I would like to help about denite.nvim HOT 14 CLOSED

You can test denite.nvim.
The customization feature is not implemented though.

My easy test:

home directory 137042 files

:Denite file_rec: 0.7902[s](The initialization until user input time)
unite.vim: N/A(freeze UI)
CtrlP: Almost 5~6[s]
fzf: Almost same time with :Denite

Note: This is no cache result. The cache feature is not implemented yet.

from denite.nvim.

I think the best way to help shougo is create a project such as Shougo-s-plugin,show more about why use Shougo's plugin ,and how to config to let vim works better，and I think it is time for create a organization for Dark Powered vimmer

from denite.nvim.

Thank you. I have not much time to develop it currently.
Because, I need fix/support for other plugins.
Can you support for me?

from denite.nvim.

I can do my best to provide support for your plugins. I am on github often and can reply to issues regularly. What would you like for me to do?

from denite.nvim.

OK. For example, the issue.
deoplete-plugins/deoplete-jedi#13

I cannot reproduce it.

from denite.nvim.

Help is great.

But as for reproducing bugs. I mean this python-client stuff is driving me nuts. How do you begin to debug stuff like this?

Do you think python-client gets in the way? Do you have any ideas for how neovim could help debugging. Why does plugins for neovim have to be so hard, when other IDE's do it easy?

Sublime 3 uses python 3 and there's no need to worry about messaging. Plugins just work.

https://www.sublimetext.com/docs/3/api_reference.html

from denite.nvim.

But as for reproducing bugs. I mean this python-client stuff is driving me nuts. How do you begin to debug stuff like this?

It may be harder. But I think it is same in external interface plugins in Vim.
For example, YouCompleteMe.

Sublime 3 uses python 3 and there's no need to worry about messaging. Plugins just work.

You can do it. I think current neovim development team has not the power to support it.

from denite.nvim.

debugging ability was improved slightly with neovim/pynvim#169 (specifically for async handlers) but there still are quite a few gotchas that make plugin development harder that it should be. Most importantly documentation (well it exists, but spread out and hard to find). I hope to improve README.md the coming days.

from denite.nvim.

I can't wait to switch to asynchronous denite! From unite I mostly use buffers, file_rec/async, tags and grep sources. Lately, however, I'm using fzf :Files since unite is much slower.

@Shougo Besides the functionality, one thing that I wish would be there in denite is testing. Being a big fan of unite and using master branch on a daily basis, what often discouraged me from using the latest commit is that things occasionally get broken. I witnessed numerous issues about broken grep source. Fortunately, Shougo is exceptionally quick and good at fixing them. I do not mean to tell you how to write software (you are true master of vim and plugins for it), but maybe more efforts could be put into thoroughly testing denite? Of course this will inevitably slow down the development speed, but at the same time it will increase stability for end users. Just my humble thoughts.

from denite.nvim.

@Shougo I just tested and denite printed this:

[denite] candidates len = 25022
[denite] 1.6065280437469482

On second call:

[denite] candidates len = 25022
[denite] 0.3120131492614746

BTW.- I am loving the new UX.

from denite.nvim.

Oh, thanks.
Because my SSD speed is faster.

BTW.- I am loving the new UX.

:-)
It will be more better than unite.

from denite.nvim.

@Shougo Besides the functionality, one thing that I wish would be there in denite is testing. Being a big fan of unite and using master branch on a daily basis, what often discouraged me from using the latest commit is that things occasionally get broken. I witnessed numerous issues about broken grep source. Fortunately, Shougo is exceptionally quick and good at fixing them. I do not mean to tell you how to write software (you are true master of vim and plugins for it), but maybe more efforts could be put into thoroughly testing denite? Of course this will inevitably slow down the development speed, but at the same time it will increase stability for end users. Just my humble thoughts.

Yes, I know the reason.
It is hard to change the development model.
The faster development is important for me.

from denite.nvim.

Oh, thanks.
Because my SSD speed is faster.

Yes. I think so, too.

Would be good if you shared how is planned some UX and we can start coding some sources.

from denite.nvim.

The specification is not fixed...

from denite.nvim.