To translate a GeoGebra activity quickly, begin by protecting the construction before editing the language. A GeoGebra activity translation may involve the worksheet title, instructions, questions, button captions, dynamic text and labels in an interactive applet. The mathematics must still respond the same way when a learner drags a point, moves a slider or submits an answer. A fluent caption is no improvement if it describes a different object from the one being displayed.
This guide shows how to localize a GeoGebra worksheet or shared classroom activity without breaking dynamic objects. It addresses related search questions such as editing GeoGebra activity text, translating a dynamic GeoGebra applet, preserving slider values during localization and testing interactive mathematics resources in another language. The main principle is to distinguish words intended for readers from names and expressions used by the construction engine.
The reader’s job is one complete conversion: make a copy of an existing interactive maths activity for a defined target-language class, translate its learner-facing content and verify that the interactive behaviour and mathematical conclusions are unchanged. This is not a general introduction to drawing geometry in GeoGebra, and it is not instructions for translating GeoGebra’s entire application interface. We will use a circle-and-radius learning activity as a running example, then transfer the checks to graphs and probability.
The central distinction: visible language and mathematical dependencies
A GeoGebra resource can combine ordinary explanatory text, applets, images, videos and questions on one page. Inside an applet, the visible words may be static labels or text objects assembled dynamically from values of other mathematical objects. A user can read both as a sentence, but the software treats them differently. Translating a static sentence is like editing a paragraph; translating a dynamic text expression is closer to editing an equation with quoted language pieces.
Imagine a slider r controlling a circle of radius r. A learner sees “Radius = 4 cm,” and the label changes when r changes. If a translator replaces the entire expression with a target-language sentence containing the current value 4, the display becomes frozen. The circle still grows and shrinks, but the text no longer reflects it. This is a classic localization regression: the surface looks correct in one screenshot and wrong a moment later.
Speed comes from classifying each item before translating. Mark ordinary activity instructions, static in-applet phrases, dynamic text, variable names, command expressions, scripts, captions, image-embedded labels and student response prompts. A translator may freely change natural-language literals within the authorized scope; object references, algebraic definitions, geometric dependencies and evaluation conditions must remain functional. That map is more useful than a single list of all visible words because it shows where a language edit might change mathematics.
Copy the resource instead of repairing the original in public
GeoGebra’s Help Center describes a Copy Activity workflow for making an editable version of another user’s resource. Use the available copy or duplication route, respecting licenses and owner permissions, and preserve the original as the comparison reference. A separate translated copy gives you a rollback path and protects existing students who may be using the original worksheet. If an institution controls the source resource, agree on who is authorized to edit the copy and when it becomes visible to learners.
Record the source resource title and URL, the copied resource identifier, the target language, the class level and the software environment used for review. Small metadata habits prevent a later teacher from confusing the source and translated editions. A title like “Circle exploration—Malay learner version, term two” identifies a learning asset; an ambiguous “Circle final v4” makes testing and distribution difficult. When a resource embeds several applets or external files, note their roles as well so each dependent asset is reviewed.
Avoid assuming every shared activity grants the right to redistribute a modified copy. Public visibility is not universal permission to reuse, translate or publish. Check the stated terms and classroom agreements. Similarly, copying a resource may duplicate its current components without guaranteeing continued links to future source changes. Treat the target as a versioned edition and plan to compare it when the source’s teacher later edits instructions or corrects a mathematical error.
Write a mathematical acceptance test before changing words
For the radius activity, write down the expected outcomes in numbers. At r = 2, a circle should have diameter 4 and circumference approximately 12.57 in the chosen unit system. At r = 5, diameter should be 10 and circumference approximately 31.42. Whether the activity actually displays these quantities depends on its design, but a reviewer needs comparable fixed test states. Otherwise an attractive target screenshot at r = 3 can conceal an incorrect link to the underlying variable.
A second acceptance test describes the learner’s action. “Move the radius slider from 2 to 5, watch the circle enlarge, observe the translated label update, and decide whether doubling radius doubles area.” This tests the instructional sequence, not merely a number. If a target-language instruction says “drag the point” when the control is a slider, the wording must be corrected even though the construction itself works. Students need an instruction they can actually execute.
A third test covers the feedback. If the worksheet asks for the relationship between radius and area, inspect how an answer is evaluated or discussed. A translated explanation must maintain the square relationship, not accidentally describe a linear one. In an open-ended question, the goal may be equivalent mathematical reasoning rather than exact matching of strings. In an automatically checked question, target-language changes can interact with answer validation, so the rule itself must be inspected rather than inferred from the visible instruction.
Static text, dynamic text and mixed text need different treatment
GeoGebra’s official manual distinguishes static text, dynamic text and mixed text. Static text remains as entered unless edited. Dynamic text contains values of mathematical objects that update with the construction. Mixed text combines natural-language text with a changing value. The documentation illustrates expressions like "Point A = " + A, in which the quoted part is language and A is a reference to a mathematical object. This is an ideal example of the translation boundary.
To localize that expression, change only the quoted string into a natural equivalent for the target class while preserving the object reference and the concatenation relationship. If the target language ordinarily places the value first, the expression may require rearranging quoted pieces around A. The mathematical object remains the same. A simple search-and-replace across the whole expression is unsafe because it might rename A, delete a quote or convert a command into a literal string.
The reviewer should test the label after changing the object. Move point A and confirm that the translated text still displays its updated coordinates or value. If the sentence stays fixed, inspect whether a dynamic reference was accidentally converted into quoted prose. If an error appears, examine the exact concatenation syntax and the object name. The correct fix is usually local: restore the reference and adjust only the natural-language segment, not rebuild the entire geometry from scratch.
Minimal construction-aware example
Original mixed text expression: "Radius = " + r + " cm"
Target schematic: "[translated Radius label] = " + r + " cm"
Mathematical object: r (unchanged)
Test states: r = 2, r = 5 (dynamic display must change)The bracketed item is a translation placeholder for demonstration, not a phrase to publish. In a real target language the unit might need localized spacing or a different accompanying word, while the mathematical unit symbol itself remains stable. Do not convert r into a translated variable name merely to make the label more readable. Variables may be used elsewhere in definitions, scripts and object captions.
Variable names are not classroom vocabulary labels
Students may read object names like A, B, f, r and slope, but those tokens often belong to the construction’s internal graph. Renaming them can propagate unpredictably through expressions and references. Some systems can update references automatically when a name changes; even then, it is an unnecessary risk for a language-only assignment. Preserve object identifiers unless there is a separate, supervised mathematical redesign.
The visible caption of an object may be a more appropriate localization surface than its identifier. For example, the internal number remains r while the learner sees a target-language phrase equivalent to “radius.” Distinguish what the teacher wants students to recognize algebraically from what must be understandable as prose. In upper-secondary mathematics, a consistent symbol like m for gradient can be pedagogically valuable across languages. A translated description can explain the symbol without altering it.
This distinction matters when reusing one construction in multiple lessons. A teacher may embed the same graph in an English revision sheet and a Chinese revision sheet. If each translator renames internal objects to match the language, formulas, scripts and imported copies become harder to compare. Stable identifiers support a shared mathematical core, while localized captions support readers. A two-column inventory listing internal ID and target display label makes this separation auditable.
Sliders are experiments with controlled mathematical state
The GeoGebra manual describes sliders as representations of free numbers or angles, with settings such as interval, increment and animation behaviour. Those settings are part of the mathematical experience. An accidental change from an increment of 0.5 to 5 can make a lesson on fine-grained variation impossible. A translated caption must not lead the reviewer to modify slider bounds just to make text fit a narrow panel.
Begin QA at the minimum, a representative middle value and the maximum of each slider. Confirm that the object moves, every dependent label updates and any explanation is true across the range. A caption that says “positive radius” is misleading if the slider permits negative values; the source construction may itself need review, but translation cannot assume a restriction absent from the actual state. If an animation is enabled, observe it over time because some defects only arise at intermediate values.
Consider a statistics activity in which one slider controls a sample size and another controls a mean. If the translated word for “sample” could be read as an individual observation rather than the set, students may misinterpret what changing the slider does. This is a semantic defect: the control works, but the instruction names the wrong quantity. The reviewer needs to know the curriculum concept and the variable’s role in the construction before approving the phrase.
Buttons and scripts sit at an especially fragile boundary
A GeoGebra resource may contain buttons, input boxes, check boxes or scripted actions. A visible button caption can be localized, while a script triggered by the button is an executable instruction to the applet. Never assume the script contains only readable prose. A command name, object identifier or string used as a conditional input must be distinguished from the text displayed on the button itself.
Suppose a button resets a slider to its starting value. A translation can change “Start again” to an equivalent phrase and leave the reset action alone. But if a global replacement also changes a variable name inside the script, pressing the button may stop working. It is easy to overlook because the default view does not require that button to be used. A serious QA plan includes every control’s trigger, expected state change and error recovery.
Input boxes require special attention. A learner may type a number, an algebraic expression or a word. Translate the surrounding explanation without changing whether the input expects a mathematical expression or a specific term. For numeric inputs, decimal separators and local conventions can affect what students type; test the actual applet’s accepted format instead of promising that every locale uses the same syntax. Distinguish a localized written example from the input parser’s real requirements.
Separate activity-page questions from embedded applet text
The resource page and embedded GeoGebra applet may each own visible text. A teacher can edit an activity title, introductory paragraph and open-response question in one place, but a dynamic label inside an applet may be controlled by a construction expression elsewhere. A copied resource can therefore contain two independent versions of the same instructional phrase. Editing only the page text leaves the applet in the source language.
Create a screen-by-screen inventory: page title, objective, initial instruction, applet labels, feedback, follow-up question and completion message. Mark the owner of each string as page editor, applet editor, image file or external content. This prevents a common error in digital education: one component looks complete, and the whole activity is released even though a student’s next click exposes untranslated instructions or a mismatched answer.
Keep the learner’s navigation path in mind. A student might enter from a learning-management system or a shared link, encounter one activity and then continue to another. A translated title in the first view is not enough if a later page still says “Submit your conjecture” in a language students cannot read. Test the route a real learner will follow, and document any deliberately bilingual elements so they are not mistaken for missed work.
Mathematical register must match the class, not the dictionary
Geometry uses precise terms. Radius is not diameter, a line is not a segment, and a gradient is not necessarily interchangeable with a colloquial word for steepness. Translation choices should be based on the target-language syllabus and the age of the learners. In a primary activity, the text may reasonably introduce a simpler description before naming a technical term. In a secondary examination revision activity, the official terminology often matters more than conversational familiarity.
A glossary can be small and highly useful. Include the source concept, preferred target term, permissible alternative, symbol, example sentence and any warning about near-synonyms. The goal is consistency across controls and explanations, not rigid word substitution. “Rotation,” “reflection” and “translation” are related transformation terms but not interchangeable; an incorrect target label may lead students to carry out the wrong construction while believing they obeyed the question.
Bilingual teachers can add a concise cross-language note if the curriculum explicitly benefits from it, but this is a separate editorial choice. Avoid silently mixing technical terms in a supposedly monolingual learning path merely because a direct equivalent was hard to find. If the target class uses both languages, define which term will appear in the prompt and which in the explanation. That makes an apparent inconsistency into an intentional teaching aid.
Units, numbers and mathematical notation deserve a separate audit
A localized sentence may alter word order around cm, m², degrees or percentages, but it should not silently alter a measured quantity. When a slider value represents a length in centimetres, verify that its target label still describes centimetres. Converting to metres would require changing the displayed numeric value or calculation, not just replacing the unit word. An activity that claims a radius of 0.05 m while displaying a raw number 5 is teaching a conversion error.
Decimal commas and decimal points can be another trap. GeoGebra internally handles numeric syntax according to its conventions and settings, while a school worksheet may display locally customary notation. Do not copy a target-language printed decimal into a formula input without checking that it parses as intended. The safest localization preserves the mathematics engine’s accepted expression and explains the expected learner input clearly. Rendering and parser syntax are related but not the same problem.
Also distinguish an angle in degrees from one in radians. A translated label might say “angle” while the original task requires degrees, or it might omit the degree sign because the sentence has become longer. That omission changes the interpretation of the value. When the applet uses π or trigonometric ratios, keep the mathematical notation stable and verify the target explanation at characteristic values. A mathematically literate reviewer is essential here; fluent translators alone cannot validate the result.
Use a controlled comparison rather than relying on memory
Open the source and target editions side by side at the same mathematical states. Compare the geometry, slider limits, labels, questions and answer feedback. Choose states that expose different behaviour: endpoints, zero when valid, negative values if allowed, and a value that produces a non-integer result. Capturing only the starting frame hides bugs that a student will encounter when exploring.
Record each discrepancy with its type. A typo in a heading is a language defect. A static value that no longer changes is a dependency defect. An “incorrect” feedback message triggered by a valid answer is an assessment defect. An image that overflows the translated layout is a presentation defect. Labeling the cause helps the right person fix it quickly and avoids a broad rebuild for a small phrase-level issue.
Compare not only what appears on screen but also what the task asks students to infer. A translated open-ended question may unintentionally ask for proof where the source asked for an observation, or for a numerical answer where the source asked for a relationship. If the activity’s applet invites exploratory learning, preserve that sequence. A better-sounding instruction that reveals the expected conclusion too early can reduce the educational value of the interaction.
A worked circle activity: translate the complete learning journey
Begin with an activity that asks students to change the radius of a circle and discover how circumference and area respond. Its introductory text says to move the radius slider slowly and observe the circumference. Its applet displays a circle, a slider r, and dynamic text showing r, 2πr and πr². Its final question asks whether doubling the radius doubles the area. This gives three different content owners: introductory prose, dynamic labels and reflective assessment.
First localize the introductory prose so students know which control to move, what observation to make and which quantities are compared. Next translate static portions of applet labels without modifying r or the formulae. Then localize the final question while keeping its cognitive demand: students should reason that doubling radius multiplies area by four. If the source asks them to investigate, do not add an explicit answer in the translated instruction, even if doing so seems more helpful. The learning journey is designed to produce discovery.
Test at radius 2 and radius 4. The displayed circumference should double and the area should become four times as large. Check that the target text expresses those relationships, that no unit changes have been introduced, and that the graph still uses the same radius. Finally ask a student or colleague who has not seen the English source to explain their findings. If they correctly manipulate the applet but misunderstand the target question, that is a translation failure even when the equations themselves are fine.
Failure mode: changing an object identifier to make it sound local
Suppose the applet contains an object named AreaA and a quoted English label saying “Area of shape A.” The translator may believe both should be changed to a target-language phrase, especially when editing text from an exported file. A broad replacement might rename an internal reference while other expressions still call AreaA. The result may be an undefined value, disappearing label or broken script. The visible wording can be localized without changing the identifier, so this failure is avoidable.
To diagnose such a bug, return to the source object list and compare definitions. Check whether the named object still exists, whether dependent expressions resolve and whether the same button state produces the same outputs in both editions. If only one label is wrong, restore the underlying object name and adjust the quoted text. If an automated rename already propagated through the construction, inspect every affected dependency and regression-test the full applet rather than assuming automatic propagation was complete.
Do not disguise a broken reference by typing a current numeric value as a static label. That might make the screenshot match the source, but it destroys the dynamic relationship. Fix the mechanism and then retest with at least two different control values. In interactive mathematics, a correct initial frame is weak evidence; the behaviour across states is the product.
Failure mode: a translated instruction changes the required action
A source lesson might say “drag point B along the segment,” while a localized phrase is closer to “move point B anywhere.” The graph may constrain point B to the segment anyway, but the target student could interpret the task differently, attempt an impossible action and conclude that the tool is malfunctioning. Similar ambiguity occurs with “click,” “tap,” “drag,” “enter” and “slide,” which describe distinct user actions.
Review verbs against the affordances actually visible on the learner’s device. On a laptop a student may drag with a mouse, while on a tablet they may touch and move the object. Choose a wording that is accurate across the supported devices or supply a short device-specific instruction. Avoid prescribing an interface route that differs between GeoGebra apps or versions. The copy should tell students what mathematical action to carry out, not pretend that every platform exposes identical controls.
A good correction is demonstrated by testing, not by debating synonyms in isolation. Give the translated instruction to a learner, observe their first attempt and ask what they expected would move. If they choose the wrong control, revise the sentence. This feedback is especially valuable in Primary mathematics, where a technically precise instruction may still be too abstract to communicate a gesture clearly.
Failure mode: a dynamic label turns into a stale statement
A common regression occurs when a translator pastes a complete sentence into a text object that previously combined quoted words with a changing value. Initially the display says “Length = 6 cm,” which appears right. Then a slider changes the length to 9 but the label stays at 6. The translator has unintentionally turned dynamic evidence into decorative prose, making the applet contradict itself.
To catch this, tag every dynamic text object in the inventory and test it under multiple states. The question is not whether the target sentence has the same words as the source; it is whether the expression still incorporates the intended mathematical variable. If the expression uses a calculated number rather than a direct slider value, check the calculation as well. It is possible to preserve the wrong dynamic reference and still see the number change.
When the target language requires a different word order, reconstruct the expression with translated static fragments around the same dynamic objects. Inspect quotation marks and concatenation, then use the applet’s own preview. Do not generalize one correct syntax example across every script without checking the actual editor. Some expression editors display mixed text visually instead of showing only its raw source, so editing methods may differ by application and version.
Revisit accessibility and reading order after language changes
Longer translated labels can overlap sliders or diagrams. A label that was readable in English may cover a crucial point or hide the scale on a coordinate axis in a target language with longer phrases. Fix spacing and positioning at the display layer where possible; do not shrink text below legibility or move mathematical objects simply to make a label fit. The student needs to see both the relationship and the explanation clearly.
Colour is another instructional cue. If the source instruction asks for “the blue line,” the translated target must still point to the blue line after any visual changes. But colour alone should not be the only cue, because students with colour-vision differences may struggle. A localized resource can use object names, dash patterns or spatial descriptions as supplementary clues if they are consistent with the construction. Such accessibility improvements should be documented separately from straight translation so the owner can assess their educational effect.
Check keyboard and touch operation where supported, visible focus, alternate instructions and any accompanying screen-reader-friendly text. Not every GeoGebra construction or embedding environment offers the same accessibility features, so do not promise universal access from a screenshot. Report the actual tested conditions and known limitations. A useful localized resource enables real students in their real devices to make sense of the mathematics, not just an ideal user in a design review.
Practical comparison matrix for the release reviewer
Prepare a small matrix that pairs each control with a test state and expected interpretation. For a radius slider, minimum and maximum values expose bounds; middle values expose calculation updates; a zero value tests whether a formula or explanatory sentence handles the edge case; an animation tests continuous behaviour. For a button, click once, click twice and reset. For a text-input task, try the intended correct value, a plausible wrong value and a malformed entry. This is compact, rigorous evidence.
Record the source output and target output at each state. For purely mathematical objects, expect equivalence, not necessarily visually identical position if language text occupies different space. For explanations, expect equivalent claims and learner instructions. For feedback, expect the same acceptance or rejection decisions unless the assignment specifically includes localizing answer rules. Keep the distinction explicit so a reviewer does not approve a mathematically altered task merely because it looks similar.
The matrix also helps teams divide labour. A bilingual educator checks terminology and whether questions elicit the same reasoning. A technically confident editor checks identifiers, scripts and dependencies. A student-view reviewer checks usability without prior source knowledge. One person can perform all three roles for a small activity, but the criteria must remain separate; skill in one area does not automatically guarantee competence in the others.
Transfer exercise: from circle geometry to a linear graph
Now consider a new worksheet that displays a line y = mx + c controlled by sliders m and c. Its source instruction asks students to change m while leaving c fixed and observe how the gradient changes. The applet contains a dynamic caption built from the current value of m, a static instruction and a button that resets both sliders. You are asked to translate this for a class that uses a target-language term for gradient rather than the common English phrase “slope.”
What should remain fixed? Keep the internal identifiers m and c, the relation y = mx + c, the slider ranges and increments, and the reset script’s mathematical actions. Translate the visible noun and the instructions. Check that the target language distinguishes the gradient from the y-intercept, since the two controls have different conceptual jobs. At m = 0, the line is horizontal; at a negative m, the line slopes downward from left to right. Those states provide immediate tests of the explanation.
The answer also requires testing student inference. If the target language’s translated instruction says “raise the line” rather than “increase the gradient,” students may change c instead of m. The applet will function normally, but the activity will lead them to investigate the wrong variable. Give the target instructions to a student and observe which slider they touch first. Correct the wording before distribution. This illustrates how an interactive resource can be mechanically sound but educationally mislocalized.
Build an efficient review workflow for many activities
For a set of twenty GeoGebra worksheets, begin with an inventory of repeated labels and common mathematical concepts. Build a small glossary for transformations, coordinates, functions, measurement and interface actions. Then establish a safe editing pattern using one representative worksheet: copy, classify text, change language literals, test dynamic references, inspect mobile layout and approve terminology. The second worksheet will move much faster because the team already knows where language resides and which objects must not be touched.
Use a reviewer checklist and a consistent version label. Track source URL, copy URL, target locale, class level, applet count, dynamic labels, controls tested and known constraints. A single resource might contain multiple independent applets, so a page-level “done” flag is insufficient. Record a completion result for each embedded component and linked assessment. Where the target version reuses an asset from the original, note whether it is linked or duplicated; later source updates may affect it differently.
Resist the temptation to automate replacement of every English-looking string inside a construction file. Translation memory can accelerate repeated phrases, but object references, formula fragments and quoted labels must be parsed in context. Even a safe-looking word like “Mean” may name a command or an object as well as a natural-language heading. Automation is best used to propose candidates and flag unchanged visible text, while a construction-aware editor decides which strings are actually replaceable.
How to decide that the target edition is finished
Declare a translated activity complete only after a reviewer can follow it from start to finish without the source language. That includes opening the copied resource, understanding the objective, operating controls, observing all dynamic labels, answering questions and receiving feedback. Test the specific devices and browser conditions learners will use. If the original has embedded media with source-language speech or on-image labels, document whether those were localized or intentionally left as bilingual context.
The handoff should include a brief release note identifying the source edition, target language and tested mathematical states. It should say which text was changed, which construction components were protected, whether any mathematical defects were corrected with permission, and what still needs review. This is especially useful for tuition centres where several tutors teach the same topic: each can rely on one validated resource rather than making small ad hoc edits before a lesson.
A translated activity is not successful because the translation sounds polished in isolation. It is successful when the same mathematical phenomenon unfolds for the target-language student and the prompts guide them toward the same intended insight. That standard keeps localization grounded in teaching instead of surface decoration.
Final quality-control checklist
Confirm that the original is intact and a licensed copy was made, the intended class and language are specified, and page-level instructions are complete. Inspect static text, mixed text, slider labels, buttons, input boxes, embedded images and any media. Preserve identifiers, formulas, dependencies, script actions and mathematical units. Where text expressions combine literals and object values, test at several control states rather than only the initial view.
Verify slider endpoints, increments, intermediate values, any animations and reset behaviour. Confirm that the target prompts ask for the same mathematical reasoning as the source. Check long labels at the actual viewport size and review not just spelling but register, symbol names, reading direction and accessibility. Use a bilingual mathematics reviewer for high-stakes terminology, not a generic spellcheck alone.
Archive the tested target edition, source reference, defect log and representative before-and-after screenshots or state descriptions. If one component could not be tested, state that limitation rather than quietly marking it complete. These checks keep the next translation faster because they leave behind an explicit, reusable method for separating language from the mathematics engine.
Summary: keep the construction alive
GeoGebra activity translation is a matter of preserving cause and effect. A slider changes a mathematical object, a dynamic expression reports its value, and a question asks the learner to interpret what happened. Translate the words connecting those stages while safeguarding the variables, formulas and controls that make them work. A bilingual reviewer should evaluate both the text and the mathematical conclusion.
The broader transferable lesson is that interactivity must be translated as an experience, not captured as one picture. An artifact can look correct at the opening screen and fail when a student acts. Testing varied states is therefore the bridge between a fast draft and a reliable educational activity.
Frequently asked questions
Can I translate a GeoGebra activity without editing its mathematical construction?
Often yes. Activity titles, instructions, questions and static text can be localized while many formulas and object identifiers remain unchanged. Mixed or dynamic labels must be edited with care so their mathematical references still work. Test the complete student interaction afterward.
What is the difference between static and dynamic text in GeoGebra?
Static text is not tied to changing construction values. Dynamic text includes values that update as objects change. Mixed text joins ordinary language with such values; the quoted language fragments can be localized while object references remain functional.
Can I translate variable names such as r, m or A?
Usually preserve them for a language-only assignment. They may be referenced throughout definitions, scripts and labels. Translate the visible captions and explanations instead, unless a separate construction redesign is explicitly authorized and fully tested.
What should I test after translating a slider label?
Try the minimum, an intermediate value, the maximum and any special case relevant to the mathematics. Check that the geometric change and dynamic text agree, that units remain correct and that the translated instruction names the intended control.
Why does a translated label show the same number after the slider moves?
The translator may have converted a dynamic expression into a fixed text string. Inspect the mixed-text formula, restore the reference to the mathematical object, and test under multiple states. Changing only the displayed numeral hides the bug.
Is copying a public activity always permitted?
No. GeoGebra provides ways to copy and adapt resources, but reuse and redistribution still depend on licensing and the resource owner’s terms. Check permission and attribute correctly where required.
How do I localize buttons without breaking their actions?
Translate visible captions separately from scripts or commands. After editing, activate each button and compare the before-and-after changes to object values. A control that looks translated but no longer resets or checks an answer is not ready.
How do I handle decimal commas and degree symbols?
Preserve the engine’s mathematical syntax and test the actual input parser. Localize explanatory text and visible number presentation only where supported. Do not assume a printed punctuation convention can be pasted unchanged into a formula input.
What if a translated question gives away the mathematical conclusion?
Revise it to preserve the original cognitive demand. An exploratory question should invite investigation rather than announce the answer. A bilingual subject reviewer should compare what students are asked to infer, not only literal wording.
When is a localized interactive activity ready for students?
When a learner can complete its intended journey using the target language, every control works, dynamic labels update, assessment feedback remains valid, and the tested edition is archived with known limitations. A single screenshot is insufficient evidence.
Verified reference documents and further reading
- GeoGebra Help Center: Create GeoGebra Resources — activity copying, editing and composition.
- GeoGebra Manual: Texts — static, dynamic and mixed text.
- GeoGebra Manual: Slider Tool — intervals, increments and control behaviour.
- GeoGebra Manual: Scripting Commands — named commands that are not prose.
Relevant eduKateSG internal reading opportunities
- How People Translate Quickly | Localization Build Validation — checking that interactive target resources still function.
- How People Translate Quickly | QA Warning Suppression — prioritizing genuine defects.
- Secondary 1 Mathematics Tutor Clementi | Small Groups Tutorials — eduKateSG’s classroom mathematics teaching context.