Adding the ability to differentiate builtin plugins when listing #3810

Neko-Box-Coder · 2025-07-20T12:22:00Z

Currently, it is quite difficult to tell which plugins are built-in and which ones are not when listing. For example:

initlua (0.0.0-unknown)
MicroOmni (0.4.1)
filemanager2 (1.3.0)
gitStatus (0.1.5)
autocomplete_tooltip (0.0.1)
misspell (0.2.0)
quoter (1.0.0)
autoclose (1.0.0)
comment (1.0.0)
diff (1.0.0)
ftoptions (1.0.0)
literate (1.0.0)
status (1.0.0)

With this change, it explicitly tells you which ones are builtin, like this

initlua (built-in) (0.0.0-unknown)
MicroOmni (0.4.1)
filemanager2 (1.3.0)
gitStatus (0.1.5)
autocomplete_tooltip (0.0.1)
misspell (0.2.0)
quoter (1.0.0)
autoclose (built-in) (1.0.0)
comment (built-in) (1.0.0)
diff (built-in) (1.0.0)
ftoptions (built-in) (1.0.0)
literate (built-in) (1.0.0)
status (built-in) (1.0.0)

JoeKar · 2025-07-21T17:12:29Z

I'd prefer to have (built-in) behind the version number (1.0.0).

...or...

Right now the version number of internal plugins is somehow useless, since we never update them and so they become superfluous.
At this point we could simply stop printing the version number of internal plugins and directly post-fix them with (built-in):

autoclose (built-in)
comment (built-in)
diff (built-in)
ftoptions (built-in)
literate (built-in)
status (built-in)

We can proceed printing their version number in case they're overwritten by the user.

Neko-Box-Coder · 2025-07-21T17:22:20Z

Right now the version number of internal plugins is somehow useless, since we never update them and so they become superfluous. At this point we could simply stop printing the version number of internal plugins and directly post-fix them with (built-in):
autoclose (built-in)
comment (built-in)
diff (built-in)
ftoptions (built-in)
literate (built-in)
status (built-in)

Sounds good to me.

We can proceed printing their version number in case they're overwritten by the user.

wdym exactly? That's possible for the internal plugins?

JoeKar · 2025-07-21T17:32:07Z

wdym exactly? That's possible for the internal plugins?

We should print the version number in case ~/.config/micro/plug/comment overwrites the built-in one and the respective flag (Default) isn't set.

BTW:
Isn't Builtin the better flag name than Default?

Neko-Box-Coder · 2025-07-21T17:40:05Z

We should print the version number in case ~/.config/micro/plug/comment overwrites the built-in one and the respective flag (Default) isn't set.

I see, if it gets overridden, it will be treated as user plugin right? So the version will be printed but the (built-in) text won't be there. Unless the built-in version of the plugin also comes up when we do -plugin list, which I don't know, will try that later.

BTW:
Isn't Builtin the better flag name than Default?

I don't mind changing it to Builtin but I am just keeping consistent with plugin.go because Default is used there which gets propagated to the plugin installer

JoeKar · 2025-07-21T19:02:35Z

I see, if it gets overridden, it will be treated as user plugin right?

Yes.

I don't mind changing it to Builtin but I am just keeping consistent with plugin.go because Default is used there which gets propagated to the plugin installer

I see and yes, we should keep them in sync.

Neko-Box-Coder · 2025-07-21T20:30:06Z

I see and yes, we should keep them in sync.

Yes as in keep it as it is or yes as in change them both to Builtin? Can't tell which one you meant 😂

JoeKar · 2025-07-22T16:21:24Z

Builtin is somehow more self explaining than Default...at least for me. It wouldn't require the additional comment.

Neko-Box-Coder · 2025-07-27T11:03:12Z

This is what the latest changes look like. I overrided "comment" with a different version just to see how it acts, seems good to me.

❯ ./micro -plugin list
The following plugins are currently installed:
initlua (built-in)
MicroOmni (0.4.1)
comment (1.1.0)
filemanager2 (1.4.0)
gitStatus (0.1.5)
autocomplete_tooltip (0.0.1)
misspell (0.2.0)
quoter (1.0.0)
autoclose (built-in)
diff (built-in)
ftoptions (built-in)
literate (built-in)
status (built-in)

internal/action/actions.go

dmaluka · 2025-07-28T21:45:13Z

internal/config/plugin.go

@@ -71,7 +71,7 @@ type Plugin struct {
 	Info    *PluginInfo   // json file containing info
 	Srcs    []RuntimeFile // lua files
 	Loaded  bool
-	Default bool // pre-installed plugin
+	Builtin bool          // pre-installed plugin


gofmt

And, what @JoeKar said in #3810 (comment) when advocating for renaming it: "It wouldn't require the additional comment."?

...IMHO Default is a fine name. But since you guys are eager to rename it, I'm fine with that too.

dmaluka · 2025-07-28T22:00:14Z

% ./micro -plugin list
The following plugins are currently installed:
initlua (built-in)
...

initlua is built-in?

Neko-Box-Coder · 2025-07-28T22:09:00Z

initlua is built-in?

It comes with micro, so I guess it is? I am not sure, it's kinda a special case.

dmaluka · 2025-07-28T22:10:49Z

It doesn't come with micro, it is created by the user, isn't it?

Neko-Box-Coder · 2025-07-28T22:21:56Z

It doesn't come with micro, it is created by the user, isn't it?

Ah, didn't notice I have an empty init.lua in my config.

Okay, if I set Builtin to false, it will show initlua (0.0.0-unknown) instead.

What should the version text be?

Neko-Box-Coder · 2025-07-28T22:26:11Z

Or I can do a special check for initlua when listing the plugin, in that case do we just show initlua?

dmaluka · 2025-07-28T22:26:16Z

It neither is built-in nor has a version, so we shouldn't show either?

Neko-Box-Coder · 2025-07-28T22:28:25Z

Something like this now:

❯ ./micro -plugin list
The following plugins are currently installed:
initlua
MicroOmni (0.4.1)
filemanager2 (1.4.0)
gitStatus (0.1.5)
autocomplete_tooltip (0.0.1)
misspell (0.2.0)
quoter (1.0.0)
autoclose (built-in)
comment (built-in)
diff (built-in)
ftoptions (built-in)
literate (built-in)
status (built-in)

dmaluka · 2025-07-28T22:39:34Z

internal/config/plugin_installer.go

@@ -665,7 +670,7 @@ func PluginCommand(out io.Writer, cmd string, args []string) {
 		for _, plugin := range args {
 			// check if the plugin exists.
 			for _, p := range Plugins {
-				if p.Name == plugin && p.Default {
+				if p.Name == plugin && p.Builtin {


And if this is initlua?

Although, this part is already buggy without this PR.

True, I have just updated initlua to be builtin then, should error out now.

So, is it built-in or not built-in?

We are introducing mess where we could easily avoid it? For example, we could just add an explicit check for "initlua"?

It neither is built-in nor has a version, so we shouldn't show either?

Actually I was thinking the same, but it has an advantage to print its presence:
You see that there might be one more callback routine affecting micro's behavior.

This we can then request directly within the issue template (when we use the new GitHub template mechanism after micro's move).

It neither is built-in nor has a version, so we shouldn't show either?

Actually I was thinking the same, but it has an advantage to print its presence: You see that there might be one more callback routine affecting micro's behavior.

I thought @dmaluka meant only the version (bracketed) text. If @dmaluka meant the whole entry like you said then I agree with you @JoeKar , I would prefer it being shown and not hidden away from the user.

Yeah, I meant only the bracketed text.

dmaluka · 2025-07-29T19:33:19Z

internal/config/plugin_installer.go

@@ -665,7 +670,7 @@ func PluginCommand(out io.Writer, cmd string, args []string) {
 		for _, plugin := range args {
 			// check if the plugin exists.
 			for _, p := range Plugins {
-				if p.Name == plugin && p.Default {
+				if (p.Name == plugin && p.Builtin) || p.Name == "initlua" {
 					fmt.Fprintln(out, "Default plugins cannot be removed, but can be disabled via settings.")


If we are renaming "default" to "built-in" all over the place anyway, don't we update to update this log message as well?

dmaluka · 2025-07-29T19:49:40Z

internal/config/plugin_installer.go

@@ -665,7 +670,7 @@ func PluginCommand(out io.Writer, cmd string, args []string) {
 		for _, plugin := range args {
 			// check if the plugin exists.
 			for _, p := range Plugins {
-				if p.Name == plugin && p.Default {
+				if (p.Name == plugin && p.Builtin) || p.Name == "initlua" {


Now, when successfully removing any plugin, micro prints "Default plugins cannot be removed..."

Although the plugin still ends up being successfully removed, thanks to continue (BTW I'm not sure why we have continue instead of break here).

What about UpdatePlugins()?

Now, when successfully removing any plugin, micro prints "Default plugins cannot be removed..."

Although the plugin still ends up being successfully removed, thanks to continue (BTW I'm not sure why we have continue instead of break here).

yeah but we do print out the plugins we have removed, not the best UI/UX but it works. Trying to not rewrite things unless necessary in the PR to not get sidetracked. I have changed continue to break tho.

2. What about UpdatePlugins()?

Okay yeah, just updated that as well.

Oh god. I was really hoping I wouldn't have to point out the obvious. It prints "Default plugins cannot be removed..." because you for some reason put || p.Name == "initlua" outside the parentheses.

(And before you rush to move the parentheses, another obvious thing: it is completely pointless to check p.Name == "initlua" at each iteration of the loop. We can just check plugin == "initlua" before entering the loop.)

So, now that you changed continue to break (which I mentioned just as a side note), without fixing the issue that you've introduced, as a result it not just needlessly says "Default plugins cannot be removed" but also fails to remove any plugins (and honestly says "No plugins removed"). Which you should have noticed if you tested your changes.

Ooops, my bad. Got a bit tired and mixed up with similar change in the UpdatePlugins while splitting the commit into two.

Should be good now.

dmaluka · 2025-07-29T19:51:10Z

internal/config/plugin_installer.go

@@ -687,7 +692,13 @@ func PluginCommand(out io.Writer, cmd string, args []string) {
 		plugins := GetInstalledVersions(false)
 		fmt.Fprintln(out, "The following plugins are currently installed:")
 		for _, p := range plugins {
-			fmt.Fprintf(out, "%s (%s)\n", p.Pack().Name, p.Version)
+			if p.Pack().Name == "initlua" {
+				fmt.Fprintf(out, "initlua\n")


nit: just to keep it in line with the below printfs, we could fmt.Fprintf(out, "%s\n", p.Pack().Name)

dmaluka · 2025-07-30T00:01:14Z

internal/config/plugin_installer.go

 			// check if the plugin exists.
 			for _, p := range Plugins {
 				if p.Name == plugin && p.Builtin {
 					fmt.Fprintln(out, "Built-in plugins cannot be removed, but can be disabled via settings.")
-					continue
+					break


If you really want to change it to break right away, what about the other continue below, in the successful case?

And before you rush to change it to break as well: have you investigated what exactly would be the consequences of these changes? What if there actually are multiple plugins with the same name? We probably don't want to allow that, but do we already prevent that? And if we don't, shouldn't we first of all implement preventing that? (In particular, a special case: we probably want to disallow ~/.config/micro/plug/initlua/, since the name initlua is reserved for init.lua? Do we already do that?)

If you don't want to investigate that now, just leave this continue alone.

... If you don't want to investigate that now, just leave this continue alone.

It's fine, I reverted the change. I thought it was an innocent change when you pointed it out initially.

And before you rush to change it to break as well: ...

This sounds quite harsh, to be honest. I did something by mistake earlier in this PR (as well as not testing things out) doesn't mean you should put it in this way, twice.

I understand those mistakes are quite dumb, but still, I don't like being talked down to like that. Stating any concerns/questions/corrections plainly is enough.

What if there actually are multiple plugins with the same name?

Anyway yes, we should leave it as continue then since if an unlikely situation of a plugin that shares the same name as the built in ones, then the user can still be able to remove it.

cutelisp · 2025-07-30T00:22:21Z

internal/config/plugin_installer.go

-				if p.Name == plugin && p.Default {
-					fmt.Fprintln(out, "Default plugins cannot be removed, but can be disabled via settings.")
+				if p.Name == plugin && p.Builtin {
+					fmt.Fprintln(out, "Built-in plugins cannot be removed, but can be disabled via settings.")


When running -plugin remove comment literate, this message is printed multiple times, which looks a bit odd IMO.

Why we show a generic error message ? Wouldn't it be better to specify which of the provided plugins are built-in?

Also, it would be helpful to display a clearer message when trying to remove a plugin that isn't installed? e.g. Unknown plugin "asd" (similar to the behavior of -plugin install asd).

Why we show a generic error message ? Wouldn't it be better to specify which of the provided plugins are built-in?

Yeah, I can specify the plugin name in the message but it dooesn't really bother me too much since it tells me what plugins are removed at the end anyway.

I can add it, maybe something like "xxxx is a built-in plugin which cannot be removed, but can be disabled via settings."?

I can add it, maybe something like "xxxx is a built-in plugin which cannot be removed, but can be disabled via settings."?

Looks good.

Nit: Should "Built-in plugins cannot be removed, but can be disabled via settings." include a period? Looking at other messages in the plugin manager, some have a period and some don't looks it's a bit inconsistent.

Side-quests

There are a couple fmt.Fprintln calls in plugin_installer.go with an extra space. example

Enhancing the final output of -plugin install * to explicitly list which plugins were installed (similar to -plugin remove) would be a good idea?

Nit: Should "Built-in plugins cannot be removed, but can be disabled via settings." include a period? Looking at other messages in the plugin manager, some have a period and some don't looks it's a bit inconsistent.

One might argue that this message is rather long (while others a rather short), so the period is useful to see that the sentence finally ends...

There are a couple fmt.Fprintln calls in plugin_installer.go with an extra space. example

Fixed in 094b02d

Enhancing the final output of -plugin install * to explicitly list which plugins were installed (similar to -plugin remove) would be a good idea?

To be fixed, if anyone is willing to.

Neko-Box-Coder force-pushed the ShowBuiltinPlugins branch from 424aefa to d058948 Compare July 27, 2025 11:02

JoeKar reviewed Jul 27, 2025

View reviewed changes

internal/action/actions.go Outdated Show resolved Hide resolved

Neko-Box-Coder force-pushed the ShowBuiltinPlugins branch from d058948 to e5e07e6 Compare July 27, 2025 18:19

JoeKar approved these changes Jul 28, 2025

View reviewed changes

dmaluka reviewed Jul 28, 2025

View reviewed changes

Neko-Box-Coder force-pushed the ShowBuiltinPlugins branch from e5e07e6 to b645494 Compare July 28, 2025 22:27

dmaluka reviewed Jul 28, 2025

View reviewed changes

Neko-Box-Coder force-pushed the ShowBuiltinPlugins branch 2 times, most recently from 131e8b5 to d3b42b6 Compare July 29, 2025 18:09

dmaluka reviewed Jul 29, 2025

View reviewed changes

Adding the ability to differenitate builtin plugins when listing

1096c4f

Neko-Box-Coder force-pushed the ShowBuiltinPlugins branch 2 times, most recently from 3951d7c to 4bd40c7 Compare July 29, 2025 23:26

dmaluka reviewed Jul 30, 2025

View reviewed changes

Adding explicit checks for initlua for different plugin commands

21edb89

Neko-Box-Coder force-pushed the ShowBuiltinPlugins branch from 4bd40c7 to 21edb89 Compare July 30, 2025 00:16

cutelisp reviewed Jul 30, 2025

View reviewed changes

Updating remove plugin message to specify plugin name

177f4d5

JoeKar approved these changes Aug 19, 2025

View reviewed changes

dmaluka merged commit 7a252f4 into zyedidia:master Aug 20, 2025
6 checks passed

Adding the ability to differentiate builtin plugins when listing #3810

Adding the ability to differentiate builtin plugins when listing #3810

Uh oh!

Conversation

Neko-Box-Coder commented Jul 20, 2025

Uh oh!

JoeKar commented Jul 21, 2025

Uh oh!

Neko-Box-Coder commented Jul 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

JoeKar commented Jul 21, 2025

Uh oh!

Neko-Box-Coder commented Jul 21, 2025

Uh oh!

JoeKar commented Jul 21, 2025

Uh oh!

Neko-Box-Coder commented Jul 21, 2025

Uh oh!

JoeKar commented Jul 22, 2025

Uh oh!

Neko-Box-Coder commented Jul 27, 2025

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dmaluka commented Jul 28, 2025

Uh oh!

Neko-Box-Coder commented Jul 28, 2025

Uh oh!

dmaluka commented Jul 28, 2025

Uh oh!

Neko-Box-Coder commented Jul 28, 2025

Uh oh!

Neko-Box-Coder commented Jul 28, 2025

Uh oh!

dmaluka commented Jul 28, 2025

Uh oh!

Neko-Box-Coder commented Jul 28, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

JoeKar Jul 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Neko-Box-Coder Jul 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Neko-Box-Coder Jul 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Neko-Box-Coder commented Jul 21, 2025 •

edited

Loading

JoeKar Jul 29, 2025 •

edited

Loading

Neko-Box-Coder Jul 29, 2025 •

edited

Loading

Neko-Box-Coder Jul 29, 2025 •

edited

Loading