FreeMind Version 1.1.0 Beta 1

[Christian Foltin]

Dear FreeMind users,

the first beta of 1.1.0 is available under

https://sourceforge.net/projects/freemind/files/freemind-unstable/1.1.0_Beta1/

The highlight is the option to support HiDPI screens under windows/linux (as Mac had this already). Moreover, linked maps are shown with a preview if you place the mouse over the link.

The new functions are not thoroughly tested. This is a beta version. Backup your data first.

Hope you like it,

Best regards, Chris

The complete list of changes:

~~~~~

Changes made in FreeMind 1.1.0 Beta 1

• Fixed https://sourceforge.net/p/freemind/discussion/22102/thread/cf032151/?limit=25#c631
• new feature: HiDPI support (for windows and linux, mac was already there)
• New feature: font color combo box

===============================
Changes made in FreeMind 1.1.0 Alpha 2
===============================

• Removed lots of the attribute functionality. This has been re-developed.
• MVC-pattern more enforced.
• Fixed: Bug 1181: Folding symbol is invisible on black background
• Search dialog on geographical maps accepts coordinates in the form "lat lon" or "lat lon;lat2 lon2;..."
• A reverse lookup is then performed unless you set do_reverse_lookup_on_lat_lon_search to false.
• Format dialog has example nodes that demonstrate the changes.
• New search&replace dialog shows tooltips of the node and note text.
• Enhanced the calendar widget to show 12 month at once.
• New feature: standalone map server created.
• new feature: icon toolbar can have multiple columns (see properties->appearance).
• fix for "Encrypted nodes can't be saved".
• Update to JMapViewer 1.03
• New method to create graphical and local links: copy nodes and then paste them as graphical or local links.
• Geographical maps don't need local png files anymore. They are directly referenced using HTML code.
• Paste of structured (HTML) data recognizes the structure and reconstructs it as hierarchical nodes.
• New feature: on save, thumbnails are created and displayed as tooltip on mindmap links to other mm-files
  ~~~~

[jzelno]

I hope that version 1.1.0 will include a fix to allow the LaTeX plugins to work?

[GeertW]

Hi Chris,

I note the mindmapcontroller editAttribute class has been removed - is there an alternative way you'd recommend to edit attributes or will it be back?

Best regards,
Geert

[Tony Davis]

Great program Chris - just one tiny, tiny update I'm desperate for. All of the secondary nodes (nodes of nodes/tier 3) on the right of the central node 'align left' perfectly. But the secondary nodes on the left of the central node won't align right. Of course, I've selected 'align right' in the long node editor and checked that there are no additional spaces after the text, but this doesn't make any difference. I've also tried using 'automatic layout', but this doesn't get it right either. When you're trying to make your work look spot on, this issue is a real pain.

Forgive me if this is user error. If it is, then perhaps it's just the user interface that could be tweaked to stop the frustration.

Very many thanks in anticipation.
Tony

[Christian Foltin]

Hi, do you have a short screen shot about what you mean? TIA, Chris

[Marcio de A Braga]

Maybe this one can represent what he described.

[Christian Foltin]

Hi,

do you mean the old LaTeX plugin? It was abandoned.
To which one do you refer?

Br, Chris

[jzelno]

I am talking about the old latex plugin, but if there's a new one that works, I'd be happy (your answer implies that there's one that does work). I don't understand why the plugin isn't supported... It used to work before v1.0. The latex menu shows up, but it didn't respond. I'm aware of 2 plugins and neither works. It seems odd that those specific plugins, and only those plugins, don't work.

[dodrg]

Hi, thank you for the HIDPI solution. This workaround came just at the right moment.
But are you working on a real solution for display resolution independent output?

To analyse and structure the problem, I am willing to offer some of my time as tester and perhaps more with Windows 8.1 and Linux as available operating systems. Unfortunatelly Java programming is not my business, but system analysis and some scripting experiences are available plus some more probably usefull skills. As the relevant hardware to mention is a 4k display with about 280 dpi.

[Christian Foltin]

Hi,

what do you mean with "real solution"? From my perspective, it has to be built in into Java itself. It can't be the task of a java developer.
But if there is something basic missing, please tell me.

BR, Chris

[dodrg]

Hi Chris,

sorry that I did not differentiate between the tasks of the OS, Java and the program FreeMind, I'm not a program developer. I am aware, that some resources required by a Java program are managed by the OS, some by Java and some the developer himself has to add. But currently, I cannot sort out fine enough, which resource provider is the owner of a specific task. - I'll try to improve myself.

From a theoretical point of view IMHO the best way to achieve an resolution independent output (a task for the operation systems), is to calculate a scale value respecting the following factors:
    - the physical screen size
    - the screen resolution
    - the user-customizable distance from the display

The result would be a radian measured value that would allow the applications to display their content objects in an appropriate scale, independently of the device (smartphone, computer monitor, multi-monitor presentation wall). This also would give the font unit "pt" its unified usability back.

Obviously, the reality of most current OS is far away from such an abstraction level for the screens. So we have to deal with the reality: The OS User settings will give us a hint which sizes are regarded as comfortable.

I think the font size of the menu text in the OS user profile should be the "100% base" used by HIDPI. If this is a wrong guess or personal preferences differ, using the HIDPI scaling factor can help to fix this.

I don't know how difficult it is to realize such a behavior. A hint that this should be possible within Java shows the attached screenshot:
    - menu and menu entries and icons are formated by Java/FreeMind
    - menu entry keyboard shortcut hints and window title bar are formatted by OS (Windows)

I've been astonished to notice that obviously parts of the menu content is handled by OS, while most of it is handled by Java. - So the key question is: Does Java exports the formatting properties of the menu entry shortcut hints to the program developer?

A second issue with HIDPI scaling are the icons. While font scaling is fine, the icons used in FreeMind are 16x16 pixel PNG-bitmaps. By up-scaling with HIDPI they appear pixelized and unpleasant. To give a try, I exchanged the icons of the left icon bar by up-scaled versions (32x32 pixel) in the "freemind.jar". The result:
    - the icon bar object also scaled up to double size and the the result remained pixelized.
    - the icons within a FreeMind map doubled their size too. As the font-size now did not optically match to the icon size, this has been an unpleasant result.

=> The displayed size of the icons object should NOT be defined by the pixel size of the icon. I think the size dependency should be
    - "HIDPI scales the icon object size"
    - "The icon object scales the used image to its required size"

Realizing a solution that is independent of the icon's pixel count would open the path to icons-sets in the SVG vector format. They would scale as seamessly as the fonts and make it unnecessary to provide huge multiple sized bitmap icon sets.
(The usage of icons drawn as vector graphics like SVGs is quite common in Linux)

How well Java supports the realization of these ideas I'm unaware. So hopefully it is possible and hopefully you like these ideas.

BR,
-do-

[Christian Foltin]

Hi,

I've already added the scaling of the accelerator key font. It will appear in the next beta version.
To use 32x32 pixel graphics, you can already add files named 'format-font-size-more_32.png' for a 32x32 icon for 'format-font-size-more.png' to the freemind.jar.

If you can send a set of 32x32 scaled icons for the left toolbar, I'll add them to the sources. But for me it is a very tedious work to find all the icons in that scale.

BR, Chris

[dodrg]

Hi Chris,

I've already added the scaling of the accelerator key font. It will appear in the next beta version.

OK, if it is possible to manipulate the accelerator key font it also should be possible to evaluate the original size (and other properties) of this font provided by the operating system. Probably also is true for the font used by OS for the menus. - Using this OS-/user-defined font size should be an appropriate approach to determine a useful default scaling for the 100% scale values used by HIDPI. A program can expect, that the defined system values have useful dimensions. By using this "calibration workaround" in most cases a manual HIDPI value tuning should not be necessary any more.
This workaround should be fine until the OS have settled their problems with high resolution displays.

=> Do you also think, this import of user-settings to FreeMind might be useful and worth to implement?

To use 32x32 pixel graphics, you can already add files named 'format-font-size-more_32.png' for a 32x32 icon for 'format-font-size-more.png' to the freemind.jar.

I'm Still testing this. The current results are not as straight forward as implied.

Where do you prefer bug reports related to HiDPI: Here in this threat or in your bug tracking tool at https://sourceforge.net/p/freemind/bugs/ ?

If you can send a set of 32x32 scaled icons for the left toolbar, I'll add them to the sources. But for me it is a very tedious work to find all the icons in that scale.

My icon-set for the testing is a quick 'n dirty upscale of the 16x16 icons. That's nothing to spread. - But I would agree to provide a set of higher scaled icons.
For the first step, I would search the web for an appropriate existing icon-set. If necessary later on I would extend the icon-set found to get all necessary icons in a homogenous style.

Most useful free icon-sets on the web use the Creative Commons license CC-BY-SA. So an entry honoring the author of the icons is required, also the note, that these icons and their derivatives can only be reused using the same license. - Quite similar to the GPL.
=> Is the CC-BY-SA license of the icon-set OK for the FreeMind project?

Icons created by me will have the same license as FreeMind.

BR
-do-

[Christian Foltin]

Hi,

look at MindMapActions:

void setAttribute(MindMapNode node, int pPosition, Attribute pAttribute);

/**

 * Inserts a new attribute at a given place of the attributes table.
 * To insert an attribute as the last item, {link {@link #addAttribute(MindMapNode, Attribute)} instead.
 */
void insertAttribute(MindMapNode node, int pPosition, Attribute pAttribute);

/**

 * Inserts a new attribute at the end of the attributes table.
 * 
 * @param node
 *            to which the attribute is added
 * @param pAttribute
 *            itself
 * @return the index of the new attribute.
 */
int addAttribute(MindMapNode node, Attribute pAttribute);

/**

 * Removes the attribute at the given position
 * 
 * @param pPosition
 *            the position to delete.
 * */
void removeAttribute(MindMapNode node, int pPosition);

BR, Chris

[Mark Aufflick]

I use editAttribute in groovy scripts attached to nodes, but am not all that familiar with the Java source, what's the best way for me to figure out how to do the following:

I have a number (I've recursively found all numbers in a node's children and summed them), I then want to show it in an attribute with key "sum". I used to just:

c.setAttribute(node, "sum", (String)sum);

So it sounds like now I need to either find an existing attribute object with key sum, or made a new one and add it to the node, then update the key/val of that object? What docs/.java files should I be looking in?

Thanks.

[Christian Foltin]

Hi,

as several keys may have the same text, the API is changed.
I admit, that it is more tedious, but you get more flexibility.
So, to set a certain attribute, search for it in order to be replaced (by setAttribute), and if you don't find it, add it (by addAttribute). Perhaps, you write a helper method (in Tools?), that does exactly that, and send me the patch, then it is easier for all.
HTH, Chris

[Mark Aufflick]

Thanks. I've been looking at this, but it seems lots of other api has changed too, for example there's no getAttributes() on a node and none of the documentation seems to have been updated (fair enough, it's in beta). What .java should I be looking at to see the relevant functions?

[LeO]

Great that now the Toolbar could be expanded to mutliple columns...... long awaited

[Milavon]

Hi, thanks for this amazing version

I just don't understand exactly what this feature means:
(* Paste of structured (HTML) data recognizes the structure and reconstructs it as hierarchical nodes.)

because I thinks it is already existing in the current version 1.0.1

Can anyone give more details about this feature, what this feature mean, is not it already included in 1.0.1 and if not, what it is the difference? because when I paste HTML Data to Freemind 1.0.1, it is already will be pasted as hierarchical nodes.

Thanks

[Christian Foltin]

Hi Milavon,
AFAIK, pasting HTML before created one node with all contained. Now, if you paste HTML, then it is tried to be cut in hierarchical pieces. For example, if you have

• text1
  • subtext1.1
  • subtext1.2
• text2
  etc., then it will be pasted as two nodes, one of them has two more grandchildren.

OK?

BR, Chris

[Milavon]

Thanks Chris,
I thinks it is not a big issue to argue about, it is only about having a better understanding.

I tried alot of pasting Html Text (of different elements) for both Freemind 1.0.1 and Freemind 1.1.0 Beta 1 and the result is the same for both versions,
there is no difference at all.
(Your example, Html Tables, Lists, Html Structured Elements,..) all are pasted in the same way. (as hierarchical nodes) whether we pasted them into Freemind 1.0.1 or 1.1.0 or even into Freeplane

Anyway, this feature may be supposed to be attributed to an older version of Freemid, not to this version.

Thanks

[Christian Foltin]

Hi,

try again for M$-Word Lists, and you'll see certainly a difference.

BR, Chris

[Mark Aufflick]

Awesome, love that it launches on MacOS without the old legacy Java installed now.

With the new attribute changes, is it still possible to display the attributes inline in the tree view? I prefer everything there and hide all toolbars/panes etc. but have scripts that show results in attributes.

[Christian Foltin]

Hi,

all documentation for groovy actions are described in freemind.modes.mindmapmode.actions.MindMapActions. To which documentation do you refer?

BR, Chris

[Matt Steven]

On larger mindmaps, I'm getting out of memory errors that seem to be preventing the saving of any maps that are open in the new beta. The logfile says this after I try to close and save other open maps.

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap spaceNov 11, 2015 5:27:31 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:27:31 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 4, pIndex 0

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap spaceNov 11, 2015 5:27:52 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:27:53 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 4, pIndex 2

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap spaceNov 11, 2015 5:27:59 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:27:59 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 4, pIndex 3

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap spaceNov 11, 2015 5:28:23 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:28:24 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 4, pIndex 1
Nov 11, 2015 5:28:25 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:28:25 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 3, pIndex 1

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap spaceNov 11, 2015 5:28:31 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:28:31 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 3, pIndex 2

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap spaceNov 11, 2015 5:28:35 PM freemind.modes.common.listeners.MindMapMouseWheelEventHandler <init>
INFO: Setting SCROLL_SKIPS to 80
Nov 11, 2015 5:28:35 PM freemind.controller.Controller numberOfOpenMapInformation
INFO: number 3, pIndex 0

STDERR: Exception in thread "AWT-EventQueue-0" 
STDERR: java.lang.OutOfMemoryError: Java heap space

(etc). System memory is plentiful.