diff mbox

[1/2] package/monolite: bump to version 1050001000

Message ID 1502439245-18039-1-git-send-email-angelo.compagnucci@gmail.com
State Accepted
Headers show

Commit Message

Angelo Compagnucci Aug. 11, 2017, 8:14 a.m. UTC
The latest version of mono carries a bit of changes in the monolite
package: the version string changes and from now on, monolite should
be installed in a subdirectory with the exact version string as a name.

Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
---
 package/monolite/monolite.hash | 2 +-
 package/monolite/monolite.mk   | 6 +++---
 2 files changed, 4 insertions(+), 4 deletions(-)

Comments

Arnout Vandecappelle Aug. 11, 2017, 9:56 a.m. UTC | #1
On 11-08-17 10:14, Angelo Compagnucci wrote:
> The latest version of mono carries a bit of changes in the monolite
> package: the version string changes and from now on, monolite should
> be installed in a subdirectory with the exact version string as a name.

 And does this work independently of the version bump of mono itself? I.e., if
we apply just this patch, will mono still build?

 If not, then the two patches should be squashed. If you don't have time we can
do that while applying.

 Regards,
 Arnout

> 
> Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
> ---
>  package/monolite/monolite.hash | 2 +-
>  package/monolite/monolite.mk   | 6 +++---
>  2 files changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/package/monolite/monolite.hash b/package/monolite/monolite.hash
> index 09f9554..7638ce2 100644
> --- a/package/monolite/monolite.hash
> +++ b/package/monolite/monolite.hash
> @@ -1,2 +1,2 @@
>  # sha256 locally computed
> -sha256 2cdf6cff1d82d76412461a4c8a3616bb2aa1e835fb55479941662dec3799c924  monolite-156-latest.tar.gz
> +sha256 365dc589e6d336530ef8efaa491e932c15163f449632daef6c41bed770d9fe53  monolite-1050001000-latest.tar.gz
> diff --git a/package/monolite/monolite.mk b/package/monolite/monolite.mk
> index 73f2352..2cc08df 100644
> --- a/package/monolite/monolite.mk
> +++ b/package/monolite/monolite.mk
> @@ -4,14 +4,14 @@
>  #
>  ################################################################################
>  
> -MONOLITE_VERSION = 156
> +MONOLITE_VERSION = 1050001000
>  MONOLITE_SITE = http://download.mono-project.com/monolite/
>  MONOLITE_SOURCE = monolite-$(MONOLITE_VERSION)-latest.tar.gz
>  MONOLITE_LICENSE = LGPL-2.0 or commercial
>  
>  define HOST_MONOLITE_INSTALL_CMDS
> -	mkdir -p $(HOST_DIR)/usr/lib/monolite
> -	cp $(@D)/* $(HOST_DIR)/usr/lib/monolite
> +	mkdir -p $(HOST_DIR)/usr/lib/monolite/$(MONOLITE_VERSION)
> +	cp -r $(@D)/* $(HOST_DIR)/usr/lib/monolite/$(MONOLITE_VERSION)
>  endef
>  
>  $(eval $(host-generic-package))
>
Angelo Compagnucci Aug. 11, 2017, 10:40 a.m. UTC | #2
Dear Arnout Vandecappelle,

2017-08-11 11:56 GMT+02:00 Arnout Vandecappelle <arnout@mind.be>:
>
>
> On 11-08-17 10:14, Angelo Compagnucci wrote:
>> The latest version of mono carries a bit of changes in the monolite
>> package: the version string changes and from now on, monolite should
>> be installed in a subdirectory with the exact version string as a name.
>
>  And does this work independently of the version bump of mono itself? I.e., if
> we apply just this patch, will mono still build?

Nope. The patches are sent in a series for the purpose to be both applied.

>  If not, then the two patches should be squashed. If you don't have time we can
> do that while applying.

Sorry for the noise, but this is not how the things worked in the
past. I was suggested several times in the past to send a patch for
each modification/package.

Sincerely, Angelo.

>  Regards,
>  Arnout
>
>>
>> Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>
>> ---
>>  package/monolite/monolite.hash | 2 +-
>>  package/monolite/monolite.mk   | 6 +++---
>>  2 files changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/package/monolite/monolite.hash b/package/monolite/monolite.hash
>> index 09f9554..7638ce2 100644
>> --- a/package/monolite/monolite.hash
>> +++ b/package/monolite/monolite.hash
>> @@ -1,2 +1,2 @@
>>  # sha256 locally computed
>> -sha256 2cdf6cff1d82d76412461a4c8a3616bb2aa1e835fb55479941662dec3799c924  monolite-156-latest.tar.gz
>> +sha256 365dc589e6d336530ef8efaa491e932c15163f449632daef6c41bed770d9fe53  monolite-1050001000-latest.tar.gz
>> diff --git a/package/monolite/monolite.mk b/package/monolite/monolite.mk
>> index 73f2352..2cc08df 100644
>> --- a/package/monolite/monolite.mk
>> +++ b/package/monolite/monolite.mk
>> @@ -4,14 +4,14 @@
>>  #
>>  ################################################################################
>>
>> -MONOLITE_VERSION = 156
>> +MONOLITE_VERSION = 1050001000
>>  MONOLITE_SITE = http://download.mono-project.com/monolite/
>>  MONOLITE_SOURCE = monolite-$(MONOLITE_VERSION)-latest.tar.gz
>>  MONOLITE_LICENSE = LGPL-2.0 or commercial
>>
>>  define HOST_MONOLITE_INSTALL_CMDS
>> -     mkdir -p $(HOST_DIR)/usr/lib/monolite
>> -     cp $(@D)/* $(HOST_DIR)/usr/lib/monolite
>> +     mkdir -p $(HOST_DIR)/usr/lib/monolite/$(MONOLITE_VERSION)
>> +     cp -r $(@D)/* $(HOST_DIR)/usr/lib/monolite/$(MONOLITE_VERSION)
>>  endef
>>
>>  $(eval $(host-generic-package))
>>
>
> --
> Arnout Vandecappelle                          arnout at mind be
> Senior Embedded Software Architect            +32-16-286500
> Essensium/Mind                                http://www.mind.be
> G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
> LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
> GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF
Arnout Vandecappelle Aug. 11, 2017, 1:38 p.m. UTC | #3
On 11-08-17 12:40, Angelo Compagnucci wrote:
> Dear Arnout Vandecappelle,
> 
> 2017-08-11 11:56 GMT+02:00 Arnout Vandecappelle <arnout@mind.be>:
>>
>>
>> On 11-08-17 10:14, Angelo Compagnucci wrote:
>>> The latest version of mono carries a bit of changes in the monolite
>>> package: the version string changes and from now on, monolite should
>>> be installed in a subdirectory with the exact version string as a name.
>>
>>  And does this work independently of the version bump of mono itself? I.e., if
>> we apply just this patch, will mono still build?
> 
> Nope. The patches are sent in a series for the purpose to be both applied.
> 
>>  If not, then the two patches should be squashed. If you don't have time we can
>> do that while applying.
> 
> Sorry for the noise, but this is not how the things worked in the
> past. I was suggested several times in the past to send a patch for
> each modification/package.

 Yes, but only if they indeed work separately. Don't worry, I'll squash while
applying.

 Regards,
 Arnout

[snip]
Arnout Vandecappelle Aug. 11, 2017, 2:07 p.m. UTC | #4
On 11-08-17 10:14, Angelo Compagnucci wrote:
> The latest version of mono carries a bit of changes in the monolite
> package: the version string changes and from now on, monolite should
> be installed in a subdirectory with the exact version string as a name.
> 
> Signed-off-by: Angelo Compagnucci <angelo.compagnucci@gmail.com>

 I've squashed both, modified the commit message a little, and applied to next,
thanks!

 Regards,
 Arnout
Angelo Compagnucci Aug. 11, 2017, 3:07 p.m. UTC | #5
Dear Arnout Vandecappelle ,

2017-08-11 15:38 GMT+02:00 Arnout Vandecappelle <arnout@mind.be>:
>
>
> On 11-08-17 12:40, Angelo Compagnucci wrote:
>> Dear Arnout Vandecappelle,
>>
>> 2017-08-11 11:56 GMT+02:00 Arnout Vandecappelle <arnout@mind.be>:
>>>
>>>
>>> On 11-08-17 10:14, Angelo Compagnucci wrote:
>>>> The latest version of mono carries a bit of changes in the monolite
>>>> package: the version string changes and from now on, monolite should
>>>> be installed in a subdirectory with the exact version string as a name.
>>>
>>>  And does this work independently of the version bump of mono itself? I.e., if
>>> we apply just this patch, will mono still build?
>>
>> Nope. The patches are sent in a series for the purpose to be both applied.
>>
>>>  If not, then the two patches should be squashed. If you don't have time we can
>>> do that while applying.
>>
>> Sorry for the noise, but this is not how the things worked in the
>> past. I was suggested several times in the past to send a patch for
>> each modification/package.
>
>  Yes, but only if they indeed work separately. Don't worry, I'll squash while
> applying.

I searched the history for the mono package just to be sure and such a
way of squashing commits for separate packages was never been done.
I'm a bit puzzled ...

Right now I'm working updating sysdig and it requires som bumps to
other packages, should I send a squashed patch?!

Thanks!

>
>  Regards,
>  Arnout
>
> [snip]
>
> --
> Arnout Vandecappelle                          arnout at mind be
> Senior Embedded Software Architect            +32-16-286500
> Essensium/Mind                                http://www.mind.be
> G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
> LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
> GPG fingerprint:  7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF
Thomas Petazzoni Aug. 11, 2017, 6:31 p.m. UTC | #6
Hello,

On Fri, 11 Aug 2017 17:07:18 +0200, Angelo Compagnucci wrote:

> >> Sorry for the noise, but this is not how the things worked in the
> >> past. I was suggested several times in the past to send a patch for
> >> each modification/package.  
> >
> >  Yes, but only if they indeed work separately. Don't worry, I'll squash while
> > applying.  
> 
> I searched the history for the mono package just to be sure and such a
> way of squashing commits for separate packages was never been done.
> I'm a bit puzzled ...

Yes, I used to apply them separately. Theoretically, if the bumps are
really both needed for the whole thing to work, Arnout is right that
both bumps should be done in the same patch.

I simply don't apply an absolutely strict rule of "everything should be
bisectable" in Buildroot, so I'm a bit more relaxed than Arnout on
this. I'm not saying Arnout isn't right, just that it wasn't something
that I thought was important enough.

> Right now I'm working updating sysdig and it requires som bumps to
> other packages, should I send a squashed patch?!

No, because the patch series is bisectable: if we apply only PATCH 1/3,
it works. If we apply patches 1/3 and 2/3, it works. And if we apply
1/3, 2/3 and 3/3, it works.

The question is not whether patches are independent, but whether the
patch series is bisectable. I.e what happens if you apply only M
patches on the total of N patches in the series.

Best regards,

Thomas
Arnout Vandecappelle Aug. 11, 2017, 8:35 p.m. UTC | #7
On 11-08-17 17:07, Angelo Compagnucci wrote:
> Dear Arnout Vandecappelle ,
> 
> 2017-08-11 15:38 GMT+02:00 Arnout Vandecappelle <arnout@mind.be>:
>>
>>
>> On 11-08-17 12:40, Angelo Compagnucci wrote:
>>> Dear Arnout Vandecappelle,
>>>
>>> 2017-08-11 11:56 GMT+02:00 Arnout Vandecappelle <arnout@mind.be>:
>>>>
>>>>
>>>> On 11-08-17 10:14, Angelo Compagnucci wrote:
>>>>> The latest version of mono carries a bit of changes in the monolite
>>>>> package: the version string changes and from now on, monolite should
>>>>> be installed in a subdirectory with the exact version string as a name.
>>>>
>>>>  And does this work independently of the version bump of mono itself? I.e., if
>>>> we apply just this patch, will mono still build?
>>>
>>> Nope. The patches are sent in a series for the purpose to be both applied.
>>>
>>>>  If not, then the two patches should be squashed. If you don't have time we can
>>>> do that while applying.
>>>
>>> Sorry for the noise, but this is not how the things worked in the
>>> past. I was suggested several times in the past to send a patch for
>>> each modification/package.
>>
>>  Yes, but only if they indeed work separately. Don't worry, I'll squash while
>> applying.
> 
> I searched the history for the mono package just to be sure and such a
> way of squashing commits for separate packages was never been done.
> I'm a bit puzzled ...
> 
> Right now I'm working updating sysdig and it requires som bumps to
> other packages, should I send a squashed patch?!

 Usually it should be one patch per package. But in this case, the bump of
monolite as such would break mono. I assume that normally that's not the case,
i.e. monolite 156 can still be used for building mono 4.6.2.16. So normally they
don't have to be updated together. It's only now that they have to be squashed.

 Now, it's not really *that* important. It's only important for being able to
trace back when something broke, and that's not needed very often. But since I
saw this potential issue and could easily "fix" it, I did.

 If in doubt, make separate patches. Squashing after the fact is easy, splitting
is not.

 Regards,
 Arnout
diff mbox

Patch

diff --git a/package/monolite/monolite.hash b/package/monolite/monolite.hash
index 09f9554..7638ce2 100644
--- a/package/monolite/monolite.hash
+++ b/package/monolite/monolite.hash
@@ -1,2 +1,2 @@ 
 # sha256 locally computed
-sha256 2cdf6cff1d82d76412461a4c8a3616bb2aa1e835fb55479941662dec3799c924  monolite-156-latest.tar.gz
+sha256 365dc589e6d336530ef8efaa491e932c15163f449632daef6c41bed770d9fe53  monolite-1050001000-latest.tar.gz
diff --git a/package/monolite/monolite.mk b/package/monolite/monolite.mk
index 73f2352..2cc08df 100644
--- a/package/monolite/monolite.mk
+++ b/package/monolite/monolite.mk
@@ -4,14 +4,14 @@ 
 #
 ################################################################################
 
-MONOLITE_VERSION = 156
+MONOLITE_VERSION = 1050001000
 MONOLITE_SITE = http://download.mono-project.com/monolite/
 MONOLITE_SOURCE = monolite-$(MONOLITE_VERSION)-latest.tar.gz
 MONOLITE_LICENSE = LGPL-2.0 or commercial
 
 define HOST_MONOLITE_INSTALL_CMDS
-	mkdir -p $(HOST_DIR)/usr/lib/monolite
-	cp $(@D)/* $(HOST_DIR)/usr/lib/monolite
+	mkdir -p $(HOST_DIR)/usr/lib/monolite/$(MONOLITE_VERSION)
+	cp -r $(@D)/* $(HOST_DIR)/usr/lib/monolite/$(MONOLITE_VERSION)
 endef
 
 $(eval $(host-generic-package))