to the user about how to change the clearance of a group
+ of objects.
+
+2010-04-23 DJ Delorie * dj AT delorie dot com *
+
+ * src/change.c, src/change.h: Handle attempts to change clearance on
+ polygons. If the user attempts to change the clearance between polygons and
+ objects within the polygon by pressing 'k' over the *polygon*,
+ notice that attemp and tell the user how to accomplish what they
+ want.
+
+2010-04-12 Kai-Martin Knaak * kmk AT lilalaser dot de *
+
+ * doc/extract-docs: expand the syntax of documentating comments This is part of the effort to derive the command line option entries
+ in the manual dynamically from the source. The expanded syntax
+ allows for more control on the actual headings rendered in the
+ manual. 1) accept keys enclosed in quotation marks ("). This allows for
+ multiple word nodes in the documentation. 2) strip leading digits of the key after sort. This allows to
+ control the order of nodes from the source.
+
+2010-04-16 Jared Casper * jaredcasper AT gmail dot com *
+
+ * doc/pcb.texi, doc/thermal.pcb: Fix pr2136131. Clarify image of
+ thermal in doc. Also added some text about changing the style of thermal using
+ shift-click in the sections that talk about it.
+
+2010-03-18 Jared Casper * jaredcasper AT gmail dot com *
+
+ * src/hid/common/actions.c: Check action name at registration. Prevent actions with spaces and '(' in their names from being
+ registered; these will cause ambiguity and problems in
+ hid_parse_actionstring.
+
+2010-03-18 Jared Casper * jaredcasper AT gmail dot com *
+
+ * doc/pcb.texi, src/action.c, src/hid.h, src/hid/batch/batch.c,
+ src/hid/common/actions.c, src/hid/common/hid_resource.c,
+ src/hid/gtk/gui-command-window.c, src/hid/gtk/gui-top-window.c,
+ src/hid/lesstif/main.c, src/hid/lesstif/menu.c, src/main.c:
+ Consolidate hid action parsing. - Create a common routine hid_parse_command, which handles both
+ action script style "action(arg1, arg2);" and command entry style
+ "action arg1 arg2". This is done by making a static
+ hid_parse_actionstring function, which takes a boolean to determine
+ whether or not it should accept command entry style strings.
+ hid_parse_actions functions as it currently does, only accepting
+ action script style, but does so by calling hid_parse_actionstring
+ with TRUE. - Use hid_parse_command across all hids for user command entry,
+ removing command_parse in lesstif/main.c and batch/batch.c. - Added extra error handling to common hid_actionv to match
+ lesstif_call_action and remove lesstif_call_action.
+
+2010-04-05 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/toporouter.c: Fix some "mixed declarations and code" issues.
+
+2010-04-04 Ineiev * ineiev AT gmail dot com *
+
+ * src/misc.c: Fix "Request for bounding box of unsupported type
+ 1024" Tracker bug # 2893717
+
+ https://sourceforge.net/tracker/?func=detail&atid=538811&aid=2893717&group_id=73743
+
+2010-04-03 DJ Delorie * dj AT duopoly dot delorie dot com *
+
+ * src/hid/lesstif/dialogs.c: Fix doc typo
+
+2010-04-03 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/gtk/gtkhid-main.c, src/hid/gtk/gui-dialog.c: Open
+ schematic file chooser dialog if PCB name is not set. (cherry picked from commit 209e1031656f58ffe40189d93893457975c4042f)
+
+2010-04-03 DJ Delorie * dj AT duopoly dot delorie dot com *
+
+ * src/hid/lesstif/dialogs.c: Implement lesstif ImportGUI()
+
+2010-04-03 DJ Delorie * dj AT duopoly dot delorie dot com *
+
+ * src/action.c: Add missing newline.
+
+2010-04-03 DJ Delorie * dj AT duopoly dot delorie dot com *
+
+ * src/action.c: Let the GUI deal with choosing schematics. If the PCB is as yet unnamed, or if the pcb name doesn't correspond
+ to a schematic file, have Import() call ImportGUI() to let the user
+ tell PCB what to do. Note: corresponding HID changes are separate commits.
+
+2010-04-03 DJ Delorie * dj AT duopoly dot delorie dot com *
+
+ * src/misc.c, src/misc.h: Add AttributePut functionality. Add the logic to put arbitrary attributes into a PCB.
+
+2010-04-03 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/file.h: Remove conflicting declaration of ParseLibraryTree. This is a static function declared in file.c. Therefore I took it
+ out of file.h.
+
+2010-04-03 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/action.c: Place components in middle of board instead of at
+ 0,0.
+
+2010-03-01 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/action.c, src/buffer.c, src/change.c, src/copy.c,
+ src/create.c, src/file.c, src/file.h, src/global.h, src/main.c:
+ Support for footprints not in subdirectories Added tons of debug spew to various files to aid in program tracing
+ Also, modify library loading stuff in file.c to allow footprints to
+ live directly in library directory (rather than only one level
+ down).
+
+2010-04-02 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/png/png.c: When built with less than all of the graphic
+ formats that the png exporter supports, don't segfault anymore. The
+ problem is an array is defined differently at compile time but the
+ default value for the format was constant and a ways into the array.
+ Problem reported by jean on irc.
+
+2010-04-01 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gtk/gui-top-window.c: Inform the user where (what file or
+ compiled in defaults) the menu config came from. Should help with
+ some of the gpcb-menu.res versus pcb-menu.res and the
+ ./gpcb-menu.res versus ~/.pcb/gpcb-menu.res confusion which comes up
+ from time to time.
+
+2010-04-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/report.c: Return non-fail for ReportAllNetLengths()
+
+2010-04-01 DJ Delorie * dj AT delorie dot com *
+
+ * src/report.c: Record the calculated length in Report(NetLength)
+
+2010-03-24 Dan McMahill * dan AT mcmahill dot net *
+
+ * : commit d9e4279e4c0af142bc776bdd4a3b905388ce6b5a Author: Dan
+ McMahill * dan AT mcmahill dot net * Date: Wed Mar 24 17:38:17
+ 2010 -0400
+
+2010-03-22 anthonix * anthonix AT anthonix-desktop dot (none) *
+
+ * : commit 6081daa8966c8c83bb734789c38dd5c42dd97443 Author:
+ anthonix * anthonix AT anthonix-desktop dot (none) * Date: Mon Mar
+ 22 12:38:01 2010 +1300
+
+2010-03-21 DJ Delorie * dj AT delorie dot com *
+
+ * src/hid/lesstif/menu.c: Fix Typo
+
+2010-03-20 anthonix * anthonix AT anthonix-desktop dot (none) *
+
+ * src/toporouter.c: Toporouter: Speccut bug fix
+
+2010-03-20 anthonix * anthonix AT anthonix-desktop dot (none) *
+
+ * : commit 1ee723379413c20973bdad27b1e1cf1d722cf3d1 Author:
+ anthonix * anthonix AT anthonix-desktop dot (none) * Date: Sat Mar
+ 20 19:45:05 2010 +1300
+
+2010-03-19 Anthony M. Blake * anthonix AT anthonix dot resnet dot scms dot waikato dot ac dot nz *
+
+ * : commit e45f4e8d15f2c30f8ccb37de5b0bfa801fca572c Author: Anthony
+ M. Blake * anthonix AT anthonix dot resnet dot scms dot waikato dot
+ ac dot nz * Date: Fri Mar 19 16:25:38 2010 +1300
+
+2010-03-18 Dan McMahill * dan AT mcmahill dot net *
+
+ * tests/run_tests.sh: Missed the copyright bump when adding the
+ gcode compare routines a few weeks back.
+
+2010-03-18 Dan McMahill * dan AT mcmahill dot net *
+
+ * configure.ac, tests/golden/Makefile.am,
+ tests/golden/hid_png2/Makefile.am,
+ tests/golden/hid_png3/Makefile.am, tests/tests.list: Add tests for
+ --outfile and --dpi arguments to the PNG exporter
+
+2010-03-13 Kai-Martin Knaak * kmk AT lilalaser dot de *
+
+ * README, doc/pcb.texi, src/main.c, win32/Readme.txt: replaces
+ "pcb.sourceforge.net" with "pcb.gpleda.org"
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid.h, src/hid/common/actions.c: Apply patch from Patrick
+ Bernaud: Add function to register single action. The possibility to register a single action has been introduced with
+ commit 337fa8ba8094cee1ed291cec7bf29895ac7c9d72 but then reverted
+ later when changing way of passing context to action callback.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/common/actions.c: Apply patch from Patrick Bernaud: Modify
+ HID actions storage. It removes the HID_ActionNode structure and directly store
+ registered actions in an array of HID_Action*. It also provides a nicer way of making sure the array is sorted and
+ the code for binary search within the array has been removed in
+ favor of a call to bsearch().
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/main.c: Apply patch from Patrick Bernaud: Set 'exporter'
+ global variable before printing defaults. In print_defaults(), when the HID to print defaults is not a gui,
+ set 'exporter' global variable prior to calling HID's
+ get_export_options() callback.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/main.c: Apply patch from Patrick Bernaud: Rewrite function
+ that print HID options. The previous version of usage_hid() had a strange search for gui HID
+ attributes through 'hid_attr_node' rather than requesting them
+ directly from the HID. As a result it was calling the get_export_options() callback on
+ exporter without prior setting of the exporter global variable.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/gtk/gui-dialog-print.c: Apply patch from Patrick Bernaud:
+ gtkhid: Avoid use of 'exporter' when testing for exporter
+ availability.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/gtk/gui-dialog-print.c: Apply patch from Patrick Bernaud:
+ gtkhid: Set and unset global variable 'exporter' when printing. The GTK hid was not setting the global variable 'exporter' before
+ exporting the layout in ghid_dialog_print(): exporter HID relying on
+ this variable being set were lost.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid.h, src/hid/common/actions.c, src/hid/lesstif/menu.c: Apply
+ patch from Patrick Bernaud: Provide context to callbacks of actions
+ through global variable. Providing a context to action is necessary for scripting languages
+ (for example). To do so, a global variable ('current_action') is
+ being defined: it is saved, set and restored when running the
+ callback of an action in a way similar to what PCB already does with
+ HIDs. The callback function may access the action it belongs to and cast
+ it the way it wants to extract information from a potentially
+ extended HID_Action structure.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/common/actions.c: Apply patch from Patrick Bernaud: Sort
+ actions by references in 'all_actions'. Previously, when building 'all_actions', a copy of every HID_Action
+ was performed for each action. Now it builds 'all_actions' with
+ pointers to the actual HID_Action structures (without reallocating).
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid.h, src/hid/common/actions.c, src/hid/hidint.h,
+ src/hid/lesstif/menu.c: Apply patch from Patrick Bernaud: Revert
+ "single-action register/deregister". This reverts commits 337fa8ba8094cee1ed291cec7bf29895ac7c9d72 and
+ b274cf8fe9c8caf5bcd4edc28935c88cf5bab7f4 before the introduction of
+ a new context passing scheme for actions.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid.h: Apply patch from Patrick Bernaud: Fix comment for
+ global variable 'gui' in hid.h. hid_start_gui() does not exist. The variable is instead set in
+ main() and hid_expose_callback().
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/draw.c: Apply patch from Patrick Bernaud: Modify
+ DrawSpecialPolygon() prototype to not take a pointer on HID. DrawSpecialPolygon() was the only function to take a HID as
+ parameter. Every other draw function works on the current HID
+ through global variable 'gui'.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/gtk/gui-dialog-print.c: Apply patch from Patrick Bernaud:
+ Support HID_Mixed options in attribute dialog Since an HID_Mixed attribute is the combination of a real and an
+ enum, reuse the code for HID_Enum to provide a combo box next to the
+ spin button for the real part.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/hid/common/hidinit.c, src/hid/gtk/gui-config.c, src/main.c:
+ Apply patch from Patrick Bernaud. Modified code to not abort on HID_Mixed attributes. Th gtk HID is
+ modified to understand such an attribute (though the attribute
+ dialog has still no support for them), the lesstif HID is unchanged.
+
+2010-03-14 Stuart Brorson * sdb AT cloud9 dot net *
+
+ * src/main.c: Applying patch from Patrick Bernaud. HIDs with a name too long were breaking alignement in usage() as the
+ code requested insertion of a tab after the name even for HID names
+ already exceeding the length of the tabulation. This is similar to how usage_attr() already does alignment of name
+ and help text.
+
+2010-03-13 Dan McMahill * dan AT mcmahill dot net *
+
+ * : commit b75d1fe8d42c138ae499898db5b926948b73f767 Author: Dan
+ McMahill * dan AT mcmahill dot net * Date: Sat Mar 13 09:08:06
+ 2010 -0500
+
+2010-03-12 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid/gcode/gcode.c: Remove invalidate_wh from the new gcode HID
+ too.
+
+2010-03-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/global.h, src/main.c, src/misc.c: Remove Settings.init_done
+ variable and its usage This variable was previously used to allow action scripts run prior
+ to loading the GUI to avoid GUI startup if they executed a "Quit"
+ action. The requirement dates back prior to the HID split, where the "Quit"
+ action would cause PCB to call gtk_main_quit(), an operation which
+ is illegal if the GUI main loop was not yet started. The init_done
+ flag was used in QuitApplication(), both to determine whether the
+ GUI had started, and if not - to note that the GUI should not be
+ initialised later. (Instead of calling gtk_main_quit()). With the HID split, the QuitApplication() function was changed to
+ call exit(0) rather than gtk_main_quit(), so this dance isn't
+ necessary at all now.. in either case, we can just exit the
+ application immediately.
+
+2010-03-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/global.h: Remove unused HistorySize field from the settings
+ structure
+
+2010-03-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/draw.c, src/hid.h, src/hid/batch/batch.c,
+ src/hid/common/hidnogui.c, src/hid/gtk/gtkhid-main.c,
+ src/hid/lesstif/main.c: Remove unused "last" parameter from HID
+ "invalidate_lr" method
+
+2010-03-11 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * src/hid.h, src/hid/batch/batch.c, src/hid/bom/bom.c,
+ src/hid/common/extents.c, src/hid/common/hidnogui.c,
+ src/hid/gerber/gerber.c, src/hid/gtk/gtkhid-main.c,
+ src/hid/lesstif/main.c, src/hid/lpr/lpr.c, src/hid/nelma/nelma.c,
+ src/hid/png/png.c, src/hid/ps/eps.c, src/hid/ps/ps.c: Remove unused
+ HID method invalidate_wh() invalidate_lr and invalidate_all are sufficient APIs for our needs.
+
+2010-03-11 Dan McMahill * dan AT mcmahill dot net *
+
+ * data/pcb.desktop.in: Revert "As silly as it seems to group any
+ sort of scientific or EDA" I've pushed really hard to sort this nonsense out, and this patch is
+ not in my opinion suitable for applcation. gEDA, gerbv and PCB all
+ deliberately omit the Education category.. this is not where we need
+ to fix this issue. I strongly object to putting bogus categories into our apps. If
+ Debian want to patch these in, let them. This reverts commit b402a144e19225f0648189bca6fbfad95d046047.
+
+2010-03-11 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/ps/ps.c: Fix an obvious bug in the previous commit.
+ Caught by harry.
+
+2010-03-11 Dan McMahill * dan AT mcmahill dot net *
+
+ * data/pcb.desktop.in: As silly as it seems to group any sort of
+ scientific or EDA software under "Education", this appears to be the
+ standard documented by
+ http://standards.freedesktop.org/menu-spec/1.0/apa.html Noted in
+ patch #2889228 by Ahmed El-Mahmoudy.
+
+2010-03-11 Dan McMahill * dan AT mcmahill dot net *
+
+ * data/pcb.desktop.in: "Encoding" is deprecated.
+ http://standards.freedesktop.org/desktop-entry-spec/1.0/apc.html
+ Noted in patch #2889228
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/Makefile.am: It is gcode/hid.conf not gcode/gcode.conf that
+ needs to be distributed. Also put this file in the correct sorted
+ location.
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * configure.ac: Add missing AC_PROG_MKDIR_P check.
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/Makefile.am: Fix building outside of the source tree with
+ dependency tracking disabled. In that case some directories need to
+ be manually created. Fixes bug #2889226.
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/ps/ps.c: In the rectangle fill code fix a bug in swapping
+ of the coordinates. It doesn't appear that this code path is
+ encountered much.
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/ps/ps.c: Add DSC to the postscript outputs to keep certain
+ windows based postscript viewers a little happier. Patch supplied
+ in patch #2700352 by Bob Paddock.
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * INSTALL, configure.ac: The gcode exporter used libgd so make sure
+ we check for it and document that we need it. Noted by Patrick
+ Bernaud in bug #2967313
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * tests/inputs/Makefile.am: Make sure the simple gcode test layout
+ makes it into the distfile.
+
+2010-03-10 Dan McMahill * dan AT mcmahill dot net *
+
+ * : commit 802576013b5d780cd87caafa31c694279ee2dc4f Author: Dan
+ McMahill * dan AT mcmahill dot net * Date: Wed Mar 10 13:15:26
+ 2010 -0500
+
+2010-03-09 Stephen Ecob * silicon dot on dot inspiration AT gmail dot com *
+
+ * src/djopt.c: Fix Freckles The autorouter sometimes ends a trace just past a pad. The
+ optimizer cuts this trace at the pad center, leaving a tiny trace
+ left over. This tiny trace is a "freckle" and it inhibits mitering.
+ Check for such freckles and remove them.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * tests/run_tests.sh: Make the xy file check actually do something.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * configure.ac, tests/golden/Makefile.am,
+ tests/golden/hid_gcode1/Makefile.am,
+ tests/golden/hid_gcode1/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode1/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode1/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode10/Makefile.am,
+ tests/golden/hid_gcode10/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode10/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode10/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode11/Makefile.am,
+ tests/golden/hid_gcode11/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode11/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode11/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode2/Makefile.am,
+ tests/golden/hid_gcode2/out.back.cnc,
+ tests/golden/hid_gcode2/out.drill.cnc,
+ tests/golden/hid_gcode2/out.front.cnc,
+ tests/golden/hid_gcode3/Makefile.am,
+ tests/golden/hid_gcode3/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode3/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode3/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode4/Makefile.am,
+ tests/golden/hid_gcode4/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode4/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode4/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode5/Makefile.am,
+ tests/golden/hid_gcode5/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode5/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode5/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode6/Makefile.am,
+ tests/golden/hid_gcode6/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode6/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode6/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode7/Makefile.am,
+ tests/golden/hid_gcode7/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode7/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode7/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode8/Makefile.am,
+ tests/golden/hid_gcode8/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode8/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode8/gcode_oneline.gcode.front.cnc,
+ tests/golden/hid_gcode9/Makefile.am,
+ tests/golden/hid_gcode9/gcode_oneline.gcode.back.cnc,
+ tests/golden/hid_gcode9/gcode_oneline.gcode.drill.cnc,
+ tests/golden/hid_gcode9/gcode_oneline.gcode.front.cnc,
+ tests/inputs/gcode_oneline.pcb, tests/run_tests.sh,
+ tests/tests.list: Add testsuite entries to test the gcode export HID
+ and all of its command line options.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/gcode.c: Remove RCSID. We don't use those anymore.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/gcode.c: Change options like 'drill depth' to
+ 'drill-depth' as a space in a command line option is non-standard.
+ Also be consistent with having or not having a "." at the end of the
+ option help strings.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/gcode.c: Remove unix2dos system() call. It was not
+ needed.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/trace.c: Get rid of compiler warnings.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/gcode.c: Get rid of some gcc warnings.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/gcode.c: C++ style comments -> C style comments
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * doc/pcb.1: Fix nroff error. Noted in sf patch #2889227 by Ahmed
+ El-Mahmoudy.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * configure.ac: add the gcode exporter to the export hid list
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * INSTALL: add notes about what libs may be needed (none) for the
+ gcode exporter.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/auxiliary.h, src/hid/gcode/bitmap.h,
+ src/hid/gcode/curve.c, src/hid/gcode/curve.h,
+ src/hid/gcode/decompose.c, src/hid/gcode/decompose.h,
+ src/hid/gcode/gcode.c, src/hid/gcode/gcode.h,
+ src/hid/gcode/lists.h, src/hid/gcode/potracelib.h,
+ src/hid/gcode/trace.c, src/hid/gcode/trace.h: indent to conform to
+ the coding style used by pcb.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/Makefile.am: gcode_lists.h should depend on gcode sources, not
+ png sources.
+
+2010-03-09 Dan McMahill * dan AT mcmahill dot net *
+
+ * src/hid/gcode/auxiliary.h, src/hid/gcode/bitmap.h,
+ src/hid/gcode/curve.c, src/hid/gcode/curve.h,
+ src/hid/gcode/decompose.c, src/hid/gcode/decompose.h,
+ src/hid/gcode/gcode.c, src/hid/gcode/gcode.h,
+ src/hid/gcode/hid.conf, src/hid/gcode/lists.h,
+ src/hid/gcode/potracelib.h, src/hid/gcode/trace.c,
+ src/hid/gcode/trace.h: Apply sf patch 2948711 from Alberto Maccioni
+ which adds a g-code exporter.
+
+2010-03-08 Dan McMahill * dan AT mcmahill dot net *
+
+ * configure.ac: When building the docs have configure check for
+ epsf.tex being in the TeX installation. Problem noted by Kai-Martin
+ Knaak and the method for checking suggested by Peter Clifton.
+
+2010-03-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * configure.ac: Fix test for whether to rebuild the shipped
+ pcblib-newlib library ${top_srcdir} is a Makefile variable, we need $srcdir in
+ configure.ac
+
+2010-03-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * lib/Makefile.am: Fix make distcheck by removing regenerated newlib
+ footprints The generated newlib files need to be removed by make distclean if
+ they are being updated by the build process. If the files aren't
+ being updated, we should not delete them as they were originally
+ distributed with the source tarball.
+
+2010-03-02 Peter Clifton * pcjc2 AT cam dot ac dot uk *
+
+ * tools/Makefile.am: Fix distribution of gnet-pcbfwd.scm in the dist
+ tarball Use the dist_ prefix to make this simple, so o...
[truncated message content] |
