8000 It's always stuck. · Issue #2352 · RooVetGit/Roo-Code · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

It's always stuck. #2352

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
msjsc001 opened this issue Apr 6, 2025 · 11 comments
Closed

It's always stuck. #2352

msjsc001 opened this issue Apr 6, 2025 · 11 comments
Labels
bug Something isn't working needs information Needs more information to be actionable

Comments

@msjsc001
Copy link
msjsc001 commented Apr 6, 2025

App Version

Version: 3.11.8

API Provider

OpenRouter

Model Used

google/gemini-2.5-pro-exp-03-25:free

Actual vs. Expected Behavior

When using Gemini 2.5 Pro to edit code exceeding 400 lines, it easily gets stuck and shows no action.

When reading long code (over 300 lines) via API, it often gets stuck, remaining stuck in the API request.

Detailed Steps to Reproduce

1.When using Gemini 2.5 Pro to edit code exceeding 400 lines, it easily gets stuck and shows no action.

2.When reading long code (over 300 lines) via API, it often gets stuck, remaining stuck in the API request.

Relevant API Request Output

[read_file for 'translations.json'] Result:

<file>
  <path>translations.json</path>
  <content>
  1 | {
  2 |     "zh": {
  3 |       "window_title": "Anki-TTS-Edge (自定义版本 v1.6)",
  4 |       "input_text_label": "📝 输入文本:",
  5 |       "tab_voices": "🔊 声音",
  6 |       "tab_settings": "⚙️ 设置",
  7 |       "tab_appearance": "🎨 外观",
  8 |       "filter_language_placeholder": "筛选语言 (如: zh)...",
  9 |       "voice_list_label_1": "声音列表 1",
 10 |       "voice_list_label_2": "声音列表 2",
 11 |       "refresh_voices_button": "🔄 刷新声音
8000
列表",
 12 |       "rate_label": "语速:",
 13 |       "volume_label": "音量:",
 14 |       "settings_output_cache_label": "输出与缓存",
 15 |       "settings_copy_label": "🔗 自动复制文件",
 16 |       "settings_autoplay_label": "▶️ 自动播放",
 17 |       "settings_max_files_label": "🔢 最大缓存数:",
 18 |       "settings_clipboard_label": "剪贴板功能",
 19 |       "settings_enable_ctrl_c_label": "🖱️启用复制生音频浮窗 (ctrl+c)",
 20 |       "settings_enable_selection_label": "启用划选生音频浮窗 (鼠标划选)",
 21 |       "appearance_theme_label": "界面主题:",
 22 |       "appearance_mode_light": "浅色模式",
 23 |       "appearance_mode_dark": "深色模式",
 24 |       "appearance_color_label": "自定义主色 (Hex):",
 25 |       "appearance_pick_color_button": "🎨",
 26 |       "appearance_apply_color_button": "应用颜色",
 27 |       "generate_button": "生成音频",
 28 |       "status_ready": "✅ 准备就绪",
 29 |       "status_error": "❌ 错误",
 30 |       "status_success": "✅ 成功",
 31 |       "status_generating": "⏳ 正在生成音频",
 32 |       "status_getting_voices": "⏳ 正在获取声音列表...",
 33 |       "status_voices_updated": "✅ 声音列表已更新",
 34 |       "status_monitor_enabled": "✅ 剪贴板/划选监控已启用",
 35 |       "status_monitor_disabled": "❌ 剪贴板/划选监控已禁用",
 36 |       "status_file_copied": "✅ 文件已复制到剪贴板",
 37 |       "status_copy_failed": "❌ 复制文件失败",
 38 |       "status_play_error": "❌ 播放错误",
 39 |       "status_generate_error": "❌ 音频生成失败",
 40 |       "status_empty_text_error": "❌ 错误：请输入文本",
 41 |       "status_no_voice_error": "❌ 错误：请选择一个声音",
 42 |       "status_settings_saved": "✅ 设置已保存",
 43 |       "status_settings_save_failed": "❌ 保存设置失败",
 44 |       "status_mixer_init_error": "❌ 播放器初始化错误",
 45 |       "status_file_not_found": "❌ 错误: 文件不存在",
 46 |       "error_invalid_color_title": "无效颜色",
 47 |       "error_invalid_color_message": "请输入有效的 6 位十六进制颜色代码 (例如 #FF5733)，而不是 '{0}'",
 48 |       "error_mixer_init_title": "初始化错误",
 49 |       "error_mixer_init_message": "无法初始化音频播放器 (pygame): {0}\n自动播放功能将不可用。",
 50 |       "debug_voices_loaded": "获取到 {0} 个声音，已按 语言->地区->名称 层级分类。",
 51 |       "debug_voices_load_failed": "获取声音列表失败: {0}",
 52 |       "debug_run_async_voices_error": "运行异步获取声音任务时出错: {0}",
 53 |       "debug_voice_ui_updated": "DEBUG: Voice UI updated. Current Voice: {0}",
 54 |       "debug_voice_ui_no_data": "DEBUG: No voice data received.",
 55 |       "debug_voice_load_failed_ui": "获取失败",
 56 |       "debug_no_matching_voices": "无匹配声音",
 57 |       "debug_voice_map_empty": "DEBUG: voice_display_to_full_map is empty in _populate_inline_voice_list",
 58 |       "debug_voice_selected": "DEBUG _select_voice_inline: Selected {0}",
 59 |       "debug_apply_color": "应用自定义颜色: {0}",
 60 |       "debug_load_settings_failed": "加载设置失败: {0}",
 61 |       "debug_invalid_color_loaded": "警告：加载的自定义颜色 '{0}' 无效，将使用默认颜色。",
 62 |       "debug_settings_saved": "设置已保存。",
 63 |       "debug_settings_save_failed": "保存设置失败: {0}",
 64 |       "debug_generating_audio": "准备生成音频: voice='{0}', rate='{1}', volume='{2}', pitch='{3}'",
 65 |       "debug_output_path": "输出路径: {0}",
 66 |       "debug_generate_thread_error": "运行异步生成任务时出错: {0}",
 67 |       "debug_edge_tts_ok": "Edge TTS 音频生成成功: {0}",
 68 |       "debug_edge_tts_fail": "Edge TTS 音频生成失败: {0}",
 69 |       "debug_edge_tts_internal_error": "Edge TTS 内部错误",
 70 |       "debug_empty_text": "文本为空，无法生成音频",
 71 |       "debug_audio_complete": "音频生成完成: {0}",
 72 |       "debug_autoplay_manual": "DEBUG: Autoplay (manual)? {0}",
 73 |       "debug_autoplay_float": "DEBUG: Autoplay (float)? {0}",
 74 |       "debug_autocopy_float": "DEBUG: Autocopy (float)? {0}",
 75 |       "debug_pygame_play_call": "DEBUG: play_audio_pygame called with path: {0}",
 76 |       "debug_pygame_stop_current": "DEBUG: 停止当前播放...",
 77 |       "debug_pygame_load_play": "DEBUG: 尝试加载并播放: {0}",
 78 |       "debug_pygame_play_start": "DEBUG: Pygame 播放已启动。",
 79 |       "debug_pygame_play_error": "ERROR: Pygame 播放错误: {0}",
 80 |       "debug_play_unknown_error": "ERROR: 播放时发生未知错误: {0}",
 81 |       "debug_float_trigger_text": "通过浮窗触发生成: {0}...",
 82 |       "debug_no_float_text": "ERROR: No text found for float trigger.",
 83 |       "debug_monitor_start": "启动剪贴板和鼠标监控...",
 84 |       "debug_poll_thread_start": "DEBUG: Clipboard polling thread started.",
 85 |       "debug_initial_paste_error": "DEBUG: Error getting initial paste for monitor start: {0}",
 86 |       "debug_initial_paste_set": "DEBUG: Initial clipboard content set: {0}...",
 87 |       "debug_new_clipboard_content": "检测到新的剪贴板内容: {0}...",
 88 |       "debug_poll_mouse_pos": "轮询检测到变化，鼠标位置: {0}",
 89 |       "debug_poll_trigger_error": "Error triggering float window from poll: {0}",
 90 |       "debug_poll_clipboard_error": "剪贴板访问错误(忽略): {0}",
 91 |       "debug_poll_generic_error": "剪贴板监控错误: {0}",
 92 |       "debug_poll_thread_stop": "剪贴板监控线程已停止。",
 93 |       "debug_mouse_pressed": "DEBUG: Mouse pressed at {0}",
 94 |       "debug_mouse_released": "DEBUG: Mouse released. Dist: {0:.2f}, Duration: {1:.2f}s",
 95 |       "debug_selection_detected": "检测到鼠标划选结束...",
 96 |       "debug_simulate_copy": "DEBUG: Simulating Ctrl+C...",
 97 |       "debug_simulate_copy_complete": "DEBUG: Ctrl+C Simulation complete.",
 98 |       "debug_simulate_copy_error": "ERROR: 模拟 Ctrl+C 时出错: {0}",
 99 |       "debug_mouse_release_error": "Error in mouse release logic: {0}",
100 |       "debug_mouse_listener_stop_error": "停止鼠标监听器时出错: {0}",
101 |       "debug_mouse_listener_thread_stop": "鼠标监听器线程已结束。",
102 |       "debug_monitor_stop": "正在停止剪贴板和鼠标监控...",
103 |       "debug_closing": "窗口关闭...",
104 |       "debug_pygame_stop_mixer": "停止 Pygame mixer...",
105 |       "debug_pygame_quit": "退出 Pygame...",
106 |       "debug_pygame_close_error": "关闭 pygame 时出错: {0}",
107 |       "debug_destroy_error": "Error during root destroy: {0}",
108 |       "debug_dpi_error": "设置 DPI 感知失败: {0}",
109 |       "debug_init_pygame_ok": "pygame mixer 初始化成功。",
110 |       "debug_init_pygame_fail": "pygame mixer 初始化失败: {0}",
111 |       "debug_delete_file": "删除旧音频文件: {0}",
112 |       "debug_delete_fail_permission": "删除文件 {0} 失败 (可能正在使用): {1}",
113 |       "debug_delete_fail_os": "删除文件 {0} 失败: {1}",
114 |       "debug_delete_busy": "播放器可能正忙，延迟删除: {0}",
115 |       "debug_manage_files_error": "文件管理出错: {0}",
116 |       "debug_file_copied": "文件已复制到剪贴板 (CF_HDROP): {0}",
117 |       "debug_copy_fail": "复制文件到剪贴板失败: {0}",
118 |       "lang_button_text": "中 / En",
119 |       "status_lang_changed": "✅ 语言已切换至 {0}",
120 |       "lang_name_zh": "中文",
121 |       "lang_name_en": "English"
122 |     },
123 |     "en": {
124 |       "window_title": "Anki-TTS-Edge (Custom Ver. v1.6)",
125 |       "input_text_label": "📝 Input Text:",
126 |       "tab_voices": "🔊 Voices",
127 |       "tab_settings": "⚙️ Settings",
128 |       "tab_appearance": "🎨 Appearance",
129 |       "filter_language_placeholder": "Filter language (e.g., en)...",
130 |       "voice_list_label_1": "Voice List 1",
131 |       "voice_list_label_2": "Voice List 2",
132 |       "refresh_voices_button": "🔄 Refresh Voice List",
133 |       "rate_label": "Rate:",
134 |       "volume_label": "Volume:",
135 |       "settings_output_cache_label": "Output & Cache",
136 |       "settings_copy_label": "🔗 Auto-copy File",
137 |       "settings_autoplay_label": "▶️ Autoplay",
138 |       "settings_max_files_label": "🔢 Max Cache Files:",
139 |       "settings_clipboard_label"
8000
: "Clipboard Features",
140 |       "settings_enable_ctrl_c_label": "🖱️Enable Copy-to-Audio Float (Ctrl+C)",
141 |       "settings_enable_selection_label": "Enable Select-to-Audio Float (Mouse)",
142 |       "appearance_theme_label": "Interface Theme:",
143 |       "appearance_mode_light": "Light Mode",
144 |       "appearance_mode_dark": "Dark Mode",
145 |       "appearance_color_label": "Custom Color (Hex):",
146 |       "appearance_pick_color_button": "🎨",
147 |       "appearance_apply_color_button": "Apply Color",
148 |       "generate_button": "Generate Audio",
149 |       "status_ready": "✅ Ready",
150 |       "status_error": "❌ Error",
151 |       "status_success": "✅ Success",
152 |       "status_generating": "⏳ Generating audio",
153 |       "status_getting_voices": "⏳ Getting voice list...",
154 |       "status_voices_updated": "✅ Voice list updated",
155 |       "status_monitor_enabled": "✅ Clipboard/Selection monitor enabled",
156 |       "status_monitor_disabled": "❌ Clipboard/Selection monitor disabled",
157 |       "status_file_copied": "✅ File copied to clipboard",
158 |       "status_copy_failed": "❌ Failed to copy file",
159 |       "status_play_error": "❌ Playback error",
160 |       "status_generate_error": "❌ Audio generation failed",
161 |       "status_empty_text_error": "❌ Error: Please enter text",
162 |       "status_no_voice_error": "❌ Error: Please select a voice",
163 |       "status_settings_saved": "✅ Settings saved",
164 |       "status_settings_save_failed": "❌ Failed to save settings",
165 |       "status_mixer_init_error": "❌ Player init error",
166 |       "status_file_not_found": "❌ Error: File not found",
167 |       "error_invalid_color_title": "Invalid Color",
168 |       "error_invalid_color_message": "Please enter a valid 6-digit hex color code (e.g., #FF5733), not '{0}'",
169 |       "error_mixer_init_title": "Initialization Error",
170 |       "error_mixer_init_message": "Failed to initialize audio player (pygame): {0}\nAutoplay feature will be unavailable.",
171 |       "debug_voices_loaded": "Found {0} voices, categorized by Language->Region->Name.",
172 |       "debug_voices_load_failed": "Failed to get voice list: {0}",
173 |       "debug_run_async_voices_error": "Error running async get voices task: {0}",
174 |       "debug_voice_ui_updated": "DEBUG: Voice UI updated. Current Voice: {0}",
175 |       "debug_voice_ui_no_data": "DEBUG: No voice data received.",
176 |       "debug_voice_load_failed_ui": "Failed to load",
177 |       "debug_no_matching_voices": "No matching voices",
178 |       "debug_voice_map_empty": "DEBUG: voice_display_to_full_map is empty in _populate_inline_voice_list",
179 |       "debug_voice_selected": "DEBUG _select_voice_inline: Selected {0}",
180 |       "debug_apply_color": "Applied custom color: {0}",
181 |       "debug_load_settings_failed": "Failed to load settings: {0}",
182 |       "debug_invalid_color_loaded": "Warning: Loaded invalid custom color '{0}', using default.",
183 |       "debug_settings_saved": "Settings saved.",
184 |       "debug_settings_save_failed": "Failed to save settings: {0}",
185 |       "debug_generating_audio": "Preparing audio generation: voice='{0}', rate='{1}', volume='{2}', pitch='{3}'",
186 |       "debug_output_path": "Output path: {0}",
187 |       "debug_generate_thread_error": "Error running async generation task: {0}",
188 |       "debug_edge_tts_ok": "Edge TTS audio generated successfully: {0}",
189 |       "debug_edge_tts_fail": "Edge TTS audio generation failed: {0}",
190 |       "debug_edge_tts_internal_error": "Edge TTS internal error",
191 |       "debug_empty_text": "Text is empty, cannot generate audio",
192 |       "debug_audio_complete": "Audio generation complete: {0}",
193 |       "debug_autoplay_manual": "DEBUG: Autoplay (manual)? {0}",
194 |       "debug_autoplay_float": "DEBUG: Autoplay (float)? {0}",
195 |       "debug_autocopy_float": "DEBUG: Autocopy (float)? {0}",
196 |       "debug_pygame_play_call": "DEBUG: play_audio_pygame called with path: {0}",
197 |       "debug_pygame_stop_current": "DEBUG: Stopping current playback...",
198 |       "debug_pygame_load_play": "DEBUG: Attempting to load and play: {0}",
199 |       "debug_pygame_play_start": "DEBUG: Pygame playback started.",
200 |       "debug_pygame_play_error": "ERROR: Pygame playback error: {0}",
201 |       "debug_play_unknown_error": "ERROR: Unknown error during playback: {0}",
202 |       "debug_float_trigger_text": "Triggering generation from float: {0}...",
203 |       "debug_no_float_text": "ERROR: No text found for float trigger.",
204 |       "debug_monitor_start": "Starting clipboard and mouse monitor...",
205 |       "debug_poll_thread_start": "DEBUG: Clipboard polling thread started.",
206 |       "debug_initial_paste_error": "DEBUG: Error getting initial paste for monitor start: {0}",
207 |       "debug_initial_paste_set": "DEBUG: Initial clipboard content set: {0}...",
208 |       "debug_new_clipboard_content": "Detected new clipboard content: {0}...",
209 |       "debug_poll_mouse_pos": "Poll detected change, mouse position: {0}",
210 |       "debug_poll_trigger_error": "Error triggering float window from poll: {0}",
211 |       "debug_poll_clipboard_error": "Clipboard access error (ignored): {0}",
212 |       "debug_poll_generic_error": "Clipboard monitor error: {0}",
213 |       "debug_poll_thread_stop": "Clipboard polling thread stopped.",
214 |       "debug_mouse_pressed": "DEBUG: Mouse pressed at {0}",
215 |       "debug_mouse_released": "DEBUG: Mouse released. Dist: {0:.2f}, Duration: {1:.2f}s",
216 |       "debug_selection_detected": "Mouse selection end detected...",
217 |       "debug_simulate_copy": "DEBUG: Simulating Ctrl+C...",
218 |       "debug_simulate_copy_complete": "DEBUG: Ctrl+C Simulation complete.",
219 |       "debug_simulate_copy_error": "ERROR: Failed to simulate Ctrl+C: {0}",
220 |       "debug_mouse_release_error": "Error in mouse release logic: {0}",
221 |       "debug_mouse_listener_stop_error": "Error stopping mouse listener: {0}",
222 |       "debug_mouse_listener_thread_stop": "Mouse listener thread stopped.",
223 |       "debug_monitor_stop": "Stopping clipboard and mouse monitor...",
224 |       "debug_closing": "Window closing...",
225 |       "debug_pygame_stop_mixer": "Stopping Pygame mixer...",
226 |       "debug_pygame_quit": "Quitting Pygame...",
227 |       "debug_pygame_close_error": "Error closing pygame: {0}",
228 |       "debug_destroy_error": "Error during root destroy: {0}",
229 |       "debug_dpi_error": "Failed to set DPI awareness: {0}",
230 |       "debug_init_pygame_ok": "pygame mixer initialized successfully.",
231 |       "debug_init_pygame_fail": "pygame mixer initialization failed: {0}",
232 |       "debug_delete_file": "Deleted old audio file: {0}",
233 |       "debug_delete_fail_permission": "Failed to delete file {0} (in use?): {1}",
234 |       "debug_delete_fail_os": "Failed to delete file {0}: {1}",
235 |       "debug_delete_busy": "Player might be busy, delaying delete: {0}",
236 |       "debug_manage_files_error": "Error managing audio files: {0}",
237 |       "debug_file_copied": "File copied to clipboard (CF_HDROP): {0}",
238 |       "debug_copy_fail": "Failed to copy file to clipboard: {0}",
239 |       "lang_button_text": "中 / En",
240 |       "status_lang_changed": "✅ Language changed to {0}",
241 |       "lang_name_zh": "中文",
242 |       "lang_name_en": "English"
243 |     }
244 |   }
  </content>
</file>

<environment_details>
# VSCode Visible Files
Anki-TTS-Edge.py

# VSCode Open Tabs
Anki-TTS-Edge.py,Anki-TTS-Edge copy.py

# Current Time
2025/4/6 下午10:26:56 (Asia/Shanghai, UTC+8:00)

# Current Context Size (Tokens)
40,832 (4%)

# Current Cost
$0.00

# Current Mode
<slug>code</slug>
<name>Code</name>
</environment_details>

[TASK RESUMPTION] This task was interrupted 11 minutes ago. It may or may not be complete, so please reassess the task context. Be aware that the project state may have changed since then. The current working directory is now 'd:/软件编写/Anki-TTS-Edge'. If the task has not been completed, retry the last step before interruption and proceed with completing the task.

Note: If you previously attempted a tool use that the user did not provide a result for, you should assume the tool use was not successful and assess whether you should retry. If the last tool was a browser_action, the browser has been closed and you must launch a new browser if needed.

<environment_details>
# VSCode Visible Files


# VSCode Open Tabs
Anki-TTS-Edge.py

# Current Time
2025/4/6 下午10:39:00 (Asia/Shanghai, UTC+8:00)

# Current Context Size (Tokens)
40,832 (4%)

# Current Cost
$0.00

# Current Mode
<slug>code</slug>
<name>Code</name>


# Current Working Directory (d:/软件编写/Anki-TTS-Edge) Files
.gitignore
Anki-TTS-Edge.py
Anki-TTS-Edge.spec
icon.ico
LICENSE
NOTICE
README.md
requirements_backup.txt
requirements.txt
translations.json
.git/
.git/AUTO_MERGE
.git/COMMIT_EDITMSG
.git/config
.git/description
.git/FETCH_HEAD
.git/HEAD
.git/index
.git/ORIG_HEAD
.git/REBASE_HEAD
.git/hooks/
.git/info/
.git/logs/
.git/objects/
.roo/
.venv/
音频/
build/
dist/
github-edge_tts/

(File list truncated. Use list_files on specific subdirectories if you need to explore further.)
</environment_details>

Additional Context

Image
@msjsc001 msjsc001 added the bug Something isn't working label Apr 6, 2025
@hannesrudolph
Copy link
Collaborator

Are you using a custom prompt or memory bank system?

Your response data says "[TASK RESUMPTION] This task was interrupted 11 minutes ago. It may or may not be complete, so please reassess the task context. Be aware that the project state may have changed since then. The current working directory is now 'd:/软件编写/Anki-TTS-Edge'. If the task has not been completed, retry the last step before interruption and proceed with completing the task." which seems that the prompting is causing problems.

@hannesrudolph hannesrudolph moved this from New to Issue [Needs Info] in Roo Code Roadmap Apr 7, 2025
@hannesrudolph hannesrudolph added the needs information Needs more information to be actionable label Apr 7, 2025
@QaDoPL
Copy link
QaDoPL commented Apr 8, 2025

Same issue here, Gemini2.5, open router and it start stuck before editing file. I tried many time.
prompt original.
DIFF - enabled and experimental search.

Image

@mattparlane
Copy link

Just chiming in, same issue here. Fairly consistent on large files. Happy to help debug if you can give me some pointers.

@hrdflt
Copy link
hrdflt commented Apr 9, 2025 •
edited
Loading

+1 for the same thing happening here exclusively with Gemini 2.5 (gemini-2.5-pro-preview-03-25)

Conditions under which this happens:

  • Loading up the context window as much as possible with documentation and code (50-90% of the 1M tokens on the sliding graph)
  • Either Architect or Code mode -- It's finally going to "do" the work after you agree to the proposal it made, whether its writing an MD file in architect mode, or writing markdown/code in Code mode.

The behavior for me is that it will:

  1. Start doing the work, writing anywhere from 5-50 lines of ANYTHING (code or markdown)
  2. Freezes, mid write (in the diff file).
  3. Pressing cancel does nothing. You need to "Close Window" in VScode and re-open.
  4. Attempting to resume the task yields the same result, in an almost identical position in the diff/file
  5. Re-attempting the same task from scratch yields the exact same result more or less

@hannesrudolph
Copy link
Collaborator

+1 for the same thing happening here exclusively with Gemini 2.5 (gemini-2.5-pro-preview-03-25)

Conditions under which this happens:

  • Loading up the context window as much as possible with documentation and code (50-90% of the 1M tokens on the sliding graph)
  • Either Architect or Code mode -- It's finally going to "do" the work after you agree to the proposal it made, whether its writing an MD file in architect mode, or writing markdown/code in Code mode.

The behavior for me is that it will:

  1. Start doing the work, writing anywhere from 5-50 lines of ANYTHING (code or markdown)
  2. Freezes, mid write (in the diff file).
  3. Pressing cancel does nothing. You need to "Close Window" in VScode and re-open.
  4. Attempting to resume the task yields the same result, in an almost identical position in the diff/file
  5. Re-attempting the same task from scratch yields the exact same result more or less

Can you pleaes share a little more about your setup? Are you running any sort of memory bank? Is this still happening with the updates we have made since you posted this? We have been focusing heavily on improving the diff editing as we want to make sure this works for you!

@msjsc001
Copy link
Author
msjsc001 commented Apr 15, 2025 •
edited
Loading

Are you using a custom prompt or memory bank system?

Your response data says "[TASK RESUMPTION] This task was interrupted 11 minutes ago. It may or may not be complete, so please reassess the task context. Be aware that the project state may have changed since then. The current working directory is now 'd:/软件编写/Anki-TTS-Edge'. If the task has not been completed, retry the last step before interruption and proceed with completing the task." which seems that the prompting is causing problems.

@hannesrudolph Hi,I've found that this issue occurs regardless of whether custom or default prompts are used. The fundamental reason is API interruption, which happens easily, especially with long contexts. Sometimes, the API might only return partial information, stopping mid-sentence without finishing, which also leads to the process getting stuck. At other times, factors like network instability might also cause this problem. Essentially, the core issue is that when the API response is returned to Roo, Roo cannot recognize that it has been interrupted. I have still been encountering this frequently recently. Would it be possible to add settings for users to specify the time interval and number of attempts for automatic retries upon API response interruption? (An API response interruption typically means that after inputting text, it stops generating further output, or there is simply no response at all). This should provide a good solution to this problem.

@hannesrudolph
Copy link
Collaborator

@mrubens thoughts?

@scpedicini
Copy link
scpedicini commented Apr 16, 2025 •
edited
Loading

For me the context size was relatively small (14k tokens) against Gemini 2.5 Pro Preview and it just gets stuck and spins endlessly. The Roo Output window doesn't show anything meaningful and it never times out either, just spins at "API Request...". Anecdotal but seems to occur more when transitioning from Architect to Code.

In some situations, Cancel + Resume has been able to get it unstuck thankfully.

@hannesrudolph
Copy link
Collaborator

What is happening at the time it spins? What is Roo trying to do at that time?

@msjsc001
Copy link
Author

Based on my own experience, Roo is usually in the middle of modifying code at this time, but if the API is interrupted, it will keep waiting for the API's information. At this point, only the user actively canceling can stop its waiting.

@github-project-automation github-project-automation bot moved this from Issue [Needs Info] to Done in Roo Code Roadmap May 1, 2025
@hannesrudolph
Copy link
Collaborator

cant repro

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
bug Something isn't working needs information Needs more information to be actionable
Projects
Roo Code Roadmap
Archived in project
Milestone
No milestone
Development

No branches or pull requests
6 participants
@mattparlane @scpedicini @hannesrudolph @hrdflt @msjsc001 @QaDoPL
0