[Feature]: UI界面问题

[java2shadow-ui]

Problem or Motivation

再听课的时候无法沉浸式观看PPT，讨论那个模块能不能隐藏？或者可以做成弹幕啊

Proposed Solution

关于讨论这块可以隐藏

Alternatives Considered

No response

Area

Classroom generation

Additional Context

No response

[YizukiAme]

Hi @java2shadow-ui 👋

This is actually already on the official roadmap for the next release — you can check out Discussion #150 for the full plan, which covers both Presentation Mode and Discussion UX enhancements.

On the presentation/immersive side, I've already submitted PR #133, which adds a fullscreen Presentation Mode — header auto-hides, both sidebars (including the discussion panel) collapse, slide fills the screen, controls auto-fade after idle, plus keyboard navigation. It partially addresses what you're looking for. If you'd like, you can pull and try that branch to see if it meets your needs.

Also, the underlying feature request has been tracked since #102 — you might want to follow that issue for ongoing updates from the maintainers.

[YizukiAme]

Another thing — the Danmaku/bullet-screen idea is cool! I actually considered it when working on #133 , but given the complexity around rendering consistency and edge cases, I went with a more fundamental and reliable approach: hiding most UI elements in fullscreen to ensure the lecture content gets maximum screen real estate.

That said, if there's enough community interest or the team decides to explore it, I (or the team) may pick it up in the future.
Stay tuned~

[java2shadow-ui]

我给你们团队发了邮件咨询关于授权商用的，我想把OpenMAIC封装到我自己的APP当中，应该怎么获得到授权呢？

[YizukiAme]

我给你们团队发了邮件咨询关于授权商用的，我想把OpenMAIC封装到我自己的APP当中，应该怎么获得到授权呢？

Hi! Quick update on the feature side — I didn't expect the team to settle on a direction so quickly, but here we are! I've already implemented what you were looking for in PR #195 today. In presentation mode, the discussion panel is fully hidden, and messages now appear as floating speech bubbles in the screen corners (teacher on the bottom-left, student/user on the bottom-right) with slide-in/out animations. Controls auto-hide after 3 seconds of idle, so you get a truly immersive PPT viewing experience! Hope this helps!

Also — it seems like you might have mistaken me for a team member, which is a big misunderstanding 😄 I'm just a random community contributor! If you need to reach the project maintainers, you can contact @cosarah and @wyuc directly here on GitHub. You could also try reaching out through the project's other community channels, or just wait for a reply to your email.

BTW, on a personal note regarding the licensing question — I took a look at the project's license, which is GNU Affero General Public License (AGPL-3.0). This license does permit commercial use, but it comes with very strict copyleft and open-source requirements. If you wrap, modify, or serve the software over a network (SaaS), you are required to release your entire application's source code under the same AGPL-3.0 license and provide users with access to the source. So I'd strongly recommend discussing licensing and authorization matters thoroughly with the team before proceeding with any commercial integration.

[java2shadow-ui]

好的谢谢

…

On Sat, Mar 21, 2026 at 8:41 PM Yizuki_Ame ***@***.***> wrote: *YizukiAme* left a comment (THU-MAIC/OpenMAIC#174) <#174 (comment)> 我给你们团队发了邮件咨询关于授权商用的，我想把OpenMAIC封装到我自己的APP当中，应该怎么获得到授权呢？ Hi! Quick update on the feature side — I didn't expect the team to settle on a direction so quickly, but here we are! I've already implemented what you were looking for in PR #195 <#195> today. In presentation mode, the discussion panel is fully hidden, and messages now appear as floating speech bubbles in the screen corners (teacher on the bottom-left, student/user on the bottom-right) with slide-in/out animations. Controls auto-hide after 3 seconds of idle, so you get a truly immersive PPT viewing experience! Hope this helps! Also — it seems like you might have mistaken me for a team member, which is a big misunderstanding 😄 I'm just a random community contributor! If you need to reach the project maintainers, you can contact @cosarah <https://github.com/cosarah> and @wyuc <https://github.com/wyuc> directly here on GitHub. You could also try reaching out through the project's other community channels, or just wait for a reply to your email. BTW, on a personal note regarding the licensing question — I took a look at the project's license, which is GNU Affero General Public License (AGPL-3.0). This license does permit commercial use, but it comes with very strict copyleft and open-source requirements. If you wrap, modify, or serve the software over a network (SaaS), you are required to release your entire application's source code under the same AGPL-3.0 license and provide users with access to the source. So I'd strongly recommend discussing licensing and authorization matters thoroughly with the team before proceeding with any commercial integration. — Reply to this email directly, view it on GitHub <#174?email_source=notifications&email_token=CAKN5R76GWCBTGDCDUMSYGT4R2EWRA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMJQGMZDMMBTGU4KM4TFMFZW63VHNVSW45DJN5XKKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-4103260358>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/CAKN5R33I22C2RZVS6363FT4R2EWRAVCNFSM6AAAAACWZQWUM6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DCMBTGI3DAMZVHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[fubaobao2023]

还有一点——弹幕/子弹屏的想法很棒！我在处理#133的时候其实也考虑过，但考虑到渲染一致性和特殊情况的复杂性，我最终选择了更基础、更可靠的方法：在全屏模式下隐藏大部分 UI 元素，以确保讲座内容占据最大的屏幕空间。

话虽如此，如果社区反响热烈，或者团队决定探索，我（或团队）未来可能会着手开发。 敬请期待~

还有一个就是老师这个脚色能否自定义,因为很多路演并非老师的脚色 也许是一个创业者的脚色 也许是一个分享会,其次工作流每次都有问答环节 这个环节可否自定义 因为根据场景可以自定选择是否要开启这个环节,第三能否生成PPTX之后能再次修改(老师)这个角色的语音回复或者文字,修改之后可以再次生成,最后能否支持下载完整的视频,因为这样就可以解决很多问题场景上的问题 比如脱离真人讲解,真人讲解毕竟在表演以及口才是都略有不足

[YizukiAme]

还有一点——弹幕/子弹屏的想法很棒！我在处理#133的时候其实也考虑过，但考虑到渲染一致性和特殊情况的复杂性，我最终选择了更基础、更可靠的方法：在全屏模式下隐藏大部分 UI 元素，以确保讲座内容占据最大的屏幕空间。
话虽如此，如果社区反响热烈，或者团队决定探索，我（或团队）未来可能会着手开发。 敬请期待~

还有一个就是老师这个脚色能否自定义,因为很多路演并非老师的脚色 也许是一个创业者的脚色 也许是一个分享会,其次工作流每次都有问答环节 这个环节可否自定义 因为根据场景可以自定选择是否要开启这个环节,第三能否生成PPTX之后能再次修改(老师)这个角色的语音回复或者文字,修改之后可以再次生成,最后能否支持下载完整的视频,因为这样就可以解决很多问题场景上的问题 比如脱离真人讲解,真人讲解毕竟在表演以及口才是都略有不足

---

Honestly, this is a bit beyond what I can answer as a random contributor, haha. But based on my understanding of the project and the team:

1. Custom roles — You can actually explicitly instruct the underlying LLM in the prompt to assign specific personas to the agents, or set the scenario context. Going forward, the team may also consider introducing customizable role templates — it's straightforward to implement and has been a frequently requested feature from the community. The default is set to "teacher" because the project is fundamentally a multi-agent classroom application, and the demo is tailored around that use case.

2. Optional Q&A session — This has indeed received quite a bit of feedback recently. I've already opened an issue to track and plan for a Solo mode. Depending on the team's direction, it may land in a future release — stay tuned!

3. Editing generated content — This is a pretty commonly asked question. I'm not heavily involved in that part of the codebase, but I'm confident the team is already aware of this direction!

4. Video download — Also a frequently asked question! In practice, this is technically very challenging — since this is fundamentally an interactive application rather than a linear video, it can't be straightforwardly exported as one (at least as things stand today). Your best bet for now would be to use a screen recorder.

[fubaobao2023]

第四点 可以临时生成一个外部访问链接,这个访问链接自定义一点个人的信息即可不展示其他的比如 名字,时间,或者公司名字 LOGO等, 基业-阿豪 ***@***.***

[fubaobao2023]

我们方便加一个即时通讯工具吗,以便我们高效沟通,我是在中国重庆的 基业-阿豪 ***@***.***

[fubaobao2023]

其实就是我在大模型里面已经定义了它的角色，也再三说了他的角色，但是整个这个老师这个角色，他在说的过程之中，他有时候在那个最后的一句话，他还是会带上，各位同学，或者说，某某教授，就是这个应该是被你们的工作流给给固定了

…

---原始邮件--- 发件人: ***@***.***&gt; 发送时间: 2026年3月23日(周一) 晚上7:01 收件人: ***@***.***&gt;; 抄送: ***@***.******@***.***&gt;; 主题: Re: [THU-MAIC/OpenMAIC] [Feature]: UI界面问题 (Issue #174) YizukiAme left a comment (THU-MAIC/OpenMAIC#174) 还有一点——弹幕/子弹屏的想法很棒！我在处理#133的时候其实也考虑过，但考虑到渲染一致性和特殊情况的复杂性，我最终选择了更基础、更可靠的方法：在全屏模式下隐藏大部分 UI 元素，以确保讲座内容占据最大的屏幕空间。 话虽如此，如果社区反响热烈，或者团队决定探索，我（或团队）未来可能会着手开发。 敬请期待~ 还有一个就是老师这个脚色能否自定义,因为很多路演并非老师的脚色 也许是一个创业者的脚色 也许是一个分享会,其次工作流每次都有问答环节 这个环节可否自定义 因为根据场景可以自定选择是否要开启这个环节,第三能否生成PPTX之后能再次修改(老师)这个角色的语音回复或者文字,修改之后可以再次生成,最后能否支持下载完整的视频,因为这样就可以解决很多问题场景上的问题 比如脱离真人讲解,真人讲解毕竟在表演以及口才是都略有不足 Honestly, this is a bit beyond what I can answer as a random contributor, haha. But based on my understanding of the project and the team: Custom roles — You can actually explicitly instruct the underlying LLM in the prompt to assign specific personas to the agents, or set the scenario context. Going forward, the team may also consider introducing customizable role templates — it's straightforward to implement and has been a frequently requested feature from the community. The default is set to "teacher" because the project is fundamentally a multi-agent classroom application, and the demo is tailored around that use case. Optional Q&amp;A session — This has indeed received quite a bit of feedback recently. I've already opened an issue to track and plan for a Solo mode. Depending on the team's direction, it may land in a future release — stay tuned! Editing generated content — This is a pretty commonly asked question. I'm not heavily involved in that part of the codebase, but I'm confident the team is already aware of this direction! Video download — Also a frequently asked question! In practice, this is technically very challenging — since this is fundamentally an interactive application rather than a linear video, it can't be straightforwardly exported as one (at least as things stand today). Your best bet for now would be to use a screen recorder. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***&gt;

[YizukiAme]

其实就是我在大模型里面已经定义了它的角色，也再三说了他的角色，但是整个这个老师这个角色，他在说的过程之中，他有时候在那个最后的一句话，他还是会带上，各位同学，或者说，某某教授，就是这个应该是被你们的工作流给给固定了
…

As a contributor, I cannot represent the official team for private discussions. For feature requests or business inquiries, please reach out to the core maintainer @wyuc directly, or join the project's official Feishu/Discord channel. Thanks!

[fubaobao2023]

好的 最后一个问题,飞书地址是多少 基业-阿豪 ***@***.*** 原始邮件 发件人：Yizuki_Ame ***@***.***&gt; 发件时间：2026年3月23日 19:53 收件人：THU-MAIC/OpenMAIC ***@***.***&gt; 抄送：fbb ***@***.***&gt;, Manual ***@***.***&gt; 主题：Re: [THU-MAIC/OpenMAIC] [Feature]: UI界面问题 (Issue #174) YizukiAme left a comment (THU-MAIC/OpenMAIC#174) 其实就是我在大模型里面已经定义了它的角色，也再三说了他的角色，但是整个这个老师这个角色，他在说的过程之中，他有时候在那个最后的一句话，他还是会带上，各位同学，或者说，某某教授，就是这个应该是被你们的工作流给给固定了 … As a contributor, I cannot represent the official team for private discussions. For feature requests or business inquiries, please reach out to the core maintainer @wyuc&nbsp;directly, or join the project's official Feishu channel. Thanks! — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.

[YizukiAme]

好的 最后一个问题,飞书地址是多少

基业-阿豪
@.***

     原始邮件

发件人：Yizuki_Ame @.>
发件时间：2026年3月23日 19:53
收件人：THU-MAIC/OpenMAIC @.>
抄送：fbb @.>, Manual @.>
主题：Re: [THU-MAIC/OpenMAIC] [Feature]: UI界面问题 (Issue #174)

YizukiAme left a comment (THU-MAIC/OpenMAIC#174)

其实就是我在大模型里面已经定义了它的角色，也再三说了他的角色，但是整个这个老师这个角色，他在说的过程之中，他有时候在那个最后的一句话，他还是会带上，各位同学，或者说，某某教授，就是这个应该是被你们的工作流给给固定了
…

As a contributor, I cannot represent the official team for private discussions. For feature requests or business inquiries, please reach out to the core maintainer @wyuc directly, or join the project's official Feishu channel. Thanks!

—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.

https://github.com/THU-MAIC/OpenMAIC/blob/main/community/feishu.md