Skip to main content

Changelog

LXT Authoring

AI will now take into account the language variant.


AI functions will now take into consideration what language variant the user currently uses in LXT and not simply use the most common variant. So for example, for Portuguese the AI will take into account whether it’s supposed to use the Brazilian or Portuguese variant, depending wholly on the variant of the language the user currently uses.

Other examples include Chinese, where it distinguishes if it’s simplified or traditional, English if it’s British or American, or German if it’s German or Swiss.

Value slider quiz visual improvements


The value slider quiz element gets two new toggles for the appearance of the selected number, which appears in the middle.

One toggle disables the selected number in the middle initially, so it is only visible once the user moves the slider to make an answer. The other toggle inverts the background around the selected number to highlight it better. The toggles are found in the appearance menu of the element, found by clicking on the green wand.


Default
Initially hidden number
Both toggles on

Visual upgrade to the evaluation in tests


The cross and checkmark icons in the test evaluation are now fittingly colored as the default. The checkmark icon, indicating a correctly answered test question, is now green, while the cross icon, which indicates an incorrectly answered test question, is now red.

Those colors are also theming-dependent and could be returned to the former single-color black. But this also means that through theming, the colors of the checkmarks and crosses can be adjusted to whatever color you wish.

These changes apply to the normal evaluation of a test as well as the extended version. They are also reflected in a SCORM export.


LanguageWire Integration


LanguageWire is a translation software similar to DeepL, which is now also supported for use in LXT.

This gives authors an alternative tool that can be used for translation within LXT.

Bugfix for text elements with drop shadow


We have done a bugfix that changes behaviour of some existing content, which is why we want to inform about this change.

If you are using text elements with drop shadows and the option “Adjust height to row” enabled, you might have noticed that the height of the drop shadow does not adjust to the height of the row.

In most cases this change will improve the layout, please find a few examples below.

However there might be cases where the fixed version might not look good. Usually it will be possible to fix those by disabling the option “Adjust height to row” which should result in a very similar look as the old version. In very rare cases this might not work as well. Feel free to contact us at support@bildungsinnovator.com if you need help to resolve such a special case.

Model for AI image creation in OpenAI updated to gpt-image-1.5


The default model for image creation in OpenAI has now been updated to gpt-image-1.5. The change won’t be reflected in the pricing. Put simply, the expenses will remain the same as they were with the older model, but the model needed to be updated.

The change comes because the model DALL-E 3 won’t be supported after May 12th of this year.

The change only applies to direct OpenAI users and not Azure users, since they have a different model deployment structure.