Testing of the portable dumper branch

Testing of the portable dumper branch

[John Wiegley]

Has anyone had a chance to play around with the portable dumper branch, and to
review the changes?  I'd love to hear some thoughts.

-- 
John Wiegley                  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com                          60E1 46C4 BD1A 7AC1 4BA2

Re: Testing of the portable dumper branch

[Philippe Vaucher]

[-- Attachment #1: Type: text/plain, Size: 250 bytes --]

> Has anyone had a chance to play around with the portable dumper branch,
> and to
> review the changes?  I'd love to hear some thoughts.
>

Sorry if I missed it, but which one is it?
http://git.savannah.gnu.org/gitweb/?p=emacs.git;a=heads

Philippe

[-- Attachment #2: Type: text/html, Size: 641 bytes --]

Re: Testing of the portable dumper branch

[ravi desai]

[-- Attachment #1: Type: text/plain, Size: 478 bytes --]

>
>
> > Has anyone had a chance to play around with the portable dumper branch,
> > and to
> > review the changes?  I'd love to hear some thoughts.
> >
>
> Sorry if I missed it, but which one is it?
> http://git.savannah.gnu.org/gitweb/?p=emacs.git;a=heads
>

I too am interested in checking out the portable dumper.

As a side-note, what do we usually do about old branches? There are
branches
on that page saying "13 years ago".  Do we delete them or just leave them
hanging?

[-- Attachment #2: Type: text/html, Size: 907 bytes --]

Re: Testing of the portable dumper branch

[Philippe Vaucher]

[-- Attachment #1: Type: text/plain, Size: 595 bytes --]

>
> As a side-note, what do we usually do about old branches? There are
> branches
> on that page saying "13 years ago".  Do we delete them or just leave them
> hanging?
>

My guess is that most people would be ok to get rid of them, but don't care
enough to argue about it when faced with the inevitable arguments like "but
it does not hurt to keep them", "it might prove useful one day" or "it has
historical value" :-)

My take on this is would be to do a giant cleanup (delete all irrelevant
branches), but archive these beforehand somewhere "just in case" to be on
the safe side.

Philippe

[-- Attachment #2: Type: text/html, Size: 1119 bytes --]

Re: Testing of the portable dumper branch

[Andreas Schwab]

On Dez 21 2016, Philippe Vaucher <philippe.vaucher@gmail.com> wrote:

> My take on this is would be to do a giant cleanup (delete all irrelevant
> branches), but archive these beforehand somewhere "just in case" to be on
> the safe side.

We archive them under old-branches.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Testing of the portable dumper branch

[John Wiegley]

>>>>> "PV" == Philippe Vaucher <philippe.vaucher@gmail.com> writes:

VP> Sorry if I missed it, but which one is it?
VP> http://git.savannah.gnu.org/gitweb/?p=emacs.git;a=heads

Daniel, I'm not seeing your dumper branch either.  What name did you give to
it?  I was expecting "pdumper" or some such...

-- 
John Wiegley                  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com                          60E1 46C4 BD1A 7AC1 4BA2

Re: Testing of the portable dumper branch

[Stefan Monnier]

>> My take on this is would be to do a giant cleanup (delete all irrelevant
>> branches), but archive these beforehand somewhere "just in case" to be on
>> the safe side.
> We archive them under old-branches.

Exactly.  I'm more annoyed at the branches we have in the top-level and
which I believe should be moved to old-branches (such as the
test-concurrency one I queried about on this list recently, with no
answer).

But if someone's worried about its impact on the total size of a "git
clone", we could move old-branches to another repository, tho I expect
those branches use up a negligible amount of space.


        Stefan

Re: Testing of the portable dumper branch

[John Wiegley]

>>>>> "SM" == Stefan Monnier <monnier@iro.umontreal.ca> writes:

SM> But if someone's worried about its impact on the total size of a "git
SM> clone", we could move old-branches to another repository, tho I expect
SM> those branches use up a negligible amount of space.

Or, move them under a different ref from "heads" on the server, i.e.:

    git push origin refs/heads/old-branches/foo:refs/old-branches/foo

Then it will be invisible by default, and only fetchable if you add
refs/old-branches to your refspec.

-- 
John Wiegley                  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com                          60E1 46C4 BD1A 7AC1 4BA2

Re: Testing of the portable dumper branch

[Eli Zaretskii]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Date: Wed, 21 Dec 2016 14:25:19 -0500
> 
> the test-concurrency one I queried about on this list recently, with
> no answer

Don't worry about that branch.  It exists for just 2 weeks, and will
disappear when the time comes.  There are much older branches in the
repository that might need some attention.

Re: Testing of the portable dumper branch

[Stefan Monnier]

>> the test-concurrency one I queried about on this list recently, with
>> no answer
> Don't worry about that branch.  It exists for just 2 weeks, and will
> disappear when the time comes.  There are much older branches in the
> repository that might need some attention.

Indeed, `test-concurrency` is the easy case because it's fresh so we
still have a good idea of its status.  I'm more worried about
the others.


        Stefan

Re: Testing of the portable dumper branch

[Phillip Lord]

On Wed, December 21, 2016 8:02 pm, Stefan Monnier wrote:
>>> the test-concurrency one I queried about on this list recently, with
>>> no answer
>> Don't worry about that branch.  It exists for just 2 weeks, and will
>> disappear when the time comes.  There are much older branches in the
>> repository that might need some attention.
>
> Indeed, `test-concurrency` is the easy case because it's fresh so we
> still have a good idea of its status.  I'm more worried about the others.
>

I have loads of feature branches. I'll delete them in the New Year.

Phil

Re: Testing of the portable dumper branch

[Daniel Colascione]

On Tue, Dec 20 2016, John Wiegley wrote:
> Has anyone had a chance to play around with the portable dumper branch, and to
> review the changes?  I'd love to hear some thoughts.

I've found that pdumper does awful things to weak hash tables restored
from dump.  I'm fixing that bug --- I'll push when that bit works.

Re: Testing of the portable dumper branch

[Philippe Vaucher]

[-- Attachment #1: Type: text/plain, Size: 704 bytes --]

> On Tue, Dec 20 2016, John Wiegley wrote:
> > Has anyone had a chance to play around with the portable dumper branch,
> and to
> > review the changes?  I'd love to hear some thoughts.
>
> I've found that pdumper does awful things to weak hash tables restored
> from dump.  I'm fixing that bug --- I'll push when that bit works.
>

...So, any news?

Or maybe are you waiting on the result of the "big elc file" approach
before deciding to continue with the portable dumper idea?

I mean, with all the discussions the portable dumper generated it's kind of
a shame not to be able to test it, or at least look at the current
implementation in some repo/branch, even if it's broken at the moment.

Philippe

[-- Attachment #2: Type: text/html, Size: 1072 bytes --]

Re: Testing of the portable dumper branch

[John Wiegley]

>>>>> "PV" == Philippe Vaucher <philippe.vaucher@gmail.com> writes:

VP> I mean, with all the discussions the portable dumper generated it's kind
VP> of a shame not to be able to test it, or at least look at the current
VP> implementation in some repo/branch, even if it's broken at the moment.

Pinging on this again... Daniel?

-- 
John Wiegley                  GPG fingerprint = 4710 CF98 AF9B 327B B80F
http://newartisans.com                          60E1 46C4 BD1A 7AC1 4BA2