root/morphix/trunk/partitionmorpher/ChangeLog

2004-10-07 Mantas Kriauciunas <mantas@akl.lt>
        * src/main.c: Check if there are at least one sufficient size Linux 
        (and swap) partition and suggest to use them.
        * src/wizard.c: 
        * src/ui.c: Replaced WarningWindow with gtk_message_dialog_new()
        and moved warnings from wizard.c to callback.c and main.c
        * src/wizard.c: Added ability to create separate /home partition in 
        Wizard mode.
        * src/wizard.c: Wizard uses existing swap partition if selected hard
        disk alredy contains sufficient size swap partition.
        * src/pm_fswrap.c: Implemented progress dialog in do_mkext3()

2004-09-21 Mantas Kriauciunas <mantas@akl.lt>
        * src/wizard.c: Improved Wizard mode to usable state ;)
        * src/main.c: Fixed a bug with parameters.
        * src/ui.c: Fixed a crash bug in show_busy_cursor()

2004-01-11 Alex de Landgraaf <alextreme@xs4all.nl>
        * src/callbacks.c: Implemented progress and error dialogs for 
        resizing and deleting. 
        * src/pm_wrap.c/h: made wrapper as to enable non-parted commandline
        tools, however timer implementing for these has yet to be done
        (impossible?). Dynamic filesystem times. Support of using 
        do_mkpartfs with parted and ntfs/ext3/xfs. Support of do_resize
        with parted and ntfs.
        * src/pm_fswrap.c/h: all non-parted calls, currently mktfs, mkfs.ext3,
        mkfs.xfs, ntfsresize. todo: xfs_grow, resize2fs for ext3
        * src/pm_libparted.c: see callbacks.c, redesigned calling of do_ 
        functions
        * src/ui.c: all non-callback gtk code to be done in here, mainly
        old callback.c code
        * some glade changes
2003-10-20  Elijah Newren  <newren@math.utah.edu>
        * src/main.c: Workaround a Parted bug.  Parted can't call the
        timer_handler with the appropriate amount of time left if stdout
        is redirected at the command line.  So, we redirect stdout to
        stderr in main.c to work around this.

        * TODO: Removed items that have been now taken care of and added
        some notes about the testing I did.

2003-10-12 Alex de Landgraaf <alextreme@xs4all.nl>
        * src/*: Heaps of updates. No, I'm not going to list them all
        See 1st law of WAL

2003-09-15  Elijah Newren  <newren@math.utah.edu>

        * partitionmorpher.glade:
        * src/interface.c: Make CreateWindow, ResizeWindow, DeleteWindow,
        ConvertWindow, and MoveWindow modal so that the user can't launch
        multiple of these windows at a time (could cause nasty problems if
        the user had a window open for resizing a partition and also
        converting the same partition).  This should also probably be done
        for the ApplyWindow, RevertWindow, DialogWindow, and perhaps
        TableWindow.

        * src/callbacks.c: Make the MainWindow also show the busy cursor
        whenever sub-windows do (since one probably should not be able to
        interact with the main window when the others are busy doing disk
        operations...).

2003-09-14  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c:
        * src/main.c: Enable startup notification to work properly.

        Note that startup notification requires both a launcher and a
        launchee, so in order to test this from the command line, one
        needs to run the test-launcher program (which is part of the
        desktop-notification tarball found on www.freedesktop.org) with
        partitionmorpher as the program to launch.  [Adding
        partitionmorpher to the gnome-panel menu would be an alternate way
        to test, but if one isn't logged in as root in a gnome session
        (which is hopefully the case), then one would also have to do
        whatever trickery is required to make partionmorpher be able to
        run as root when launched]

        The changes made here may be unnecessary in the future;
        startup-notification launchers are the ones required by the spec
        to do most of the work and they are required to be pretty smart.
        The only reason additional code is needed at all inside
        PartitionMorpher at this time is that we start displaying the main
        window a couple seconds before we are actually finished starting
        up.

2003-09-13  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c:
        * src/pm_libparted.c: Added a busy cursor when performing
        operations that take longer than a second.  Currently only added
        for Create & Delete windows.  Make the app seem much more
        responsive, though (in the Delete Window, the operation would
        always take about two seconds, but the user wasn't given any
        notification that the program was acting on their input)

        * TODO: Updated with some more bugs and stuff

2003-09-12  Elijah Newren  <newren@math.utah.edu>

        * TODO: Updated (I'm really starting to think I should break out
        the known bugs into a separate file...)

2003-09-12  Elijah Newren  <newren@math.utah.edu>

        * src/pm_libparted.[ch]:
        * src/callbacks.c: Only display debugging info if
        PRINT_DEBUGGING_INFO is defined (currently located in
        pm_libparted.h).  The debugging info is still potentially useful,
        but sometimes I don't want to see it.  Also, adding this flag
        means that it will be easier to clean out all these messages when
        PartitionMorpher is ready to ship to everyone.  I've probably
        missed marking numerous debugging messages, but I'll add them as I
        find them.

2003-09-12  Elijah Newren  <newren@math.utah.edu>

        * partitionmorpher.glade:
        * translation:
        * src/callbacks.[ch]:
        * src/interface.c: Since extended partitions can't have a
        filesystem, inactive the filesystem selector in the CreateWindow
        whenever the Extended partition radiobutton is active.  Also,
        update the tooltips for the Extended and Logical partition types.

        * TODO: Updated, specifically with info about how
        on_CreateTypeRadiobutton_clicked is called twice for every click.
        (I don't understand why it does this, but it doesn't hurt
        anything)

2003-09-12  Elijah Newren  <newren@math.utah.edu>

        * partitionmorpher.glade:
        * src/interface.c: Closing certain windows with the X button
        (located in the upper right corner of each window) was causing
        PartitionMorpher to either segfault, hang, or not respond when
        attempting to show the window again.  I tested and fixed this for
        these windows:
                CreateWindow
                ResizeWindow
                HelpWindow
                DeleteWindow
                ConvertWindow
                MoveWindow
                ErrorWindow
                ProgressWindow
        The fix was to set the signal handler for the delete_event to
        gtk_widget_hide (all but one case) or gtk_true (in the case of the
        ProgressWindow, which should not be able to be hidden by the
        user).  Other windows (Apply/Revert/Dialog/Table) may still need
        this fix.

2003-09-08  Elijah Newren  <newren@math.utah.edu>

        * TODO: Updated

2003-09-08  Elijah Newren  <newren@math.utah.edu>

        * partitionmorpher.glade:
        * translation:
        * src/callbacks.c:
        * src/interface.[ch]:
        * src/main.c
        * src/pm_libparted.c: Add a progress bar to provide visual
        feedback during long operations.  Currently, this has only been
        tested and worked in with creating partitions, but it shouldn't
        require many changes for other operations.  However, if we
        implement NTFS creation/resize (and perhaps other operations) via
        external tools, we will need to change the display from a
        percentage progress bar to an indeterminate moving back and forth
        bar.

2003-09-08  Elijah Newren  <newren@math.utah.edu>

        * src/pm_libparted.c : Small change to make MIN_FREESPACE = 1024
        k, not 1000 k.  This is an issue because users are only allowed to
        select partition sizes in terms of MBs (which, perhaps more
        appropriately would be labelled MiBs).  Thus, on one partition on
        my laptop, which is 1913.99 MB (isn't it nice to have a
        pathological laptop that provides nice corner cases?), I can only
        choose to create a partition of size 1913 MB.  The MIN_FREESPACE
        corresponds to about .98 MB, and thus my partition was left with
        an empty space at the end of almost 1 MB in size.  This would mean
        that although I thought I had used all the space on my disk, a
        view of the sectors would show that I had left gaps.

        Note: This is not an issue for Parted or QTParted (both of which
        use 1000 k for this parameter) because they both require/allow the
        user to specify to higher precision the size of their partition.
        We only allow integer sizes in the interface.

2003-09-08  Elijah Newren  <newren@math.uath.edu>

        * src/callbacks.c:
        * src/pm_libparted.[ch]: Remove duplicate (unused) code

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: Change the display in the main menu so that
        intead of showing "Size" we show "Size (MB)".

========== I was lame when starting and didn't keep a            ==========
========== proper ChangeLog.  So the dates below are inaccurate. ==========

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * Makefile:
        * Makefile.in:
        * aclocal.m4:
        * config.log:
        * config.status:
        * configure:
        * autom4te.cache/:
        * po/Makefile:
        * po/Makefile.in
        * po/Makefile.in.in:
        * src/Makefile:
        * src/Makefile.in
        * src/partitionmorpher: Remove all these files from CVS as they
        are automatically generated build files.

        * .cvsignore:
        * po/.cvsignore:
        * src/.cvsignore: Add these files to make sure that files like the
        above don't get added to CVS.
        
2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * AUTHORS, README: Added some basic content
        * TODO: Created and started it with some basic things

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: Remove first column of main window
        (corresponding to PART_NUMBER) from the interface; this was
        basically an artifact of our implementation and provides no useful
        information to the end user.  It was helpful for debugging but is
        merely clutter otherwise.

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: For extended partitions or when choosing no
        filesystem, call do_mkpart

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/pm_libparted.[ch]: do_mkpart != do_mkpartfs, so don't
        require a filesystem type to be passed.  NOTE: From looking at the
        original parted.c file, I'm not so sure this function can be
        trusted.  Alex made this code way better (it used to be absolutely
        hideous), but I'm still a little unsure about it all, so I don't
        know if my changes are either correct, or enough.

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/pm_libparted.c: Make sure a message is printed when there's
        an error.  Also, print error messages to stderr instead of stdout
        so it can be captured by my stderr redirection

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c:
        * src/pm_libparted.c: Simple fix to keep the same selected
        partition instead of always resetting to the first

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * partitionmorpher.glade:
        * translation:
        * src/callbacks.[ch]:
        * src/interface.[ch]:
        * src/main.c: Add an error dialog for when Parted aborts
        due to an error.  Use it in the case of creating partitions.

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: Add stderr redirection in order to capture
        Parted error messages

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: Get rid of unused variables and unused code;
        NOTE: The add_swap variable wasn't ever being set to true, so that
        meant I removed the change_part_flag call (dead code removal).
        I'm a little unsure this is the right thing to do (the code was
        probably there for a reason), but the current version after this
        change seems to work for me.

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: In CreateOk callback, make sure the switch
        statement numbers match the values reported by gtk.  (Apparently,
        line separators get a number)

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: Be careful about the size of *printf format
        flags.  This fixes at least one bug in the printing of debugging
        messages.  It may fix other stuff as well.

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: Minor clean-ups

2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/pm_libparted.c: Not incrementing the partition count
        (part_count) for freespace partitions caused early termination of
        loops in callbacks.c (e.g. update_partition_list). So, this patch
        always increments part_count
        
2003-09-05  Elijah Newren  <newren@math.utah.edu>

        * src/callbacks.c: My previous patch (commited by paul_c)
        introduced the variable num_free_partitions, which turned out to
        be identical to the already existing part_freespace_count.  This
        patch uses part_freespace_count instead and removes
        num_free_partitions.

=== As I don't keep changelogs personally, check the CVS logs for info ===
=== When thy are in a state of CodeRage, thy shall not document!       ===
===  -- 1st law, WAL                                                   ===
        
2003-00-00 Alex de Landgraaf <alextreme@xs4all.nl>
        * *: Creation of pm, heaps of bugs, partly working