Re: Issue 5099 - staging broken?

2017-03-26 Thread Thomas Morley
2017-03-26 13:30 GMT+02:00 David Kastrup :
> Thomas Morley  writes:
>
>> 2017-03-26 11:11 GMT+02:00 David Kastrup :
>>> Thomas Morley  writes:
>>>
 2017-03-26 10:06 GMT+02:00 David Kastrup :

> Backing out that patch from staging, and taking a further look.

 How to proceed?
 Upload a fixed patch for review?
>>>
>>> Probably.
>>
>> Meanwhile I tried to upload a fixed patch to
>> https://sourceforge.net/p/testlilyissues/issues/5101/
>> https://codereview.appspot.com/318680043/
>>
>> But I had closed the Rietveld-issue already, so I get from git-cl:
>> [...]
>> Issue creation errors: {'issue': ['This issue is closed (318680043)']}
>>
>> Nevertheless
>> https://github.com/rietveld-codereview/rietveld/wiki
>> says:
>> 'Closing the Issue
>> [...] An issue's owner can reopen the issue by unchecking the "Closed" box.'
>>
>> But I can't, the "Closed box" is not longer present.
>
> The "Close" box isn't.  I think editing the summary (possibly from the
> "post comments" page) also permitted changing the "closed" status.



Aaah, found it directly above the "Publish+Mail Comments ('m')
"-button. It's called "Edit Issue" or "Can't Edit" if you're not the
owner. No need to go to "post comments".

Though, meanwhile I created issue 5107. I think I stick to this one now.

Thanks,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread David Kastrup
Thomas Morley  writes:

> 2017-03-26 11:11 GMT+02:00 David Kastrup :
>> Thomas Morley  writes:
>>
>>> 2017-03-26 10:06 GMT+02:00 David Kastrup :
>>>
 Backing out that patch from staging, and taking a further look.
>>>
>>> How to proceed?
>>> Upload a fixed patch for review?
>>
>> Probably.
>
> Meanwhile I tried to upload a fixed patch to
> https://sourceforge.net/p/testlilyissues/issues/5101/
> https://codereview.appspot.com/318680043/
>
> But I had closed the Rietveld-issue already, so I get from git-cl:
> [...]
> Issue creation errors: {'issue': ['This issue is closed (318680043)']}
>
> Nevertheless
> https://github.com/rietveld-codereview/rietveld/wiki
> says:
> 'Closing the Issue
> [...] An issue's owner can reopen the issue by unchecking the "Closed" box.'
>
> But I can't, the "Closed box" is not longer present.

The "Close" box isn't.  I think editing the summary (possibly from the
"post comments" page) also permitted changing the "closed" status.

-- 
David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread Thomas Morley
2017-03-26 11:11 GMT+02:00 David Kastrup :
> Thomas Morley  writes:
>
>> 2017-03-26 10:06 GMT+02:00 David Kastrup :
>>
>>> Backing out that patch from staging, and taking a further look.
>>
>> How to proceed?
>> Upload a fixed patch for review?
>
> Probably.

Meanwhile I tried to upload a fixed patch to
https://sourceforge.net/p/testlilyissues/issues/5101/
https://codereview.appspot.com/318680043/

But I had closed the Rietveld-issue already, so I get from git-cl:
[...]
Issue creation errors: {'issue': ['This issue is closed (318680043)']}

Nevertheless
https://github.com/rietveld-codereview/rietveld/wiki
says:
'Closing the Issue
[...] An issue's owner can reopen the issue by unchecking the "Closed" box.'

But I can't, the "Closed box" is not longer present.


So if noone has a better hint I'll set 5101 to abandoned and do all
from scratch.

Cheers,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread James
On Sun, 26 Mar 2017 10:53:19 +0200
Thomas Morley  wrote:

> 2017-03-26 10:06 GMT+02:00 David Kastrup :
> > Thomas Morley  writes:
> >  
> >> 2017-03-26 0:48 GMT+01:00 James :  
> >>> Hello
> >>>
> >>> After 5099 was pushed I am now seeing problems with basic
> >>> 'make'.  
> >  
> >>
> >> accidently I permanently deleted the branch where the patch for
> >> issue 5099 was. But I had it as git-formated patch, so I reapplied
> >> it to a new branch and pushed from there.
> >>
> >> I've now retested make successfully, though.
> >>
> >> So I've no clue what's wrong.  
> >
> > Huh?
> >
> > diff --git a/scm/lily-library.scm b/scm/lily-library.scm
> > index c6f066ca32..1b3926af7f 100644
> > --- a/scm/lily-library.scm
> > +++ b/scm/lily-library.scm
> > @@ -778,9 +778,9 @@ as rectangular coordinates @code{(x-length .
> > y-length)}."
> >
> >  (define-public (remove-whitespace strg)
> >  "Remove characters satisfying @code{char-whitespace?} from string
> > @var{strg}"
> > -  (string-delete
> > -strg
> > -char-whitespace?))
> > +  (if (guile-v2)
> > +  (string-delete char-whitespace? strg)
> > +  (string-delete strg char-whitespace?))
> >
> >  (define-public (string-encode-integer i)
> >(cond
> >
> > The deleted lines have a net paren change of -1 (one paren more
> > closed than opened).  The added lines have a net paren change of 0.
> >
> > This cannot possibly have worked.  
> 
> Indeed.
> How could it survive automatic tests then?
> 
> >
> > But that's issue 5101 according to the commit message.  
> 
> Yep.
> 5099 looks still ok to me.
> 
> > Backing out that patch from staging, and taking a further look.  
> 
> How to proceed?
> Upload a fixed patch for review?
> 
> I'm terribly sorry for the additional work for you and James,
>   Harm
> 

It wasn't much additional work is exactly why we moved to a
staging branch workflow in the first place. You aren't the first to
'break staging' and you won't be the last and as of 'now' it looks like
it is all merged successfully.

James




___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread David Kastrup
Thomas Morley  writes:

> 2017-03-26 10:06 GMT+02:00 David Kastrup :
>
>> Backing out that patch from staging, and taking a further look.
>
> How to proceed?
> Upload a fixed patch for review?

Probably.

> I'm terribly sorry for the additional work for you and James,

Not much work in my case.  Just fired off lilypond-patchy-staging,
noticed it barfing and decided I was too lazy to look in the logs, so
instead looked at the commit diff.  And that was pretty obvious to
obsessive-compulsive paren-counters.

Don't count your parens until they are patchied.

-- 
David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread David Kastrup
Thomas Morley  writes:

> Sorry, for empty mail, misclicked.
> Obviously I need more coffee ...
>
> 2017-03-26 10:17 GMT+02:00 David Kastrup :
>> Thomas Morley  writes:
>>
>>> accidently I permanently deleted the branch where the patch for issue
>>> 5099 was.
>>
>> That would be a rather difficult feat (you don't do things like that
>> "accidentally" with Git since you have to work very hard to do them,
>
> There were too much unneeded branches in my local repro, mostly
> containing already pushed commits.
> So I decided to do some clean up with
> git branch -D branch-name
> and out of a c/p-error the wrong branch was taken.

Well, as I explained in the rest of the mail, that's only telling Git
your intent to have that branch deleted, and it removes the reference to
the branch.  But the material and the history of the material lingers on
long enough that you can reclaim it until either a lot of time or much
more dangerous commands have passed.

-- 
David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread Urs Liska


Am 26. März 2017 10:59:03 MESZ schrieb Thomas Morley :
>Sorry, for empty mail, misclicked.
>Obviously I need more coffee ...
>

Yep.
"Sometimes I make me a coffee in the morning - but usually coffee makes *me*.

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread Thomas Morley
Sorry, for empty mail, misclicked.
Obviously I need more coffee ...

2017-03-26 10:17 GMT+02:00 David Kastrup :
> Thomas Morley  writes:
>
>> accidently I permanently deleted the branch where the patch for issue
>> 5099 was.
>
> That would be a rather difficult feat (you don't do things like that
> "accidentally" with Git since you have to work very hard to do them,

There were too much unneeded branches in my local repro, mostly
containing already pushed commits.
So I decided to do some clean up with
git branch -D branch-name
and out of a c/p-error the wrong branch was taken.

> with a set of very explicit commands unless you are going to wait for
> several months as part of the recipe).
>
> Try
>
> git reflog
>
> and write down the hash for the first occurence of your branch in the
> list (which would be last in time).
>
> Then do
>
> git branch hash-that-you-wrote-down
>
> and the branch is back in life.  Of course, since this will be an
> unpublished branch, you have to do this in the repository where you
> "permanently" deleted the branch.
>
> Git tends to keep "permanently deleted" stuff around for 3 months or so
> before it really removes them for good.


Many thanks for the info.

Cheers,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread Thomas Morley
2017-03-26 10:06 GMT+02:00 David Kastrup :
> Thomas Morley  writes:
>
>> 2017-03-26 0:48 GMT+01:00 James :
>>> Hello
>>>
>>> After 5099 was pushed I am now seeing problems with basic 'make'.
>
>>
>> accidently I permanently deleted the branch where the patch for issue 5099 
>> was.
>> But I had it as git-formated patch, so I reapplied it to a new branch
>> and pushed from there.
>>
>> I've now retested make successfully, though.
>>
>> So I've no clue what's wrong.
>
> Huh?
>
> diff --git a/scm/lily-library.scm b/scm/lily-library.scm
> index c6f066ca32..1b3926af7f 100644
> --- a/scm/lily-library.scm
> +++ b/scm/lily-library.scm
> @@ -778,9 +778,9 @@ as rectangular coordinates @code{(x-length . y-length)}."
>
>  (define-public (remove-whitespace strg)
>  "Remove characters satisfying @code{char-whitespace?} from string @var{strg}"
> -  (string-delete
> -strg
> -char-whitespace?))
> +  (if (guile-v2)
> +  (string-delete char-whitespace? strg)
> +  (string-delete strg char-whitespace?))
>
>  (define-public (string-encode-integer i)
>(cond
>
> The deleted lines have a net paren change of -1 (one paren more closed
> than opened).  The added lines have a net paren change of 0.
>
> This cannot possibly have worked.

Indeed.
How could it survive automatic tests then?

>
> But that's issue 5101 according to the commit message.

Yep.
5099 looks still ok to me.

> Backing out that patch from staging, and taking a further look.

How to proceed?
Upload a fixed patch for review?

I'm terribly sorry for the additional work for you and James,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread Thomas Morley
2017-03-26 10:17 GMT+02:00 David Kastrup :
> Thomas Morley  writes:
>
>> accidently I permanently deleted the branch where the patch for issue
>> 5099 was.
>
> That would be a rather difficult feat (you don't do things like that
> "accidentally" with Git since you have to work very hard to do them,
> with a set of very explicit commands unless you are going to wait for
> several months as part of the recipe).
>
> Try
>
> git reflog
>
> and write down the hash for the first occurence of your branch in the
> list (which would be last in time).
>
> Then do
>
> git branch hash-that-you-wrote-down
>
> and the branch is back in life.  Of course, since this will be an
> unpublished branch, you have to do this in the repository where you
> "permanently" deleted the branch.
>
> Git tends to keep "permanently deleted" stuff around for 3 months or so
> before it really removes them for good.
>
> --
> David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread David Kastrup
Thomas Morley  writes:

> accidently I permanently deleted the branch where the patch for issue
> 5099 was.

That would be a rather difficult feat (you don't do things like that
"accidentally" with Git since you have to work very hard to do them,
with a set of very explicit commands unless you are going to wait for
several months as part of the recipe).

Try

git reflog

and write down the hash for the first occurence of your branch in the
list (which would be last in time).

Then do

git branch hash-that-you-wrote-down

and the branch is back in life.  Of course, since this will be an
unpublished branch, you have to do this in the repository where you
"permanently" deleted the branch.

Git tends to keep "permanently deleted" stuff around for 3 months or so
before it really removes them for good.

-- 
David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-26 Thread David Kastrup
Thomas Morley  writes:

> 2017-03-26 0:48 GMT+01:00 James :
>> Hello
>>
>> After 5099 was pushed I am now seeing problems with basic 'make'.

>
> accidently I permanently deleted the branch where the patch for issue 5099 
> was.
> But I had it as git-formated patch, so I reapplied it to a new branch
> and pushed from there.
>
> I've now retested make successfully, though.
>
> So I've no clue what's wrong.

Huh?

diff --git a/scm/lily-library.scm b/scm/lily-library.scm
index c6f066ca32..1b3926af7f 100644
--- a/scm/lily-library.scm
+++ b/scm/lily-library.scm
@@ -778,9 +778,9 @@ as rectangular coordinates @code{(x-length . y-length)}."
 
 (define-public (remove-whitespace strg)
 "Remove characters satisfying @code{char-whitespace?} from string @var{strg}"
-  (string-delete
-strg
-char-whitespace?))
+  (if (guile-v2)
+  (string-delete char-whitespace? strg)
+  (string-delete strg char-whitespace?))
 
 (define-public (string-encode-integer i)
   (cond

The deleted lines have a net paren change of -1 (one paren more closed
than opened).  The added lines have a net paren change of 0.

This cannot possibly have worked.

But that's issue 5101 according to the commit message.

Backing out that patch from staging, and taking a further look.

-- 
David Kastrup

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel


Re: Issue 5099 - staging broken?

2017-03-25 Thread Thomas Morley
2017-03-26 0:48 GMT+01:00 James :
> Hello
>
> After 5099 was pushed I am now seeing problems with basic 'make'.
>
> --snip--
>
> og level set to 287
> Relocation: is absolute:
> argv0=/tmp/build-lilypond-autobuild/out/bin/lilypond
> PATH=/tmp/build-lilypond-autobuild/out/bin (prepend) Setting PATH
> to 
> /tmp/build-lilypond-autobuild/out/bin:/tmp/build-lilypond-autobuild/lily/out:/tmp/build-lilypond-autobuild/scripts/build/out:/tmp/build-lilypond-autobuild/scripts/out:/tmp/build-lilypond-autobuild/lily/out:/tmp/build-lilypond-autobuild/scripts/build/out:/tmp/build-lilypond-autobuild/scripts/out:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/home/james/.local/bin:/home/james/bin:/home/james/lilypond-git/build/out/bin::
> Relocation: compile datadir=, new
> datadir=/tmp/build-lilypond-autobuild/out/share/lilypond//current
> Relocation: framework_prefix=/tmp/build-lilypond-autobuild/out/bin/..
> Setting INSTALLER_PREFIX to /tmp/build-lilypond-autobuild/out/bin/..
> PATH=/tmp/build-lilypond-autobuild/out/bin/../bin (prepend) Setting
> PATH
> to 
> /tmp/build-lilypond-autobuild/out/bin/../bin:/tmp/build-lilypond-autobuild/out/bin:/tmp/build-lilypond-autobuild/lily/out:/tmp/build-lilypond-autobuild/scripts/build/out:/tmp/build-lilypond-autobuild/scripts/out:/tmp/build-lilypond-autobuild/lily/out:/tmp/build-lilypond-autobuild/scripts/build/out:/tmp/build-lilypond-autobuild/scripts/out:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/home/james/.local/bin:/home/james/bin:/home/james/lilypond-git/build/out/bin::
> Setting GUILE_MIN_YIELD_1 to 65 Setting GUILE_MIN_YIELD_2 to 65 Setting
> GUILE_MIN_YIELD_MALLOC to 65 Setting GUILE_INIT_SEGMENT_SIZE_1 to
> 10485760 Setting GUILE_MAX_SEGMENT_SIZE to 104857600
>
> LILYPOND_DATADIR="/usr/local/share/lilypond/2.19.58"
> LOCALEDIR="/usr/local/share/locale"
>
> Effective prefix:
> "/tmp/build-lilypond-autobuild/out/share/lilypond/current"
> PATH="/tmp/build-lilypond-autobuild/out/bin/../bin:/tmp/build-lilypond-autobuild/out/bin:/tmp/build-lilypond-autobuild/lily/out:/tmp/build-lilypond-autobuild/scripts/build/out:/tmp/build-lilypond-autobuild/scripts/out:/tmp/build-lilypond-autobuild/lily/out:/tmp/build-lilypond-autobuild/scripts/build/out:/tmp/build-lilypond-autobuild/scripts/out:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/home/james/.local/bin:/home/james/bin:/home/james/lilypond-git/build/out/bin::"
>  []
> Guile 1.8
> [/tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily-library.scm/tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily.scmBacktrace:
> In /tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily.scm:
>  658: 0* [for-each # #]
> In unknown file:
>?: 1* [ly:load "lily-library.scm"]
> In /tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily.scm:
>  507: 2* (let* ((file-name (%search-load-path x))) (ly:debug "[~A"
> file-name) ...) 511: 3* [primitive-load-path
> "/tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily-library.scm"]
>
> /tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily.scm:511:5:
> In procedure scm_i_lreadparen in expression (primitive-load-path
> file-name): 
> /tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily.scm:511:5: 
> /tmp/build-lilypond-autobuild/out/share/lilypond/current/scm/lily-library.scm:1051:1:
> end of file
>
>
>
> --snip--
>
> Obviously this went through a normal countdown and unless I made a
> mistake with my patch testing (remember it is all still manually done)
> then I don't know why it was not picked up at the time.
>
> Could someone else verify this on their own machines?
>
> Else check git master and if staging is still not merged and the time
> difference is more than 2 hrs old (the cycle that my server at work
> merges staging) then it isn't just me after all.
>
>
> James

Hi James,

accidently I permanently deleted the branch where the patch for issue 5099 was.
But I had it as git-formated patch, so I reapplied it to a new branch
and pushed from there.

I've now retested make successfully, though.

So I've no clue what's wrong.

Cheers,
  Harm

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel