Feedback on scratch/igc Branch

Feedback on scratch/igc Branch

[Eval EXEC]

Hello,

Thank you for your excellent work on the scratch/igc branch. Over the
past few days, I have been using this branch extensively for various
tasks including reading and sending mail with mu4e, programming with
lsp-mode and rust-analyzer , and managing my notes using org-roam and
org-journal. I am pleased to report that the scratch/igc branch is
remarkably stable and performs exceptionally well.

Could you please share the next steps for the scratch/igc branch?
Additionally, do you have any estimated timeline for when it will be
merged into the master branch?

I greatly appreciate your efforts and am looking forward to the future
developments. 

Best regards,

Eval Exec
-- 

Re: Feedback on scratch/igc Branch

[Eli Zaretskii]

> From: Eval EXEC <execvy@gmail.com>
> Date: Sun, 08 Sep 2024 13:26:17 +0800
> 
> Thank you for your excellent work on the scratch/igc branch. Over the
> past few days, I have been using this branch extensively for various
> tasks including reading and sending mail with mu4e, programming with
> lsp-mode and rust-analyzer , and managing my notes using org-roam and
> org-journal. I am pleased to report that the scratch/igc branch is
> remarkably stable and performs exceptionally well.

Thanks, this is very good news.

> Could you please share the next steps for the scratch/igc branch?

Keep using the branch on the various platforms and different build
configurations, to reveal the issues that are still left to be handled
and fixed.  Also, we need better GC debugging tools and means, to ease
the investigation of problems we bump into.

> Additionally, do you have any estimated timeline for when it will be
> merged into the master branch?

It's IMO too early to talk about landing this on master.  We still
have unresolved issues, like what to do with signals delivered to
Emacs when MPS is collecting garbage and the arena is locked.  We also
need to decide whether we want to fork MPS and make changes in it on
our own or not.

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Eli Zaretskii <eliz@gnu.org> [2024-09-08 08:47]:
> It's IMO too early to talk about landing this on master.  We still
> have unresolved issues, like what to do with signals delivered to
> Emacs when MPS is collecting garbage and the arena is locked.  We also
> need to decide whether we want to fork MPS and make changes in it on
> our own or not.

I see no scratch/igc branch, how do I get it?

Jean

Re: Feedback on scratch/igc Branch

[Eli Zaretskii]

> Date: Wed, 23 Oct 2024 16:22:11 +0300
> From: Jean Louis <bugs@gnu.support>
> Cc: Eval EXEC <execvy@gmail.com>, emacs-devel@gnu.org
> 
> * Eli Zaretskii <eliz@gnu.org> [2024-09-08 08:47]:
> > It's IMO too early to talk about landing this on master.  We still
> > have unresolved issues, like what to do with signals delivered to
> > Emacs when MPS is collecting garbage and the arena is locked.  We also
> > need to decide whether we want to fork MPS and make changes in it on
> > our own or not.
> 
> I see no scratch/igc branch, how do I get it?

"git pull" followed by "git checkout scratch/igc".

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Eli Zaretskii <eliz@gnu.org> [2024-10-23 20:28]:
> > Date: Wed, 23 Oct 2024 16:22:11 +0300
> > From: Jean Louis <bugs@gnu.support>
> > Cc: Eval EXEC <execvy@gmail.com>, emacs-devel@gnu.org
> > 
> > * Eli Zaretskii <eliz@gnu.org> [2024-09-08 08:47]:
> > > It's IMO too early to talk about landing this on master.  We still
> > > have unresolved issues, like what to do with signals delivered to
> > > Emacs when MPS is collecting garbage and the arena is locked.  We also
> > > need to decide whether we want to fork MPS and make changes in it on
> > > our own or not.
> > 
> > I see no scratch/igc branch, how do I get it?
> 
> "git pull" followed by "git checkout scratch/igc".

$ git checkout scratch/igc
error: pathspec 'scratch/igc' did not match any file(s) known to git

I cannot see that branch.

Re: Feedback on scratch/igc Branch

[Philip Kaludercic]

Jean Louis <bugs@gnu.support> writes:

> * Eli Zaretskii <eliz@gnu.org> [2024-10-23 20:28]:
>> > Date: Wed, 23 Oct 2024 16:22:11 +0300
>> > From: Jean Louis <bugs@gnu.support>
>> > Cc: Eval EXEC <execvy@gmail.com>, emacs-devel@gnu.org
>> > 
>> > * Eli Zaretskii <eliz@gnu.org> [2024-09-08 08:47]:
>> > > It's IMO too early to talk about landing this on master.  We still
>> > > have unresolved issues, like what to do with signals delivered to
>> > > Emacs when MPS is collecting garbage and the arena is locked.  We also
>> > > need to decide whether we want to fork MPS and make changes in it on
>> > > our own or not.
>> > 
>> > I see no scratch/igc branch, how do I get it?
>> 
>> "git pull" followed by "git checkout scratch/igc".
>
> $ git checkout scratch/igc
> error: pathspec 'scratch/igc' did not match any file(s) known to git
>
> I cannot see that branch.

Perhaps a "git fetch --all" could help?

-- 
	Philip Kaludercic on siskin

Re: Feedback on scratch/igc Branch

[Eli Zaretskii]

> Date: Wed, 23 Oct 2024 23:40:54 +0300
> From: Jean Louis <bugs@gnu.support>
> Cc: emacs-devel@gnu.org
> 
> > > I see no scratch/igc branch, how do I get it?
> > 
> > "git pull" followed by "git checkout scratch/igc".
> 
> $ git checkout scratch/igc
> error: pathspec 'scratch/igc' did not match any file(s) known to git
> 
> I cannot see that branch.

Then your clone of the Emacs Git repository is incomplete, for some
reason.

Re: Feedback on scratch/igc Branch

[tomas]

[-- Attachment #1: Type: text/plain, Size: 1103 bytes --]

On Thu, Oct 24, 2024 at 09:51:22AM +0300, Eli Zaretskii wrote:
> > Date: Wed, 23 Oct 2024 23:40:54 +0300
> > From: Jean Louis <bugs@gnu.support>
> > Cc: emacs-devel@gnu.org
> > 
> > > > I see no scratch/igc branch, how do I get it?
> > > 
> > > "git pull" followed by "git checkout scratch/igc".
> > 
> > $ git checkout scratch/igc
> > error: pathspec 'scratch/igc' did not match any file(s) known to git
> > 
> > I cannot see that branch.
> 
> Then your clone of the Emacs Git repository is incomplete, for some
> reason.

The error message gives me the impression that git is trying to
interpret "scratch/igc" as a file path and not as a branch. So
that might be due to some uncommon/unexpected configuration on
OP's side.

Jean-Louis: what does the command

  git branch -a

say on your side?

You might have more luck by stating explicitly which branch you
want to track with

  git checkout -b scratch/igc --track origin/scratch/igc

(I'm now assuming that your upstream repo is called "origin" on
your side, which is convention, but no law).

HTH
-- 
tomás

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Philip Kaludercic <philipk@posteo.net> [2024-10-24 01:19]:
> > I cannot see that branch.
> 
> Perhaps a "git fetch --all" could help?

I did that, and I cannot see igc branch with git branch -r

-- 
Jean

Take action in Free Software Foundation campaigns:
https://www.fsf.org/campaigns

✡️🛡️ Proudly standing with Israel, a nation rooted in history and culture. Let's condemn hatred and promote understanding.

In support of Richard M. Stallman
https://stallmansupport.org/

Re: Feedback on scratch/igc Branch

[Jean Louis]

[-- Attachment #1: Type: text/plain, Size: 1540 bytes --]

* tomas@tuxteam.de <tomas@tuxteam.de> [2024-10-24 10:48]:
> The error message gives me the impression that git is trying to
> interpret "scratch/igc" as a file path and not as a branch. So
> that might be due to some uncommon/unexpected configuration on
> OP's side.
> 
> Jean-Louis: what does the command
> 
>   git branch -a
> 
> say on your side?

*   remotes/origin/scratch/hard-narrow
*   remotes/origin/scratch/highlight-n-windows
*   remotes/origin/scratch/isearch-show-toggles
*   remotes/origin/scratch/jit-lock-antiblink
*   remotes/origin/scratch/jit-lock-antiblink-cleaned-up
*   remotes/origin/scratch/joaot/make-completion-at-point-function

That is where I expected "igc" to be alphabetically sorted, is not there.

I am using git from git://git.sv.gnu.org/emacs.git

> You might have more luck by stating explicitly which branch you
> want to track with
> 
>   git checkout -b scratch/igc --track origin/scratch/igc
> 
> (I'm now assuming that your upstream repo is called "origin" on
> your side, which is convention, but no law).


$ git checkout -b scratch/igc --track origin/scratch/igc
fatal: 'origin/scratch/igc' is not a commit and a branch 'scratch/igc' cannot be created from it


-- 
Jean

Take action in Free Software Foundation campaigns:
https://www.fsf.org/campaigns

✡️🛡️ Proudly standing with Israel, a nation rooted in history and culture. Let's condemn hatred and promote understanding.

In support of Richard M. Stallman
https://stallmansupport.org/

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: Feedback on scratch/igc Branch

[tomas]

[-- Attachment #1: Type: text/plain, Size: 1156 bytes --]

On Thu, Oct 24, 2024 at 10:14:24PM +0300, Jean Louis wrote:
> * tomas@tuxteam.de <tomas@tuxteam.de> [2024-10-24 10:48]:
> > The error message gives me the impression that git is trying to
> > interpret "scratch/igc" as a file path and not as a branch. So
> > that might be due to some uncommon/unexpected configuration on
> > OP's side.
> > 
> > Jean-Louis: what does the command
> > 
> >   git branch -a
> > 
> > say on your side?
> 
> *   remotes/origin/scratch/hard-narrow
> *   remotes/origin/scratch/highlight-n-windows
> *   remotes/origin/scratch/isearch-show-toggles
> *   remotes/origin/scratch/jit-lock-antiblink
> *   remotes/origin/scratch/jit-lock-antiblink-cleaned-up
> *   remotes/origin/scratch/joaot/make-completion-at-point-function
> 
> That is where I expected "igc" to be alphabetically sorted, is not there.

Hm. But it seems to be there:

  tomas@trotzki:~$ git ls-remote --heads git://git.sv.gnu.org/emacs.git | grep igc
  d864c26b207362b4591dc821fe8bae1394c61281        refs/heads/scratch/igc

For some reason your clone isn't fetching it, At the moment I'm
out of ideas on why.

Cheers
-- 
tomás

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: Feedback on scratch/igc Branch

[Andreas Schwab]

On Okt 24 2024, Jean Louis wrote:

> That is where I expected "igc" to be alphabetically sorted, is not there.

What's your fetch spec?

$ git config get --all remote.origin.fetch

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Re: Feedback on scratch/igc Branch

[Eli Zaretskii]

> From: Andreas Schwab <schwab@linux-m68k.org>
> Cc: emacs-devel@gnu.org
> Date: Thu, 24 Oct 2024 22:26:13 +0200
> 
> On Okt 24 2024, Jean Louis wrote:
> 
> > That is where I expected "igc" to be alphabetically sorted, is not there.
> 
> What's your fetch spec?
> 
> $ git config get --all remote.origin.fetch

I think you meant

  $ git config --get-all remote.origin.fetch

Re: Feedback on scratch/igc Branch

[Andreas Schwab]

On Okt 25 2024, Eli Zaretskii wrote:

> I think you meant
>
>   $ git config --get-all remote.origin.fetch

That is deprecated.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Re: Feedback on scratch/igc Branch

[Eli Zaretskii]

> From: Andreas Schwab <schwab@linux-m68k.org>
> Cc: tomas@tuxteam.de,  emacs-devel@gnu.org
> Date: Fri, 25 Oct 2024 10:06:02 +0200
> 
> On Okt 25 2024, Eli Zaretskii wrote:
> 
> > I think you meant
> >
> >   $ git config --get-all remote.origin.fetch
> 
> That is deprecated.

Maybe so, but versions of Git to which at least I have access (the
latest being 2.34.1) reject the format you posted:

  $ git config get --all remote.origin.fetch
  error: key does not contain a section: get

So I thought the deprecated format will have a better luck.

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Andreas Schwab <schwab@linux-m68k.org> [2024-10-24 23:28]:
> On Okt 24 2024, Jean Louis wrote:
> 
> > That is where I expected "igc" to be alphabetically sorted, is not there.
> 
> What's your fetch spec?
> 
> $ git config get --all remote.origin.fetch

$ git config get --all remote.origin.fetch
error: key does not contain a section: get

-- 
Jean

Take action in Free Software Foundation campaigns:
https://www.fsf.org/campaigns

✡️🛡️ Proudly standing with Israel, a nation rooted in history and culture. Let's condemn hatred and promote understanding.

In support of Richard M. Stallman
https://stallmansupport.org/

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Eli Zaretskii <eliz@gnu.org> [2024-10-25 09:26]:
> > From: Andreas Schwab <schwab@linux-m68k.org>
> > Cc: emacs-devel@gnu.org
> > Date: Thu, 24 Oct 2024 22:26:13 +0200
> > 
> > On Okt 24 2024, Jean Louis wrote:
> > 
> > > That is where I expected "igc" to be alphabetically sorted, is not there.
> > 
> > What's your fetch spec?
> > 
> > $ git config get --all remote.origin.fetch
> 
> I think you meant
> 
>   $ git config --get-all remote.origin.fetch

+refs/heads/feature/integration-of-dictionary-el

Re: Feedback on scratch/igc Branch

[Eli Zaretskii]

> Date: Fri, 1 Nov 2024 22:31:06 +0300
> From: Jean Louis <bugs@gnu.support>
> Cc: Andreas Schwab <schwab@linux-m68k.org>, tomas@tuxteam.de,
>   emacs-devel@gnu.org
> 
> * Eli Zaretskii <eliz@gnu.org> [2024-10-25 09:26]:
> > > From: Andreas Schwab <schwab@linux-m68k.org>
> > > Cc: emacs-devel@gnu.org
> > > Date: Thu, 24 Oct 2024 22:26:13 +0200
> > > 
> > > On Okt 24 2024, Jean Louis wrote:
> > > 
> > > > That is where I expected "igc" to be alphabetically sorted, is not there.
> > > 
> > > What's your fetch spec?
> > > 
> > > $ git config get --all remote.origin.fetch
> > 
> > I think you meant
> > 
> >   $ git config --get-all remote.origin.fetch
> 
> +refs/heads/feature/integration-of-dictionary-el

Your clone includes only a single branch, so it's a small wonder you
don't see others.  I suggest to remove it and clone again, this time
using the command shown on the project's Savannah page.

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Eli Zaretskii <eliz@gnu.org> [2024-11-01 23:03]:
> Your clone includes only a single branch, so it's a small wonder you
> don't see others.  I suggest to remove it and clone again, this time
> using the command shown on the project's Savannah page.

Alright, I do not know how it happened in past.

I have now made it right and it seems that igc does not work, when I do M-x igc-stats, I see nothing.

Maybe it was not compiled, which outside library is needed for it?

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Eli Zaretskii <eliz@gnu.org> [2024-11-01 23:04]:
> > +refs/heads/feature/integration-of-dictionary-el
> 
> Your clone includes only a single branch, so it's a small wonder you
> don't see others.  I suggest to remove it and clone again, this time
> using the command shown on the project's Savannah page.

Alright, I have now got new Emacs git and then installed mps library,
and now is building I hope with the igc.

Jean

Re: Feedback on scratch/igc Branch

[Jean Louis]

* Eli Zaretskii <eliz@gnu.org> [2024-11-01 23:04]:
> Your clone includes only a single branch, so it's a small wonder you
> don't see others.  I suggest to remove it and clone again, this time
> using the command shown on the project's Savannah page.

Now it works well, I am on igc branch, and I can see M-x igc-stats 

Just by feeling it looks kind of faster, I do not see any more much
garbage collecting, I had that always turned on.

Jean

Re: Feedback on scratch/igc Branch

[Jean Louis]

I can see weird things happening after long session. I do not know if it relates to this branch or not.

I was doing a lot of minibuffer prompts, and then I could see I cannot write in Emacs, it looks like something is kind of preventing me. Truly like a poltergeist effect.

I write normally, and fast, but when I write letters get changed, like instead of u letter I get i letter, and such. I am blind typing, but still. I have seen words being replaced with other words. I was demonstrating here in house what is going on.

Then I was thinking it could be keyboard, so I turned it off and on.

But every writing in other windows was without disturbance, just within Emacs not working.

Then I removed frame.

Then opened new frame, this time keyboard was not working at all inside of Emacs. But mouse worked, but what do I do with mouse?

So I tried saving buffers, nothing from keyboard worked, and I pkilled Emacs.

Now I am updating to see new version.

It seems something is wrong with memory. I cannot know. If I span -q Emacs now to write this, it works well.

Jean

Re: Feedback on scratch/igc Branch

[Jean Louis]

I see when some error in my script prevent loading, attempting to load it 2-3 times, removed Emacs from memory.

Re: Feedback on scratch/igc Branch

[Jean Louis]

For some while I have to stop using igc and try the master branch
again.

Each time over the night I put computer in sleep mode. Next following
day, I restart using same Emacs session, but soon I can see there are
delays, as when I type I have to wait longer time that key get
registered.

I am fast typer, when I do M-x kill-emacs I know it will be M-x
kill-emacs, but I have seen something like `M-x kli-l' in the
minibuffer, and that was not my typo.

It means that one key like "k" get registered, but "i" is delayed for
some reason, and "l" get displayed before the "i" but I wrote "i"
eearler than "l", then again dash get displayed before the second "l".

This is because of some delays internally, which slow down Emacs,
there is lag of maybe half second something for the key to appear.

As I am suspecting that igc branch is culprit, I am now moving back to
master branch to see if that same behavior would come again in few
days.

Sleep mode of computer could be related to it, because those delays I
do not see late evening before it goes to sleep mode.

Jean