Removed “Force Greyscale” option when saving images, and set it to always output greyscale
Added “Font…” to the edit menu
Added “+” to the center of the selected area when moving it
Fixed pictures not being released after they have been loaded
Fixed brightness/contrast not being applied to the colours when saving as a colour image
Fixed saving as html/colour text with ansi and other non-standard characters
Small improvement to outputted css
First, a note to people still downloading the previous version: STOP IT. That is what happens when you link directly to a file instead of the download page.
Things I forgot to mention:
- The save as 8-bit colour output is currently using the inbuilt gif compression to reduce it down to <256 colours. It isn’t very good at it, so the output will be much better if I ever get around to writing my own octree quantization code (or find some gpl code to do it).
- To get the best results, check the invert output button so it has white text on black, then “save as colour text”, and save as 8-bit XHTML. I’ve been meaning to write some tutorials for a long time.
- If you want to save it as an image, to print or whatever, I would suggest the Save As Image plugin for firefox. Then just load the html into firefox and you rightclick to save the page as a png.
Changes in this version:
Added “Save as Colour Text” to output as html and rtf files with the Color information added
Added menu links to the report a bug and feature request pages
Changed batch conversion form from contextmenus to contextmenustrips
Added saving as html files
Now don’t say I don’t do anything for you people.
Either sourceforge’s stats are messed up, or the program was downloaded 10,068 times yesterday (up from the usual ~150 a day). Since it says that only 2838 people have downloaded the latest exe, I think it’s a bug.
No promises, but I’ve just read an article on colouring richtextbox text that I might be able to adapt for a certain oft-requested feature. I tried something very similar using the same method, and it looks like I gave up just before I found the answer. If it updates with a good speed everything else is easy; we’ll see…
I’m manually restoring everything, so things might be a little weird until I’m finished.
Edit: Ok, everything I could recover is back exactly how it was. I think there were a couple of comments that google hadn’t cached and are gone forever, but since one of them was a “do color ascii art!” I’m not that bothered.
I will be making regular backups from now on. 😳
It’ll be at least a few more weeks before I get around to finishing the edit settings form, so here is everything I have so far:
Ported code to .Net framework 2.0
Menu and toolbar changed to MenuStrip and multiple ToolStrips (all rearrangable)
Replaced native measure and draw text code with TextRenderer functions
Added “???? ” ansi ramp and valid characters – Select “Plain Text (Unicode)” as the type when saving the text
Changed “Close without saving the output?” to “Save the output before closing?”
Added menu icons
Replaced context menus with context menu strips
The program has been converted to .net 2.0. The menus have all been replaced with draggable menustrips, and I’m using the Silk icons set for most of the items. The exe size has dropped by about 80k.
All but one of the native winapi functions have been replaced with .net functions (for the last I need to either find a way to check the text metrics in .net or an accurate method to detect if the font is fixed width).
About having accurate output for all the fonts, and this is a real kicker, they were always accurate. The richtextbox control doesn’t display them correctly. Even in the older version, if the text is all messed up and you save it as an image it will save correctly. Using ANSI characters, some of the fonts display the characters in the control as a square, but still save the output correctly. We will just have to live with it until I sort out a way to fix the output, or write my own text control.
I want to add one more thing before I release the new version (a form to edit the settings), so unless I get distracted by another project it’ll be out sometime this week.
Boring technical stuff that I couldn’t find the solution for online and had to workout for myself, part 1
You have a toolStripContainer with a MenuStrip and a set of ToolStrips. After spending a couple of minutes neatly lining them all up how you want, you save and compile and everything’s good. Except that every single time you reload the form, or just switch to the code then back to the designer, all the controls are rearranged to fit one per line. EVERY SINGLE TIME.
Here is how to set the control positions at runtime:
// First clear the container
// Then add the strips one by one in _reverse_ order using:
// Finally add the menustrip:
I know I’ve said I wouldn’t release a new version until the ansi/non-standard characters problem is fixed, but I’ve been unable to find a solution that doesn’t break the api calls needed to accurately measure the fonts, so I gave up and worked on something else for a month, then decided to just go ahead and release the code as it currently stands.
This will be the final version released… using the .net 1.? framework.
After having another look at the TextRenderer class in 2.0, I worked out why the sizes weren’t matching up to what they should be, and I’m fairly certain that I can use it in my program. More importantly, it can handle any characters, and should (theoretically) allow the program to work correctly for every possible font with any character.
Added right-click menu items to show a dialog for changing the selection area’s fill and border colours
Added saving/loading of the selection area fill and border colours from the settings file
Added confirmation dialog for closing an image
Added confirm on close to the settings file (default: true)
Fixed double-click on a selected area loses the selection area
Changed clicking outside the image to not remove the selection area
Moved image and text brightness/contrast settings into a sub-category
Added variables and settings to specify whether to load image and text brightness/contrast and levels from the settings file (default: false)
Added variables and settings to specify whether to load the output size from the settings file (default: true)
Let’s see, I spent a week or so writing a MVC framework in php, then I found Code Igniter and decided to use that instead.
Just as a test, I’ve remade the old ascgen website using the framework, and recreated the design totally in CSS. It doesn’t look that much, but it validates as XHTML strict and the examples gallery is all created on the fly from a mysql database. The framework is really easy to use; most of the time was spent on the css. When I get around to doing the bigger web project I’ve got in mind I’ll have a good idea about how to do it.
Not much progress on this program. I’ve worked out exactly what the problem with non-standard characters is, but fixing it isn’t straightforward thanks to the native methods I had to use to get accurate measurements of the fonts.
Worked around the main problem with resizing the output images when Cleartype is on by converting the output to greyscale before saving it. People without XP, or who have cleartype turned off, can ignore the strange new and seemingly useless checkbox.
I’ve also fixed the problem with saving batch conversions with a variable width default font, hopefully for real this time.
Changes in this version:
Fixed phantom colours problem when saving as images with Cleartype on:
– Added code for converting the output images to greyscale
– Added force greyscale checkbox to the Save as Image… magnification dialog
Properly fixed the batch conversion problems with variable width default fonts