expecco Wiki (Version 25.x) - Letzte Änderungen [de]

Expecco API/en

Cg — Thu, 25 Jun 2026 07:14:22 GMT

Pin Value

← Nächstältere Version		Version vom 25. Juni 2026, 07:14 Uhr
(2 dazwischenliegende Versionen desselben Benutzers werden nicht angezeigt)
Zeile 3.984:		Zeile 3.984:

	*jsonObject '''jsonValueIfPresent''' (''inPin'') <br>Returns the pin-datum as a [[Embedded_Systems_C_Bridge_API#JSON_Library\|jsonObject]] if it has one, NULL otherwise. Similar to "<code>jsonValue()</code>", but avoids the exception. The returned pointer becomes invalid after the action's execution.		*jsonObject '''jsonValueIfPresent''' (''inPin'') <br>Returns the pin-datum as a [[Embedded_Systems_C_Bridge_API#JSON_Library\|jsonObject]] if it has one, NULL otherwise. Similar to "<code>jsonValue()</code>", but avoids the exception. The returned pointer becomes invalid after the action's execution.

			void '''keepValue''' (''inPin'') <br> (rel. 26.1) Returns a pointer to the malloc's bulk data and tells the bridge that the underlying memory shall NOT be freed when execute() returns.
			This avoids the need to memcpy big data blocks from the incoming memory buffer; however it is now in your responsibility to free that memory eventually.<br>
			This is only to be used for strings, bytearrays and other vectors.

	Variable number of input/output pins are not supported. You should pass an array or structure if required.		Variable number of input/output pins are not supported. You should pass an array or structure if required.
Zeile 4.148:		Zeile 4.152:
	If a huge number of data elements (known as "''Bulk Data''") are to be transferred, transmission times may get much longer than the actual execution time of the action (keep in mind that the communication times take most of the time if the executed computation is simple).		If a huge number of data elements (known as "''Bulk Data''") are to be transferred, transmission times may get much longer than the actual execution time of the action (keep in mind that the communication times take most of the time if the executed computation is simple).

		⚫	If the cBridge runs on the same machine as expecco, shared memory will be used to exchange bulk data. This may reduce the transmission times drastically, depending on the amount of data (there are still other messages exchanged via the socket communication).
	There are multiple possible ways to deal with this situation:

			===== Data from C to Expecco =====
	* Passing a Pointer to expecco		* Passing a Pointer to expecco
	:if you can leave the data inside the bridge (eg. as malloc'd data vector), '''and''' expecco does not need to access the elements of the vector, you can leave the data inside the cBridge and only pass a handle to expecco. This handle may later be sent to another C-action (on the same bridge) to process the data at non-critical times (eg. at the end of the test).		:if you can leave the data inside the bridge (eg. as malloc'd data vector), '''and''' expecco does not need to access the elements of the vector, you can leave the data inside the cBridge and only pass a handle to expecco. This handle may later be sent to another C-action (on the same bridge) to process the data at non-critical times (eg. at the end of the test).
Zeile 4.161:		Zeile 4.166:
	:This is probably the fastest method to transmit bytes, but comes with those possible pitfalls. We therefore recommend to try the "Bulk Data as Vector" method first. In most situations that is either within the acceptable timing constraints, or too far off, such that the ByteArray transmission will also not be fast enough, and you should use the pointer passing method anyway.		:This is probably the fastest method to transmit bytes, but comes with those possible pitfalls. We therefore recommend to try the "Bulk Data as Vector" method first. In most situations that is either within the acceptable timing constraints, or too far off, such that the ByteArray transmission will also not be fast enough, and you should use the pointer passing method anyway.

			===== Passing Data from Expecco to C =====
⚫	If the cBridge runs on the same machine as expecco, shared memory will be used to exchange bulk data. This may reduce the transmission times drastically, depending on the amount of data (there are still other messages exchanged via the socket communication).
			Be aware that the bulk data pointer you receive at an input pin has been allocated
			in malloc'd memory which is freed when the execute() function returns. If the data is to be retained inside the C bridge (or passed to a function which keeps a reference), you have two choices:
			* memcpy the data into an internal buffer
			* tell the caller that the buld data should not be freed on return from execute
			The second variant is available with expecco 26.1, and preferred.<br>In either case, it is in your responsibility to free the memory when no longer needed.

	<small><sup>1)</sup></small>Starting with expecco 26.2, this is no longer required - there are now additional fast bulk transfer entries named: <code>putFloatsN</code>, <code>putDoublesN</code> etc.		<small><sup>1)</sup></small>Starting with expecco 26.1, this is no longer required - there are now additional fast bulk transfer entries named: <code>putFloatsN</code>, <code>putDoublesN</code> etc.

	=== Example1: Writing and Calling C Code in Expecco ===		=== Example1: Writing and Calling C Code in Expecco ===

Release Notes 26.x

Alkurz — Wed, 24 Jun 2026 09:39:00 GMT

Release 26.1 (Q2 2026)

← Nächstältere Version		Version vom 24. Juni 2026, 09:39 Uhr
Zeile 17:		Zeile 17:
	* Feature: SSH builtin, plus SSH and SFTP clients (both via [[Tools_FileBrowser/en\|FileBrowser]])		* Feature: SSH builtin, plus SSH and SFTP clients (both via [[Tools_FileBrowser/en\|FileBrowser]])
	* Feature: public/private keypair generation via a [[Tools_FileBrowser/en\|FileBrowser]] menu (for easy SSH setup)		* Feature: public/private keypair generation via a [[Tools_FileBrowser/en\|FileBrowser]] menu (for easy SSH setup)
			* Improvement: Bundled python3 installer updated to version 3.13.14
	* Feature: Qt-Plugin supports Qt6.8 ([[QT_Testing/en#ExpeccoTestService_Library%3A_Delivery_in_Expecco_Versions\|Delivered versions for QT and build environment]])		* Feature: Qt-Plugin supports Qt6.8 ([[QT_Testing/en#ExpeccoTestService_Library%3A_Delivery_in_Expecco_Versions\|Delivered versions for QT and build environment]])
	* Feature: improved search text box behavior in text editors (type RETURN, CMD-f or CMD-b while box is open) and back to original position button added.		* Feature: improved search text box behavior in text editors (type RETURN, CMD-f or CMD-b while box is open) and back to original position button added.

KI Coding Plugin

Cg — Tue, 23 Jun 2026 13:16:42 GMT

LLaMa.cpp

← Nächstältere Version		Version vom 23. Juni 2026, 13:16 Uhr
Zeile 390:		Zeile 390:
	llama-server.exe -hf unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF:Q4_K_M		llama-server.exe -hf unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF:Q4_K_M

	Das Modell läuft auf der lokalen CPU (Laptop, ohne GPU), und zeigt dabei ~~ein~~ ~~hervorragende~~ Antworten (wenn man Geduld hat). Zugegeben, wir haben auch einiges an Wissen in den Speicher vorgepackt.		Das Modell läuft auf der lokalen CPU (Laptop, ohne GPU), und zeigt dabei durchaus gute Antworten (wenn man Geduld hat). Zugegeben, wir haben auch einiges an Wissen in den Speicher vorgepackt.

	Im expecco Chat Fenster:		Im expecco Chat Fenster:

CVS Plugin/en

Robin — Tue, 23 Jun 2026 09:02:07 GMT

Introduction

← Nächstältere Version		Version vom 23. Juni 2026, 09:02 Uhr
Zeile 2:		Zeile 2:

	The plugin provides a user freindly interface to cvs, for managing expecco testsuites in a CVS repository (version management system).		The plugin provides a user freindly interface to cvs, for managing expecco testsuites in a CVS repository (version management system).
	~~THe~~ typical operations to checkin/checkout/compare/merge revisions are provided via convenient menu functions in expecco's "Plugin" → "Productivity" → "CVS" menu.		The typical operations to checkin/checkout/compare/merge revisions are provided via convenient menu functions in expecco's "Plugin" → "Productivity" → "CVS" menu.

	=== Prerequisites ===		=== Prerequisites ===

Benutzer:Robin

Sv — Tue, 23 Jun 2026 08:02:08 GMT

Benutzerkonto Robin wurde von Sv erstellt und das Passwort wurde per E-Mail zugesandt

KI Coding Plugin

Cg — Sat, 20 Jun 2026 13:06:21 GMT

LLaMa.cpp

Änderungen zeigen

Expecco API/en

Cg — Fri, 19 Jun 2026 23:47:26 GMT

Environment Access (Bridged Scheme)

KI Coding Plugin

Sv — Fri, 19 Jun 2026 13:11:46 GMT

add AI Coding license-feature note at top

@@ Zeile 1: / Zeile 1: @@
 = KI Coding Plugin =
 Das KI Coding Plugin bindet einen Large Language Model (LLM) basierten
@@ Zeile 122: / Zeile 124: @@
 == Coding Assistenten ==
 === Claude Code ===
@@ Zeile 157: / Zeile 161: @@
 Sollten Sie ein Komma vergessen haben wird beim Start eine Fehlermeldung erscheinen. Mit "/mcps" sollten die beiden aufgelistet werden.
 === Tabnine ===
 == Tipps ==

KI Coding Plugin

Cg — Fri, 19 Jun 2026 10:16:59 GMT

Versuche mit qwen2.5-coder:7b

@@ Zeile 117: / Zeile 117: @@
 als Kontext versendet.
-Bei Benutzung von Anthropic API-Tokens - wie hier der Fall - werden laut Anthropic die Daten nicht zum Training des KI-Modells genutzt ([https://privacy.claude.com/de/articles/7996868-werden-meine-daten-fur-das-modelltraining-verwendet Anthropic Erkärung dazu]). Für die Wahrheit dieser Aussage könne wir nicht garantieren und haften.
+Bei Benutzung von Anthropic API-Tokens - wie hier der Fall - werden laut Anthropic die Daten nicht zum Training des KI-Modells genutzt ([https://privacy.claude.com/de/articles/7996868-werden-meine-daten-fur-das-modelltraining-verwendet Anthropic Erkärung dazu]).<br>Für die Wahrheit dieser Aussage können wir nicht garantieren und haften.<br>Zu bedenken ist auch, daß die KI-Modelle nicht in inländischen Rechenzentren laufen, und eventuell weitere "Mitleser" involviert sind. Dies gilt insbesondere für Hightech, KI, Software und Defence Unternehmen, wo Fälle von Wirtschaftsspionage bekannt sind.
-Wenn Sie sicher sein wollen, daß keine Information Ihr Haus verlässt empfehlen wir eine eigenem geschlossene in-house KI Umgebung, wie sie bei einigen unserer Kunden bereits existiert. Sie dürfen dann selbstverständlich nur die URLs dieser Umgebung konfigurieren. Expecco kommuniziert ausschließlich mit der von Ihnen eingegebenen KI.
+⚠ Wenn Sie sicher sein wollen, daß keine Information Ihr Haus verlässt empfehlen wir eine eigenem geschlossene in-House KI Umgebung, wie sie bei einigen unserer Kunden bereits existiert. Sie dürfen dann selbstverständlich nur die URLs dieser Umgebung konfigurieren. Expecco kommuniziert ausschließlich mit der von Ihnen eingegebenen KI.
 == Tipps ==
@@ Zeile 125: / Zeile 165: @@
 * '''Mehrere Konversationen''': der Chat ist ein Singleton — eine neue Anfrage über ein Browser-/Editor-Menü startet jedesmal eine '''neue''' Konversation. Folge-Fragen (Klärung, Vertiefung) gehen über das Eingabefeld in derselben Konversation.
 * '''Vorsicht:''' Die Größe der übertragenen Daten (auch die aus dem Chatfenster) geht in die Kostenberechnung ein. Deshalb immer eine neue Konversation starten, wenn es um ein neues Thema geht.
-* '''Modell wechseln''': Für schnelle Routine-Antworten Sonnet, für schwierige Refactorings Opus. Über das '''Set model'''-Untermenü direkt aus der Toolbar oder über die Settings.
+* '''Modell wechseln''': Für schnelle Routine-Antworten Haiku oder Sonnet, für schwierige Refactorings Opus. Über das '''Set model'''-Untermenü direkt aus der Toolbar oder über die Settings.
 * '''Anbieter wechseln''': Sind Schlüssel für beide Anbieter konfiguriert, schaltet das '''Set provider'''-Untermenü ohne Umweg über den Settings-Dialog um.
 * '''Code direkt ausführen''': In Code-Karten kann markierter Smalltalk-Code per doIt / printIt / inspectIt direkt evaluiert werden — praktisch z.B. um vom Modell vorgeschlagene SUnit-Asserts gleich auszuprobieren, ohne den Code erst in einen Workspace zu kopieren.
@@ Zeile 131: / Zeile 171: @@
 == Tipps für die Arbeit mit claudeCode ==
-Wenn Sie claudeCode verwenden, stehen ihnen noch weit mehr Funktionen zur Verfügung: sie können zum Beispiel im claudeCode Terminal eingeben:
+Wenn Sie ClaudeCode oder Opencode verwenden, stehen ihnen noch weit mehr Funktionen zur Verfügung: sie können zum Beispiel im claudeCode Terminal eingeben:
 * "mach mir eine neue Suite in einem neuen Fenster",
 * "Mache eine Aktion zur Berechnung des Umfangs eines Kreises bei gegebenen Radius",
@@ Zeile 291: / Zeile 331: @@
 === Versuche mit qwen2.5-coder:7b ===
-Dieser zeigt erstaunlich gute Antworten auf unserer lokalen ollama installtaion mit einer (ncah heutigem Maßstab) durchschnittlichen GPU (Preis heute ca. 1500.-)
+Dieser zeigt erstaunlich gute Antworten auf unserer lokalen ollama installtaion mit einer (nach heutigem Maßstab) durchschnittlichen GPU (Preis heute ca. 1500.-)
 == Wahl der Modellgrösse ==

Settings LicenseInstallation

Cg — Thu, 18 Jun 2026 17:20:58 GMT

Detached Floating License

← Nächstältere Version		Version vom 18. Juni 2026, 17:20 Uhr
Zeile 48:		Zeile 48:

	;'''Request a new detached license'''		;'''Request a new detached license'''
	:Request a new detached license from the license server.		:Request a new detached license (aka ''road warrior license'') from the license server.
	:A detached license allows for offline operation for a predetermined amount of time (eg. for homework, on-the-road jobs or offline work).		:A detached license allows for offline operation for a predetermined amount of time (eg. for homework, on-the-road jobs or offline work). You can get a license from the server for a limited time and take it with your machine.

	;'''Install already detached floating license'''		;'''Install already detached floating license'''

KI Coding Plugin

Cg — Tue, 16 Jun 2026 15:32:04 GMT

Sitzung 2

← Nächstältere Version		Version vom 16. Juni 2026, 15:32 Uhr
(Eine dazwischenliegende Version desselben Benutzers wird nicht angezeigt)
Zeile 288:		Zeile 288:
	richtig!		richtig!

	Man muss den ~~Context~~ also mit Wissen füllen (fein tuning), dann kommt durchaus sinnvolles heraus. Wir sind dran, das zu evaluieren damit möglicherweise sogar eine lokale KI (allerdings mit GPU) hilfreich unterstützen kann.		Man muss den Kontext also mit Wissen füllen (oder fein tuning, wozu aber unsere GPU vermutlich nicht in der Lage ist), dann kommt durchaus sinnvolles heraus. Wir sind dran, das zu evaluieren damit möglicherweise sogar eine lokale KI (allerdings mit GPU) hilfreich unterstützen kann.

			=== Versuche mit qwen2.5-coder:7b ===
			Dieser zeigt erstaunlich gute Antworten auf unserer lokalen ollama installtaion mit einer (ncah heutigem Maßstab) durchschnittlichen GPU (Preis heute ca. 1500.-)

	== Wahl der Modellgrösse ==		== Wahl der Modellgrösse ==

Release Notes 26.x

Cg — Sun, 14 Jun 2026 15:33:16 GMT

Release 26.1 (Q2 2026)

← Nächstältere Version		Version vom 14. Juni 2026, 15:33 Uhr
Zeile 8:		Zeile 8:

	* Feature: '''NEW''' [[KI Coding Plugin\|'''AI Coding''']] plugin (chat-assistant integration for the activity editor and class browser):		* Feature: '''NEW''' [[KI Coding Plugin\|'''AI Coding''']] plugin (chat-assistant integration for the activity editor and class browser):
	** supports ~~both~~ '''Anthropic Claude''', '''OpenAI ChatGPT''' or '''~~Llama~~''' as backends, switchable from the settings dialog (Plugins → AI Coding);<br>API keys are stored per provider so you can flip between them without re-entering		** supports '''Anthropic Claude''', '''OpenAI ChatGPT''', '''ollama''' and others as backends, switchable from the settings dialog (Plugins → AI Coding);<br>API keys are stored per provider so you can flip between them without re-entering
	** Toolbar / class-browser menu adapts to the active provider — reads "''Ask Claude''" or "''Ask ChatGPT''", updates live when the provider is switched		** Toolbar / class-browser menu adapts to the active provider — reads "''Ask Claude''" or "''Ask ChatGPT''", updates live when the provider is switched
	** Menu actions: Explain code/method, Suggest improvement, Generate test, Generate doc-comment (fills the Documentation tab and pin comments), Find bugs, Custom prompt; "[Apply]" can install proposed code directly into the activity body or compile a proposed helper method into a class		** Menu actions: Explain code/method, Suggest improvement, Generate test, Generate doc-comment (fills the Documentation tab and pin comments), Find bugs, Custom prompt; "[Apply]" can install proposed code directly into the activity body or compile a proposed helper method into a class
Zeile 37:		Zeile 37:
	* Feature: OLE for 64 bit architectures		* Feature: OLE for 64 bit architectures
	* Feature: optional HTTPS for the AIDYMO and license server — drop a PEM cert+key into <code>--workDir</code> (combined <code>server.pem</code>, or split<code>fullchain.pem</code>+<code>privkey.pem</code> / <code>cert.pem</code>+<code>key.pem</code>) and the service binds TLS automatically; informational hostname is derived from the certificate (SAN-aware, wildcard- and multi-SAN-safe)		* Feature: optional HTTPS for the AIDYMO and license server — drop a PEM cert+key into <code>--workDir</code> (combined <code>server.pem</code>, or split<code>fullchain.pem</code>+<code>privkey.pem</code> / <code>cert.pem</code>+<code>key.pem</code>) and the service binds TLS automatically; informational hostname is derived from the certificate (SAN-aware, wildcard- and multi-SAN-safe)
	* Feature: more search options in the [[How_to_Program/en#MethodFinder:_Find_Functions_by_Example \| MethodFinder]].		* Feature: more search options in the [[How_to_Program/en#MethodFinder:_Find_Functions_by_Example \| MethodFinder]]
			* UI: Transcript (the console) always appends text at end.
			* UI: Transcript does not scroll to end on incoming text while moving the scrollbar
	* Fix: many fixes related to DPI scaling. I.e. when multiple monitors are configured with different scaling (especially different from 100%). Includes scaling of fonts, bitmap and UI components (widgets).		* Fix: many fixes related to DPI scaling. I.e. when multiple monitors are configured with different scaling (especially different from 100%). Includes scaling of fonts, bitmap and UI components (widgets).
	* Fix: display of very long lines in a text editor/inspector (workaround a Windows 16bit line limit)		* Fix: display of very long lines in a text editor/inspector (workaround a Windows 16bit line limit)

Expecco API/en

Cg — Sat, 13 Jun 2026 11:48:08 GMT

Debugging Python Actions

← Nächstältere Version		Version vom 13. Juni 2026, 11:48 Uhr
Zeile 3.095:		Zeile 3.095:
	The following describes the first kind of blocks (bridged). Script blocks are described elsewhere, in the [[ElementaryBlock_Element/en#Python_Script_Blocks \| Python-Script Blocks]] chapter.		The following describes the first kind of blocks (bridged). Script blocks are described elsewhere, in the [[ElementaryBlock_Element/en#Python_Script_Blocks \| Python-Script Blocks]] chapter.

	=== Debugging Python Actions ===		=== Debugging Bridged Python Actions ===
	Debug support for Python actions is still being ~~developped~~ and more features are added with newer versions. However, there are probably better IDEs available and it might be a good idea to develop and debug Python code in your preferred IDE, and only add interface code to already debugged modules as expecco actions (this is certainly a matter of your personal preferences; in the past, interfaces to many modules have been implemented using only expecco's debug facilities).		Debug support for bridged Python actions is still being developed and more features are added with newer versions. However, there are probably better IDEs available and it might be a good idea to develop and debug Python code in your preferred IDE, and only add interface code to already debugged modules as expecco actions (this is certainly a matter of your personal preferences; in the past, interfaces to many modules have been implemented using only expecco's debug facilities).

	In order to support debugging of Python actions, you have to install either the "''debugpy''" or the "''ptvsd''" package (*) with:		In order to support debugging of Python actions, you have to install either the "''debugpy''" or the "''ptvsd''" package (*) with:

KI Coding Plugin

Cg — Fri, 12 Jun 2026 12:25:47 GMT

Modellwahl; Sonnet, Heiku und Fable

← Nächstältere Version		Version vom 12. Juni 2026, 12:25 Uhr
Zeile 154:		Zeile 154:
	==== Modellwahl; Sonnet, Heiku und Fable ====		==== Modellwahl; Sonnet, Heiku und Fable ====
	Sonnet-4 ist gefühlt weit intelligenter als Heiku-4.5,		Sonnet-4 ist gefühlt weit intelligenter als Heiku-4.5,
	aber auch weit teurer. Fable ist für sehr komplexe Aufgaben, aber sehr teuer. Ihr Budget kann sehr schnell schmelzen mit Sonnet - insbesondere wenn man auf medium oder high Intelligenzlevel arbeitet. Haiku macht weit mehr Fehler und man sollte genau hinschauen was er so generiert. Gefühlt war Sonnet selbst im low level besser als Haiku. Aber der Preis...		aber auch weit teurer. Fable ist für sehr komplexe Aufgaben, aber sehr teuer (ich meine: wirklich teuer). Ihr Budget kann sehr schnell schmelzen mit Sonnet - insbesondere wenn man auf medium oder high Intelligenzlevel arbeitet. Haiku macht weit mehr Fehler und man sollte genau hinschauen was er so generiert. Gefühlt war Sonnet selbst im low level besser als Haiku. Aber der Preis...
	Noch besser ist Fable, aber extrem teuer; in einer halben Stunde kann da schon mal das Tagesbudget an Tokens verbraucht werden.		Noch besser ist Fable, aber extrem teuer; in einer halben Stunde kann da schon mal das Tagesbudget an Tokens verbraucht werden.

Zeile 176:		Zeile 176:

	* wechseln sie das Modell oder den level, je nach Komplexität der Aufgabe. Aber bedenken Sie dass Wissen von Session zu Session verloren geht, wenn es nicht erinnert wird. Technisch sind Erinnerungen nichts anderes als das Lesen einer Liste der vormaligen remember Anweisungen. Kosten also auch Tokens. Sie könnten also die Erinnerungen auch einfach neu eintippen ;-)		* wechseln sie das Modell oder den level, je nach Komplexität der Aufgabe. Aber bedenken Sie dass Wissen von Session zu Session verloren geht, wenn es nicht erinnert wird. Technisch sind Erinnerungen nichts anderes als das Lesen einer Liste der vormaligen remember Anweisungen. Kosten also auch Tokens. Sie könnten also die Erinnerungen auch einfach neu eintippen ;-)

			* Fable ist ein echter Star, wenn es darum geht, z.B. eine komplexe Architektur zu entwerfen. Mein Tipp: lassen sie ihn einen Plan machen, und von einem einfacheren Modell Schritt für Schritt umsetzen.

	* auch wenn es so scheint: das "Ding" ist kein Mensch; lassen sie "bitte, danke, oder Begründungen - es sei denn sie dienen der Entscheidung - weg. Tatsächlich erhöhen sie den Tokenverbrauch u.U.		* auch wenn es so scheint: das "Ding" ist kein Mensch; lassen sie "bitte, danke, oder Begründungen - es sei denn sie dienen der Entscheidung - weg. Tatsächlich erhöhen sie den Tokenverbrauch u.U.

Recover unsaved versions after a crash/en

Cg — Thu, 11 Jun 2026 22:47:19 GMT

Die Seite wurde neu angelegt: „== Recovering Unsaved Versions after a Crash == In case of a severe error, power outage, or other event which prevented you from saving your work, chose the "''Restore Unsaved Versions''" menu item from the "''File''" menu. Expecco saves snapshots in the background of your work on your local machine in regular intervals (typically 30s to a minute after you last modification). These older unsaved versions can be recovered in this menu. It offers compare…“

Neue Seite

== Recovering Unsaved Versions after a Crash ==
In case of a severe error, power outage, or other event which prevented you from saving your work, chose the "''Restore Unsaved Versions''" menu item from the "''File''" menu.

Expecco saves snapshots in the background of your work on your local
machine in regular intervals (typically 30s to a minute after you last modification). These older unsaved versions can be recovered in this menu.

It offers compare and load functions and should be self explaining. It even keeps versions of suites which have been
finished intentionally by the user, in case you deleted a suite
or left without saving.

Be aware that saving is performed in the background, and only if
the autosave got a change to finish his work before you change the suite again will a saved suite be complete. It stops saving whenever the next change is done, and starts again.
This is done to avoid pause times, because the saved suite must be consistent, and autosave is not smart enough to handle the suite being changed while it is saved.
Thus, pause from time to time for a few seconds, to allow autosave to finish its saving.

FileMenu

Cg — Thu, 11 Jun 2026 22:42:41 GMT

← Nächstältere Version		Version vom 11. Juni 2026, 22:42 Uhr
(Eine dazwischenliegende Version desselben Benutzers wird nicht angezeigt)
Zeile 3:		Zeile 3:
	The "''File''" menu functions are:		The "''File''" menu functions are:

	* "''Restore Unsaved Versions''"<br>To recover recent edits after a crash.		* "''Restore Unsaved Versions''"<br>To [[Recover_unsaved_versions_after_a_crash/en \| recover recent edits after a crash]].

	* "''New Testsuite''"<br>Creates a new test suite. As specified in the settings, the new suite is either empty or contains a template structure with some sample actions and tests. These are useless demo actions, so you may want to immediately click "''Undo''" to get rid of it, or disable the template suite generation alltogether in the "''Look & Feel''" settings.		* "''New Testsuite''"<br>Creates a new test suite. As specified in the settings, the new suite is either empty or contains a template structure with some sample actions and tests. These are useless demo actions, so you may want to immediately click "''Undo''" to get rid of it, or disable the template suite generation alltogether in the "''Look & Feel''" settings.

KI Coding Plugin

Cg — Thu, 11 Jun 2026 21:51:13 GMT

KI Coding Plugin

← Nächstältere Version		Version vom 11. Juni 2026, 21:51 Uhr
(3 dazwischenliegende Versionen desselben Benutzers werden nicht angezeigt)
Zeile 10:		Zeile 10:
	* '''OpenAI ChatGPT''' (gpt-4o, gpt-4o-mini, gpt-4.1, gpt-4.1-mini, gpt-4.1-nano, o1, o3)		* '''OpenAI ChatGPT''' (gpt-4o, gpt-4o-mini, gpt-4.1, gpt-4.1-mini, gpt-4.1-nano, o1, o3)
	* '''Ollama''' lokal oder remote (codellama:7b, codellama:13b)		* '''Ollama''' lokal oder remote (codellama:7b, codellama:13b)
			Zusätzlich gibt es noch einen weiteren Eintrag '''Inhause'''; dieser ist für eigene, firmeninterne KI vorgesehen.
			:: Anm.: tatsächlich sprechen all hinterher das selbe OpenAI chatGPT Protokoll; sie können also auch unter irgend einem anderen Anbieter einfach die URL zu ihrer eignene KI eingeben. Dieser Eintrag dient also lediglich der Übersichtlichkeit, und einem Platz wo Sie Ihre Zugangsdaten ablegen können, wenn Sie doch noch zwischen durch mit anderen Anbietern chatten können und wollen ohne jedesmal die URLs und API keys neu eingeben zu müssen.

	Je nach gewähltem Anbieter erscheint die Toolbar-Schaltfläche als		Je nach gewähltem Anbieter erscheint die Toolbar-Schaltfläche als
	<code>Ask Claude</code> bzw. <code>Ask ChatGPT</code> bzw. <code>Ask ...</code> ; das Einstellungs-Tab		<code>Ask Claude</code> bzw. <code>Ask ChatGPT</code> bzw. <code>Ask ...</code> ; das Einstellungs-Tab
Zeile 287:		Zeile 290:
	== Wahl der Modellgrösse ==		== Wahl der Modellgrösse ==

	Versuche mit gemma3:270m (sehr klein; 270Mio), gemma3:1b (1 Mrd) und gemma3:4b (4 Mrd). Alle auf lokalem Rechner ohne GPU (was keinen Einfluss auf die Antwort hat, aber die Antwortzeit bestimmt).		Versuche mit gemma3:270m (sehr klein; 270Mio), gemma3:1b (1 Mrd) und gemma3:4b (4 Mrd).<br>Alle auf lokalem Rechner ohne GPU (was keinen Einfluss auf die Antwort hat, aber die Antwortzeit bestimmt).

	Versuch: Übersetzung einer Bausteindokumentation.		Versuch: Übersetzung einer Bausteindokumentation.

← Nächstältere Version		Version vom 19. Juni 2026, 23:47 Uhr
(11 dazwischenliegende Versionen desselben Benutzers werden nicht angezeigt)
Zeile 4.597:		Zeile 4.597:
	==== Reporting (Bridged Scheme) ====		==== Reporting (Bridged Scheme) ====

	*'''error'''(''infoString'' [ ,''arg1'' ... ~~,''arg4''~~ ]) <br>Report a defect (in the test). ~~Up to 4 optional~~ args are sliced into the string, if it contains "%i" placeholders. Stops execution.		*('''error''' ''infoString'' [ ''arg1'' ... ]) <br>Report a defect (in the test). Optional args are sliced into the string, if it contains "%i" placeholders. Stops execution.

	*'''fail'''(''infoString'' [ ,''arg1'' ... ~~,''arg4''~~ ]) <br>Report a failure (in the SUT). Stops execution.		*('''fail''' ''infoString'' [ ''arg1'' ... ]) <br>Report a failure (in the SUT). Stops execution.

	*'''inconclusive'''(''infoString'' [ ,''arg1'' ... ~~,''arg4''~~ ]) <br>Report an inconclusive test. Stops execution.		*('''inconclusive''' ''infoString'' [ ''arg1'' ... ]) <br>Report an inconclusive test. Stops execution.

	*'''activitySuccess'''(''infoString'' [ ,''arg1'' ... ~~,''arg4''~~ ]) <br>Finishes the current activity with success (same as "success()").		*('''activitySuccess''' ''infoString'' [ ''arg1'' ... ]) <br>Finishes the current activity with success (same as "success()").

	*'''pass'''(''infoString'' [ ,''arg1'' ... ~~,''arg4''~~ ]) <br>Finishes the current testCase with success.		*('''pass''' ''infoString'' [ ''arg1'' ... ]) <br>Finishes the current testCase with success.


Zeile 4.630:		Zeile 4.630:

	--- unfinished --		--- unfinished --
	*'''confirm'''(''msg'')<br>Opens a simple yes/no dialog; returns a boolean.		*('''confirm''' ''msg'')<br>Opens a simple yes/no dialog; returns a boolean.

	*'''request'''(''msg'')<br>Opens a simple string-input dialog; returns a string or nil if cancel was pressed.		*('''request''' ''msg'')<br>Opens a simple string-input dialog; returns a string or nil if cancel was pressed.

	*'''requestPassword'''(''msg'')<br>same, but the entered string is not shown on the screen		*('''requestPassword''' ''msg'')<br>same, but the entered string is not shown on the screen

	*'''requestFilename'''(''msg'')<br>asks for a filename		*('''requestFilename''' ''msg'')<br>asks for a filename


	==== Logging (Bridged Scheme) ====		==== Logging (Bridged Scheme) ====

	*'''logFail'''(''messageString'') <br>Adds a fail message to the activity log.		*('''logFail''' ''messageString'') <br>Adds a fail message to the activity log.

	*'''logError'''(''messageString'') <br>Adds a error message to the activity log.		*('''logError''' ''messageString'') <br>Adds a error message to the activity log.

	*'''logWarning'''(''messageString'') <br>Adds a warning to the activity log.		*('''logWarning''' ''messageString'') <br>Adds a warning to the activity log.

	*'''logInfo'''(''messageString'') <br>Adds an info message to the activity log.		*('''logInfo''' ''messageString'') <br>Adds an info message to the activity log.

	*'''alert'''(''messageString'')<br>Adds a warning message to the activity log, and also shows a DialogBox, which has to be confirmed by the operator. The dialog box and confirmation can be disabled by a settings flag in the "''Execution''" → "''Log Settings''" dialog (by default it is disabled).		*('''alert''' ''messageString'')<br>Adds a warning message to the activity log, and also shows a DialogBox, which has to be confirmed by the operator. The dialog box and confirmation can be disabled by a settings flag in the "''Execution''" → "''Log Settings''" dialog (by default it is disabled).

	*'''warn'''(''messageString'')<br>Same as ''alert:'' (for JavaScript compatibility).		*('''warn''' ''messageString'')<br>Same as ''alert:'' (for JavaScript compatibility).


	==== Environment Access (Bridged Scheme) ====		==== Environment Access (Bridged Scheme) ====
	*'''environmentAt'''(<varName>);<br>Fetches and returns a value from the expecco environment which is in scope of the current activity.<br>You can only read simple objects (numbers, booleans and strings) from ruby actions.		*('''environmentAt''' <varName>)<br>Fetches and returns a value from the expecco environment which is in scope of the current activity.<br>You can only read simple objects (numbers, booleans and strings) from ruby actions.

	*'''environmentAtPut'''(<varName>, <newValue>);<br>Writes a value into the expecco environment which is in scope of the current activity.<br>The variable must be writable.<br>You can only write simple objects (numbers, booleans and strings) from ruby actions.		*('''environmentAtPut!''' <varName> <newValue>)<br>Writes a value into the expecco environment which is in scope of the current activity.<br>The variable must be writable.<br>You can only write simple objects (numbers, booleans and strings) from ruby actions.
	<!--		<!--
	*'''eval''' (''smalltalkCodeString'') <br>Evaluate a piece of Smalltalk code inside expecco.		*'''eval''' (''smalltalkCodeString'') <br>Evaluate a piece of Smalltalk code inside expecco.
Zeile 4.666:		Zeile 4.667:
	==== Input Pins (Bridged Scheme) ====		==== Input Pins (Bridged Scheme) ====

	*('''pin-hasValue? ''pin'')<br>Returns #tif the pin has received a value		*('''pin-hasValue?''' ''pin'')<br>Returns #tif the pin has received a value

	*('''pin-value ''pin'')<br>Returns the value of the pin. Raises an error if the pin did not receive any value.		*('''pin-value''' ''pin'')<br>Returns the value of the pin. Raises an error if the pin did not receive any value.

	*('''pin-valueIfAbsent''' ''pin'' ''alternativeValue'') <br>Returns the value of a pin or the value from alternativeValue if the pin did not receive any value.		*('''pin-valueIfAbsent''' ''pin'' ''alternativeValue'') <br>Returns the value of a pin or the value from alternativeValue if the pin did not receive any value.
Zeile 4.679:		Zeile 4.680:
	==== Output Pins (Bridged Scheme) ====		==== Output Pins (Bridged Scheme) ====

	*''pin''.'''~~value~~'''(''someValue'')<br>Writes the value to the pin. Only simple object can be transferred by value (nil, booleans, integers, floats, strings).<br>Anything else should be passed by reference (see <code>makeRef</code> below).		*('''pin-value!''' ''pin'' ''someValue'')<br>Writes the value to the pin. Only simple object can be transferred by value (nil, booleans, integers, floats, strings).<br>Anything else should be passed by reference (see <code>makeRef</code> below).


	==== Passing Objects by Reference (Bridged Scheme) ====		==== Passing Objects by Reference (Bridged Scheme) ====
	By default, objects written to output pins via "'''value:'''" will be marshalled to JSON, transferred to expecco and decoded there.		By default, objects written to output pins via "('''pin-value!''' ...)" will be marshalled to JSON, transferred to expecco and decoded there.
	Effectively, a copy of the object is created, which looses its identity when sent back later to the remote ruby in another action.		Effectively, a copy of the object is created, which looses its identity when sent back later to the remote ruby in another action.
	This would make it impossible to get a handle from a remote action, which is to be sent to another action on the same remote machine.		This would make it impossible to get a handle from a remote action, which is to be sent to another action on the same remote machine.
	The "<code>makeRef</code>" method solves this, by creating a "pointer" or "handle" to the object inside the remote machine.		The "<code>makeRef</code>" method solves this, by creating a "pointer" or "handle" to the object inside the remote machine.

	*'''makeRef'''(''object'')<br>creates a reference object, which can be written to a pin.		*('''makeRef''' ''object'')<br>creates a reference object, which can be written to a pin.

	*'''makeRef'''(''object'', '''name''')<br>ditto, but gives it a user friendly name (eg. for the expecco log)		*('''makeRef''' ''object'', '''name''')<br>ditto, but gives it a user friendly name (eg. for the expecco log)
	:I.e. to write a reference to a pin, use:		:I.e. to write a reference to a pin, use:
	::: ''~~somePin~~''.'''~~value~~'''('''makeRef'''(''someObject''))		::: ('''pin-value!''' ''somePin'' '''('''makeRef''' ''someObject''))

	== DotNET Elementary Blocks ==		== DotNET Elementary Blocks ==