"unbundle" mkpasswd from whois.

"unbundle" mkpasswd from whois.

[ng0]

This is one solution to the whois+mkpasswd bundle.
Another solution could be to move this patch content to a package "mkpasswd" which inherits from whois.

What do you think?

[PATCH] gnu: whois: Move mkpasswd to its own output.

[ng0]

* gnu/packages/networking.scm (whois): Move mkpasswd to output "mkpasswd".
[outputs]: Add "mkpasswd".
[arguments]: Add new phase "move-mkpasswd".
Remove TODO comment about unbundling.
---
 gnu/packages/networking.scm | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)

diff --git a/gnu/packages/networking.scm b/gnu/packages/networking.scm
index 3c7c07349..03165e62d 100644
--- a/gnu/packages/networking.scm
+++ b/gnu/packages/networking.scm
@@ -8,7 +8,7 @@
 ;;; Copyright © 2016 John Darrington <jmd@gnu.org>
 ;;; Copyright © 2016 Nicolas Goaziou <mail@nicolasgoaziou.fr>
 ;;; Copyright © 2016 Eric Bavier <bavier@member.fsf.org>
-;;; Copyright © 2016 ng0 <ng0@we.make.ritual.n0.is>
+;;; Copyright © 2016, 2017 ng0 <ng0@libertad.pw>
 ;;; Copyright © 2016 Arun Isaac <arunisaac@systemreboot.net>
 ;;; Copyright © 2016 Benz Schenk <benz.schenk@uzh.ch>
 ;;;
@@ -440,7 +440,6 @@ and up to 1 Mbit/s downstream.")
         (base32
          "0r3bbxpwsxssa99d8dcjnp79mw1cjwqxgmqjm1537q277jwsk0yc"))))
     (build-system gnu-build-system)
-    ;; TODO: unbundle mkpasswd binary + its po files.
     (arguments
      `(#:tests? #f ; Does not exist
        #:make-flags (list "CC=gcc"
@@ -452,7 +451,21 @@ and up to 1 Mbit/s downstream.")
            (lambda _
              (setenv "HAVE_ICONV" "1")
              (setenv "HAVE_LIBIDN" "1")
-             #t)))))
+             #t))
+         (add-after 'install 'move-mkpasswd
+                    (lambda* (#:key outputs #:allow-other-keys)
+                      ;; Move mkpasswd related files to "mkpasswd".
+                      (let ((out (assoc-ref outputs "out"))
+                            (mp (assoc-ref outputs "mkpasswd")))
+                        (mkdir-p (string-append mp "/share/man/man1"))
+                        (mkdir-p (string-append mp "/bin"))
+              (rename-file (string-append out "/share/man/man1/mkpasswd.1")
+                           (string-append mp "/share/man/man1/mkpasswd.1"))
+              (rename-file (string-append out "/bin/mkpasswd")
+                           (string-append mp "/bin/mkpasswd"))
+              #t))))))
+    (outputs '("out"
+               "mkpasswd")) ; 28KiB binary, 16KiB man1
     (inputs
      `(("libidn" ,libidn)))
     (native-inputs
-- 
2.11.0

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[Ludovic Courtès]

Hi!

ng0 <ng0@libertad.pw> skribis:

> * gnu/packages/networking.scm (whois): Move mkpasswd to output "mkpasswd".
> [outputs]: Add "mkpasswd".
> [arguments]: Add new phase "move-mkpasswd".
> Remove TODO comment about unbundling.
> ---
>  gnu/packages/networking.scm | 19 ++++++++++++++++---
>  1 file changed, 16 insertions(+), 3 deletions(-)
>
> diff --git a/gnu/packages/networking.scm b/gnu/packages/networking.scm
> index 3c7c07349..03165e62d 100644
> --- a/gnu/packages/networking.scm
> +++ b/gnu/packages/networking.scm
> @@ -8,7 +8,7 @@
>  ;;; Copyright © 2016 John Darrington <jmd@gnu.org>
>  ;;; Copyright © 2016 Nicolas Goaziou <mail@nicolasgoaziou.fr>
>  ;;; Copyright © 2016 Eric Bavier <bavier@member.fsf.org>
> -;;; Copyright © 2016 ng0 <ng0@we.make.ritual.n0.is>
> +;;; Copyright © 2016, 2017 ng0 <ng0@libertad.pw>
>  ;;; Copyright © 2016 Arun Isaac <arunisaac@systemreboot.net>
>  ;;; Copyright © 2016 Benz Schenk <benz.schenk@uzh.ch>
>  ;;;
> @@ -440,7 +440,6 @@ and up to 1 Mbit/s downstream.")
>          (base32
>           "0r3bbxpwsxssa99d8dcjnp79mw1cjwqxgmqjm1537q277jwsk0yc"))))
>      (build-system gnu-build-system)
> -    ;; TODO: unbundle mkpasswd binary + its po files.
>      (arguments
>       `(#:tests? #f ; Does not exist
>         #:make-flags (list "CC=gcc"
> @@ -452,7 +451,21 @@ and up to 1 Mbit/s downstream.")
>             (lambda _
>               (setenv "HAVE_ICONV" "1")
>               (setenv "HAVE_LIBIDN" "1")
> -             #t)))))
> +             #t))
> +         (add-after 'install 'move-mkpasswd
> +                    (lambda* (#:key outputs #:allow-other-keys)
> +                      ;; Move mkpasswd related files to "mkpasswd".
> +                      (let ((out (assoc-ref outputs "out"))
> +                            (mp (assoc-ref outputs "mkpasswd")))
> +                        (mkdir-p (string-append mp "/share/man/man1"))
> +                        (mkdir-p (string-append mp "/bin"))
> +              (rename-file (string-append out "/share/man/man1/mkpasswd.1")
> +                           (string-append mp "/share/man/man1/mkpasswd.1"))
> +              (rename-file (string-append out "/bin/mkpasswd")
> +                           (string-append mp "/bin/mkpasswd"))
> +              #t))))))
> +    (outputs '("out"
> +               "mkpasswd")) ; 28KiB binary, 16KiB man1

Why to you call that “unbundling”?

Also, I don’t find it useful to have a separate output just for a few
kilobytes.  Could you explain?

Thanks,
Ludo’.

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[ng0]

Ludovic Courtès <ludo@gnu.org> writes:

> Hi!
>
> ng0 <ng0@libertad.pw> skribis:
>
>> * gnu/packages/networking.scm (whois): Move mkpasswd to output "mkpasswd".
>> [outputs]: Add "mkpasswd".
>> [arguments]: Add new phase "move-mkpasswd".
>> Remove TODO comment about unbundling.
>> ---
>>  gnu/packages/networking.scm | 19 ++++++++++++++++---
>>  1 file changed, 16 insertions(+), 3 deletions(-)
>>
>> diff --git a/gnu/packages/networking.scm b/gnu/packages/networking.scm
>> index 3c7c07349..03165e62d 100644
>> --- a/gnu/packages/networking.scm
>> +++ b/gnu/packages/networking.scm
>> @@ -8,7 +8,7 @@
>>  ;;; Copyright © 2016 John Darrington <jmd@gnu.org>
>>  ;;; Copyright © 2016 Nicolas Goaziou <mail@nicolasgoaziou.fr>
>>  ;;; Copyright © 2016 Eric Bavier <bavier@member.fsf.org>
>> -;;; Copyright © 2016 ng0 <ng0@we.make.ritual.n0.is>
>> +;;; Copyright © 2016, 2017 ng0 <ng0@libertad.pw>
>>  ;;; Copyright © 2016 Arun Isaac <arunisaac@systemreboot.net>
>>  ;;; Copyright © 2016 Benz Schenk <benz.schenk@uzh.ch>
>>  ;;;
>> @@ -440,7 +440,6 @@ and up to 1 Mbit/s downstream.")
>>          (base32
>>           "0r3bbxpwsxssa99d8dcjnp79mw1cjwqxgmqjm1537q277jwsk0yc"))))
>>      (build-system gnu-build-system)
>> -    ;; TODO: unbundle mkpasswd binary + its po files.
>>      (arguments
>>       `(#:tests? #f ; Does not exist
>>         #:make-flags (list "CC=gcc"
>> @@ -452,7 +451,21 @@ and up to 1 Mbit/s downstream.")
>>             (lambda _
>>               (setenv "HAVE_ICONV" "1")
>>               (setenv "HAVE_LIBIDN" "1")
>> -             #t)))))
>> +             #t))
>> +         (add-after 'install 'move-mkpasswd
>> +                    (lambda* (#:key outputs #:allow-other-keys)
>> +                      ;; Move mkpasswd related files to "mkpasswd".
>> +                      (let ((out (assoc-ref outputs "out"))
>> +                            (mp (assoc-ref outputs "mkpasswd")))
>> +                        (mkdir-p (string-append mp "/share/man/man1"))
>> +                        (mkdir-p (string-append mp "/bin"))
>> +              (rename-file (string-append out "/share/man/man1/mkpasswd.1")
>> +                           (string-append mp "/share/man/man1/mkpasswd.1"))
>> +              (rename-file (string-append out "/bin/mkpasswd")
>> +                           (string-append mp "/bin/mkpasswd"))
>> +              #t))))))
>> +    (outputs '("out"
>> +               "mkpasswd")) ; 28KiB binary, 16KiB man1
>
> Why to you call that “unbundling”?

Wrong choice of words.

> Also, I don’t find it useful to have a separate output just for a few
> kilobytes.  Could you explain?

Mkpasswd has nothing to do with whois, and upstream says so. I
think we should be able to install whois and just have whois, not
some random "historical bundled" binary in addition to whois.
I'm just stating the sizes as in comparison to the other outputs
they sometimes matter even if its just KiB.

Like I've written in the opening email, another approach could be
to simply delete mkpasswd from the output build and create an
'mkpasswd' package (inheriting from whois) which does the same
for whois.

> Thanks,
> Ludo’.
>

-- 
♥Ⓐ  ng0 -- https://www.inventati.org/patternsinthechaos/

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[Ludovic Courtès]

Hey,

ng0 <ng0@libertad.pw> skribis:

> Mkpasswd has nothing to do with whois, and upstream says so. I
> think we should be able to install whois and just have whois, not
> some random "historical bundled" binary in addition to whois.
> I'm just stating the sizes as in comparison to the other outputs
> they sometimes matter even if its just KiB.
>
> Like I've written in the opening email, another approach could be
> to simply delete mkpasswd from the output build and create an
> 'mkpasswd' package (inheriting from whois) which does the same
> for whois.

I might be overlooking something, but my gut reaction is that (1) it’s
not “bundling” in that we don’t have any other package providing
mkpasswd, and (2) it’s up to upstream to decide whether or not to
distribute mkpasswd.

WDYT?

Ludo’.

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[ng0]

Ludovic Courtès <ludo@gnu.org> writes:

> Hey,
>
> ng0 <ng0@libertad.pw> skribis:
>
>> Mkpasswd has nothing to do with whois, and upstream says so. I
>> think we should be able to install whois and just have whois, not
>> some random "historical bundled" binary in addition to whois.
>> I'm just stating the sizes as in comparison to the other outputs
>> they sometimes matter even if its just KiB.
>>
>> Like I've written in the opening email, another approach could be
>> to simply delete mkpasswd from the output build and create an
>> 'mkpasswd' package (inheriting from whois) which does the same
>> for whois.
>
> I might be overlooking something, but my gut reaction is that (1) it’s
> not “bundling” in that we don’t have any other package providing
> mkpasswd, and (2) it’s up to upstream to decide whether or not to
> distribute mkpasswd.
>
> WDYT?
>
> Ludo’.

It's annoying that two unrelated applications are thrown into one
distribution/package.

I'm contacting upstream developer later this week to see wether I
can get them to move mkpasswd elsewhere or why it's still there.
-- 
♥Ⓐ  ng0 -- https://www.inventati.org/patternsinthechaos/

Whois: Update to 5.2.14

[ng0]

This updates whois to 5.2.14.

[PATCH] gnu: whois: Update to 5.2.14.

[ng0]

* gnu/packages/networking.scm (whois): Update to 5.2.14.
---
 gnu/packages/networking.scm | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/gnu/packages/networking.scm b/gnu/packages/networking.scm
index 3c7c07349..8258b3967 100644
--- a/gnu/packages/networking.scm
+++ b/gnu/packages/networking.scm
@@ -8,7 +8,7 @@
 ;;; Copyright © 2016 John Darrington <jmd@gnu.org>
 ;;; Copyright © 2016 Nicolas Goaziou <mail@nicolasgoaziou.fr>
 ;;; Copyright © 2016 Eric Bavier <bavier@member.fsf.org>
-;;; Copyright © 2016 ng0 <ng0@we.make.ritual.n0.is>
+;;; Copyright © 2016, 2017 ng0 <ng0@libertad.pw>
 ;;; Copyright © 2016 Arun Isaac <arunisaac@systemreboot.net>
 ;;; Copyright © 2016 Benz Schenk <benz.schenk@uzh.ch>
 ;;;
@@ -430,7 +430,7 @@ and up to 1 Mbit/s downstream.")
 (define-public whois
   (package
     (name "whois")
-    (version "5.2.13")
+    (version "5.2.14")
     (source
      (origin
        (method url-fetch)
@@ -438,7 +438,7 @@ and up to 1 Mbit/s downstream.")
                            name "_" version ".tar.xz"))
        (sha256
         (base32
-         "0r3bbxpwsxssa99d8dcjnp79mw1cjwqxgmqjm1537q277jwsk0yc"))))
+         "1lmh7168yby1ma8r1svgvmv9hbgjyniy9c64r6lby3zdmd0sy7d4"))))
     (build-system gnu-build-system)
     ;; TODO: unbundle mkpasswd binary + its po files.
     (arguments
-- 
2.11.0

Re: [PATCH] gnu: whois: Update to 5.2.14.

[Leo Famulari]

On Mon, Jan 09, 2017 at 02:45:25PM +0000, ng0 wrote:
> * gnu/packages/networking.scm (whois): Update to 5.2.14.

Pushed, thanks!

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[ng0]

Ludovic Courtès <ludo@gnu.org> writes:

> Hey,
>
> ng0 <ng0@libertad.pw> skribis:
>
>> Mkpasswd has nothing to do with whois, and upstream says so. I
>> think we should be able to install whois and just have whois, not
>> some random "historical bundled" binary in addition to whois.
>> I'm just stating the sizes as in comparison to the other outputs
>> they sometimes matter even if its just KiB.
>>
>> Like I've written in the opening email, another approach could be
>> to simply delete mkpasswd from the output build and create an
>> 'mkpasswd' package (inheriting from whois) which does the same
>> for whois.
>
> I might be overlooking something, but my gut reaction is that (1) it’s
> not “bundling” in that we don’t have any other package providing
> mkpasswd, and (2) it’s up to upstream to decide whether or not to
> distribute mkpasswd.
>
> WDYT?
>
> Ludo’.
>
And upstreams decision is not to do it (on making it a choice not
to build it or to move it to another tarball):

"No, but thank you for asking."

I think it's informative and neutral enough that I can just share
this.

So we stick with the default case? I'm still in favor for making
it separate guix packages.
-- 
♥Ⓐ  ng0 -- https://www.inventati.org/patternsinthechaos/

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[Hartmut Goebel]

Am 20.01.2017 um 11:41 schrieb ng0:
> I'm still in favor for making it separate guix packages.

+1

-- 
Regards
Hartmut Goebel

| Hartmut Goebel          | h.goebel@crazy-compilers.com               |
| www.crazy-compilers.com | compilers which you thought are impossible |

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[Ludovic Courtès]

ng0 <contact.ng0@cryptolab.net> skribis:

> Ludovic Courtès <ludo@gnu.org> writes:
>
>> Hey,
>>
>> ng0 <ng0@libertad.pw> skribis:
>>
>>> Mkpasswd has nothing to do with whois, and upstream says so. I
>>> think we should be able to install whois and just have whois, not
>>> some random "historical bundled" binary in addition to whois.
>>> I'm just stating the sizes as in comparison to the other outputs
>>> they sometimes matter even if its just KiB.
>>>
>>> Like I've written in the opening email, another approach could be
>>> to simply delete mkpasswd from the output build and create an
>>> 'mkpasswd' package (inheriting from whois) which does the same
>>> for whois.
>>
>> I might be overlooking something, but my gut reaction is that (1) it’s
>> not “bundling” in that we don’t have any other package providing
>> mkpasswd, and (2) it’s up to upstream to decide whether or not to
>> distribute mkpasswd.
>>
>> WDYT?
>>
>> Ludo’.
>>
> And upstreams decision is not to do it (on making it a choice not
> to build it or to move it to another tarball):
>
> "No, but thank you for asking."

So upstream’s decision is to keep mkpasswd in whois?

> So we stick with the default case? I'm still in favor for making
> it separate guix packages.

I’m still in favor of sticking to upstream’s decision, whatever that
is.  But honestly, I fail to see why this is a big deal.  :-)

Thanks,
Ludo’.

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[Ricardo Wurmus]

ng0 <contact.ng0@cryptolab.net> writes:

>> I might be overlooking something, but my gut reaction is that (1) it’s
>> not “bundling” in that we don’t have any other package providing
>> mkpasswd, and (2) it’s up to upstream to decide whether or not to
>> distribute mkpasswd.
>>
>> WDYT?
>>
>> Ludo’.
>>
> And upstreams decision is not to do it (on making it a choice not
> to build it or to move it to another tarball):
>
> "No, but thank you for asking."
>
> I think it's informative and neutral enough that I can just share
> this.
>
> So we stick with the default case? I'm still in favor for making
> it separate guix packages.

I’d rather stick with what upstream provides.  I don’t see an advantage
of separating these tools.

--
Ricardo

GPG: BCA6 89B6 3655 3801 C3C6  2150 197A 5888 235F ACAC
https://elephly.net

Re: [PATCH] gnu: whois: Move mkpasswd to its own output.

[ng0]

Ludovic Courtès <ludo@gnu.org> writes:

> ng0 <contact.ng0@cryptolab.net> skribis:
>
>> Ludovic Courtès <ludo@gnu.org> writes:
>>
>>> Hey,
>>>
>>> ng0 <ng0@libertad.pw> skribis:
>>>
>>>> Mkpasswd has nothing to do with whois, and upstream says so. I
>>>> think we should be able to install whois and just have whois, not
>>>> some random "historical bundled" binary in addition to whois.
>>>> I'm just stating the sizes as in comparison to the other outputs
>>>> they sometimes matter even if its just KiB.
>>>>
>>>> Like I've written in the opening email, another approach could be
>>>> to simply delete mkpasswd from the output build and create an
>>>> 'mkpasswd' package (inheriting from whois) which does the same
>>>> for whois.
>>>
>>> I might be overlooking something, but my gut reaction is that (1) it’s
>>> not “bundling” in that we don’t have any other package providing
>>> mkpasswd, and (2) it’s up to upstream to decide whether or not to
>>> distribute mkpasswd.
>>>
>>> WDYT?
>>>
>>> Ludo’.
>>>
>> And upstreams decision is not to do it (on making it a choice not
>> to build it or to move it to another tarball):
>>
>> "No, but thank you for asking."
>
> So upstream’s decision is to keep mkpasswd in whois?
>
>> So we stick with the default case? I'm still in favor for making
>> it separate guix packages.
>
> I’m still in favor of sticking to upstream’s decision, whatever that
> is.  But honestly, I fail to see why this is a big deal.  :-)
>
> Thanks,
> Ludo’.

Okay. If someone comes up with it again, we can discuss it. For
me it is done.
-- 
♥Ⓐ  ng0 -- https://www.inventati.org/patternsinthechaos/