12 karma · joined June 1, 2019
Although I haven't considered creating a framework for a simple linter, but WindTerm and WindEdit are originally one during development. In fact, there are four kits in the unreleased tool: terminal (WindTerm), editor (WindEdit), browser and file manager (including local filer like windows explorer and remote filer like Sftp client), They complement each other and form an efficient tool.
For example, WindEdit has a special function that can read virtual huge files. This feature is especially designed for remote files. Imagine that there is a very large file on the server that can be read and written directly without downloading it. Just like playing a video online, you only need to get a small piece of content currently read and written. That would be very efficient, especially when reading huge logs.
I don't think putting a stub on the server is the correct solution, because we cannot put stubs on every server.
I will continue to work hard to build better tools and hope you can use them happily.
You can modify the configuration file to highlight any words you want to highlight.
And windterm repository will still exists on github, as I said, many code of it will be gradually released, like some library, classes of data structs, schemes, etc. That is why I created windterm repository on github. I just need time to clean these code when many new features are still waiting to be realized.
I love github and open source, but after experienced last failed open source project of mine, I won't release the whole windterm project any more, lots of feature requests and several tough issues make me exhausted.
I’m assuming this only kicks in if color is not applied inside the term?
You are right, if the text has color from server, the color will be respected and will not be styled again. Or many tools like vi will show a mess.
Being able to cat an xml/json/source file and get decent styling sounds quite cool.
Think a little more boldly. If we can piped the output of cat to a real text editor and let you view and modify and save it more easier:) Just pop up its content when you cat a file.
Have you thought of offering an embeddable component?
No, never. I think a good tool must guarantee its integrity. But the idea and method can be offered to others, like google introduced MapReduce and BigTable.
Perhaps offer the terminal for free and commercialize the component?
Windterm will always free for personal usage. I am very happy if someone will like to use my terminal. For companies, if they don't sell their products for money, they can use windterm for free too.
I believe this is a trend. Maybe one day just like IBM CEO Watson said he thought there was a world market "for maybe five computers".
For example, text editors have created a lot of schemes for python, java, c++, even ada, shell script and etc, so they can provide autocomplete, color styling, folding, lint and so on. But what have done for the systems? Do we have an autocomplete list for Ubuntu's commands of each version? Do we have a whole scheme for every command and every option of Windows's cmd or Macos's bash? How can terminals provide features without these?
What is the soul of a terminal? The soul is only lies in the text in them. If we want to click a traceback line in a terminal and then open the error source file and jump to the error line to instead of a lot redundant operations, copy the path, open it in an editor , scroll to the error line, we must need to know what does the tranceback mean. So everything I talked, we need a lot of people's help, just like they have helped text editors in past half of century.
When all this have done, every terminal can share the schemes library to provide good features, and everyone has a lot of good tools to choose. I think it is more important than I just open the windterm's source. I call this true open source too.
It's not shameful for me to talk about money if I can make a lot of people happier when they using a more useful terminal, but I do care the terminal more than the money. Money is only a byproduct, with or without it, I am still very happy if I make my terminal dreams come to true.
Everyone has their own selection, I respect and also keep mine. So don't worry, everything is just the beginning, I will try my best to create a good terminal, just like what I did in the past.
I have been developing windterm for many many years. Based on my data analytics, terminal only spend 2% of the time to show text if you use threads to separate the socket read and text display. So it doesn't matter if you use GPU or not. Of course, use GPU is a very good thing, especially it can display more smoothly and draw emoji characters correctly. But as you can see, windterm still has very high performance without GPU acceleration. GPU also has its disadvantage, most virtual machines on cloud have no GPU emulated, so you can not use Alacritty on them.
The last thing, coloring, This is why I have spent so much time creating windterm in the past many years. The emulator traditionally displays the colors from server, but you know only a few poor text has its color. So for a long time, terminals should depends ZSH or auto-color-ls plugin and etc. to provide additional colors to make text more readable.
But why do we have to endure this? Text editors have already know how to styling a code file or script and done it well for more than 40 years!
So it took me some time to implement a text editor, and then I use its styling engine to coloring the text in terminal. That's why you can see a gorgeous colorful terminal. Prompt, numbers, strings and so on all have its color while maintaining the color from server if they have.
We deserve to have a colorful world, even in terminals, is it not?