News archive
JPEG Saver 4.14
JPEG Saver 4.14 doesn't contain anything truly amazing, but it does have a new item for adding to the display: an analog clock. I first played with an analog clock item back in 2010, drawing the hands as Direct3D meshes. It worked, but it looked awful so I shelved the code. The analog clock in version 4.14 works the same way that the text clock does - drawing to an off-screen bitmap and then pasting it on top. It looks a lot better and makes it easier to preview the clock in the item configuration dialog.
I thought about adding a shadow option to the analog clock, but then I realised it was already easy enough to do using JPEG Saver's items list. Set up your clock the way you want it to look then create a copy, set the colour of the copy to grey (or whatever colour you want the shadow to be), change the offset by a few pixels and then drag the copy above the original clock in the items list. Items are drawn in the order they appear in the list, so the shadow should be above the main clock to be drawn first.
This version also adds some extra format tokens for the “Image info” item, for displaying some properties that can be updated in Windows Explorer. These are the title, subject, keywords (or tags), comments and author. The properties are actually stored in the Exif metadata, but are not part of the Exif standard (which is why I hadn't thought to include them before).
SVGGraph 2.18
In previous versions of SVGGraph there were labels for some bars and labels for pie slices, but that was about it. They used different code, had different option names, and supported different styles and things. Version 2.18 attempts to bring the labels together, support labelling all the data items, and provide more options for what the labels look like.
This example StackedBarGraph shows some of the new features. The labels are
displayed using a data_label_type of “bubble”, with the
data_label_filter option used to only create 1 in 3 labels for
the second dataset and all labels for the first. What do you mean, you
can't see the labels for the first dataset?
You can't see the labels for the first dataset because the
data_label_fade_in_speed and data_label_fade_out_speed
are both set to allow the labels to fade in and out, and the
data_label_click option is set to array('hide','show').
This hides the labels of the first dataset and shows the labels of the second from
the start, but they are all still there.
TagCanvas 2.6.1
TagCanvas 2.6.1 is another quick bug-fix release, just fixing one bug that could be quite annoying. The new canvas tooltip supported by version 2.6 would display its div at the end of the page when TagCanvas started up. This version fixes that, and nothing else. So if you are not using div-based tooltips, this new version should not act any differently to version 2.6.
In other news, I've had a growing number of people contacting me about problems with using TagCanvas on the Android browser. There appear to be quite a few issues with using an HTML5 canvas on the Android browser, with all sorts of workarounds that may or may not solve the problems: Google search for “android browser canvas problems”.
I don't have anything that will run the Android browser, so I haven't made any attempt to solve these problems in the TagCanvas script. My advice for now is to search the web for similar Android canvas bugs and see if there are any changes that you can make to your HTML or CSS that will fix things for your application.
TagCanvas 2.6
TagCanvas 2.6 really only adds one new feature, with a few options to go with it. The example cloud below shows the new mixed text and image tags in action.
A piece of paper with a clip on it
Speech bubble
A calendar
A monitor
A box with tape on it
A shiny carrier bag
A clock
Whirrr... click
You have mail
The first new option is imageMode and determines how TagCanvas
deals with text and images found in the links it uses to create the tags. The
default behaviour is to act the way it always used to do, displaying an image
if there is one, or the link text otherwise. The imageMode option
allows you to choose
to only display images (imageMode = 'image'), only display text
(imageMode = 'text'), or display text and images on a single tag
(imageMode = 'both'). The default fallback option is used when
imageMode is null.
The imagePosition option specifies where the image goes in
relation to the text, and imagePadding is the amount of space
between the image and the text. For positioning the text and image within the
bounding box of the tag, there are imageAlign,
imageVAlign, textAlign and textVAlign
options. Apart from textAlign, these options are only useful when
mixed image and text tags are in use.
SVGGraph 2.17
Quite often changes happen to SVGGraph, or any of the software I release, due to requests from users. Sometimes users send me code that they have used to improve the software themselves - version 2.17 of SVGGraph includes both of these types of update.
The most obvious change in this version is the addition of a load of new functions for setting the colours used to draw the different bars or pie slices. I've been unhappy with the way SVGGraph deals with colour selection for quite a long time, but I haven't done anything about it until now because it worked well enough. Recently someone sent me through some code that added support for using a range of colours, and I liked it enough to incorporate it into the system. It meant overhauling the way that colours are picked, and it also gave me the chance to add a function for setting different colours for each dataset on a multi-dataset graph. This example shows two different colour ranges on one stacked bar graph.
The old $graph->Colours() function still works the same way,
and setting $graph->colours directly still works too. Setting
individual bar colours using structured data will still override any colour
settings as well. The new colour functions are listed on the
functions page, and are demonstrated on the
colours, gradients and patterns
page.