Replies: 1 comment
-
Such a mechanism for 'problem-solution' pairs already exists as possible quick-fixes for problem markers: And Eclipse already contains many quick-fixes, for example for Java developing: They are implemented in the traditional way, but everyone interested can create an Eclipse plugin that uses AI to create quick-fixes for problems and can for example publish it at the Eclipse market-place: https://marketplace.eclipse.org/ |
Beta Was this translation helpful? Give feedback.
-
**This is one of the issues I mentioned earlier where the developer answered some of my questions and then invited me to give some options for the user experience. I thought that all these options might not be very good, so I suddenly came up with a new idea. Here are some of my answers to that question, I think it's a brilliant idea. This idea could definitely push Eclipse itself into a better platform for applications and development. But I think the cost of this idea will be huge, so feel free to discuss it.
I think it is not necessary to Eclipse itself to do this work , third-party developers can also go to do - just rely on Eclipse traditional plug-in development , then the traditional and AI will touch sparks , Eclipse will be reborn in blood!**
Below is My reply for my issue
I think the first solution you proposed is better, this way will have a better user experience, disallowing certain operations in advance will reduce unnecessary redundant operations, i.e. if the user operates a certain number of steps before they can see that they can't export both the plugin and the feature at the same time, then the user may wonder why they can't export both at the same time.
This leads to a new question, what if the user creates a project that wants to export both plugins and features at the same time? So a better way is in the user to create the project stage, give the user a choice of intent, that is, through the configurable pop-up window to show the user to create the project supports all the exportable configuration options, or just pop-up window prompts all the exportable configuration options, so that the user will be in advance to control their own project can be exported to what things, to give control to the user rather than the programmer's later maintenance.
I don't think Eclipse's internal errors are what users or developers want to know about, it's just an issue that Eclipse developers want to collect, so I don't think it's desirable to throw out errors in any way, even though every plugin has its own internal errors, Eclipse and these plugin vendors shouldn't be able to show them to the user by way of error alerts.
Although I know Eclipse itself is not a specialised product like MyEclipse, it still has product attributes of its own, Eclipse is designed for developers, but developers are also users, and with the current advances in AI, developers can now become more of a black-box user, so a developer using Eclipse would probably want a more more automated experience , so will not pay attention to Eclipse and third-party plug-ins within the error prompts , but hope to have a direct solution to the problem , or even say that I hope Eclipse itself through the background tasks to provide some solutions will be directly solved or give some tips to give some automated configuration . I think you can collect those bugs submitted by past users and use AI technology to achieve this automation function I proposed. Problem-solution data pairs, if there is some kind of problem, then through Eclipse internal AI, give the reason for this problem and possible solutions. It would also be possible to expose interfaces to third party plugin developers, allowing them to also use Eclipse's internal AI to train their own plugin's problem-solution data pair models, just like ChatGPT allows multiple users to have their own models. This is my simple idea, just for reference.
Beta Was this translation helpful? Give feedback.
All reactions