[ITP] neomutt

[ITP] neomutt

[Federico Kircheis]

Hello to everyone,

I'm interested in becoming a package maintainer for the program neomutt.
(see https://www.neomutt.org/)

It would be a new package for the cygwin distribution, but it is already
distributed on different systems, like Arch, Debian, Fedora, openSUSE,
Gentoo and many others, including BSD and Mac OS (more information can
be found here: https://www.neomutt.org/distro.html).


Currently there is no Windows port (and it is not planned), therefore I
would like to maintain a cygwin port, since i was able to compile and
use the program without any patch.


neomutt is a fork of the mutt mail client (already provided by cygwin),
and its licensed GPL2 or later.
(see https://www.neomutt.org/guide/intro#copyright).



Best regards

Federico Kircheis

Re: [ITP] neomutt

[Marco Atzeri]

On 20/01/2018 08:43, Federico Kircheis wrote:
> Hello to everyone,
> 
> I'm interested in becoming a package maintainer for the program neomutt.
> (see https://www.neomutt.org/)
> 
> It would be a new package for the cygwin distribution, but it is already
> distributed on different systems, like Arch, Debian, Fedora, openSUSE,
> Gentoo and many others, including BSD and Mac OS (more information can
> be found here: https://www.neomutt.org/distro.html).
> 
> 
> Currently there is no Windows port (and it is not planned), therefore I
> would like to maintain a cygwin port, since i was able to compile and
> use the program without any patch.
> 
> 
> neomutt is a fork of the mutt mail client (already provided by cygwin),
> and its licensed GPL2 or later.
> (see https://www.neomutt.org/guide/intro#copyright).
> 
> 
> 
> Best regards
> 
> Federico Kircheis

Federico,
thanks for you offer.

Please follow the guideline
https://cygwin.com/packaging-contributors-guide.html#submitting
to prepare the package and allow us to assess its consistency
with the expect package structure.

Feel free to ask guidance here if you have any doubt

Regards
Marco

Re: [ITP] neomutt

[Federico Kircheis]

On 01/20/2018 07:22 PM, Marco Atzeri wrote:
> [...]
> 
> Please follow the guideline
> https://cygwin.com/packaging-contributors-guide.html#submitting
> to prepare the package and allow us to assess its consistency
> with the expect package structure.
> 
> Feel free to ask guidance here if you have any doubt
> 
> [...]
> 

Thank you,

Since the package was not part of cygwin I was following the three
points described in the "Submission rules".


I'm not a developer of the neomutt project, and the project is hosted on
GitHub. Therefore I simply copied the latest release on a separate
server, in order to put it together with the .hint and .cygport file:

https://fekir.info/public/neomutt

Would it have been OK to provide the .hint and .cygport file per email
(they are not big) and the link to the original package?

(the official link to the latest release is
https://github.com/neomutt/neomutt/archive/neomutt-20171215.tar.gz)

I did no made the structure with "x86" as in the sample since the
structure for both for x86 and x64 with exact the same source code,
.hint and .cygport file.

Federico

Re: [ITP] neomutt

[Jon Turney]

On 20/01/2018 19:52, Federico Kircheis wrote:
> On 01/20/2018 07:22 PM, Marco Atzeri wrote:
[...]
> I'm not a developer of the neomutt project, and the project is hosted on
> GitHub. Therefore I simply copied the latest release on a separate
> server, in order to put it together with the .hint and .cygport file:
> 
> https://fekir.info/public/neomutt
> 
> Would it have been OK to provide the .hint and .cygport file per email
> (they are not big) and the link to the original package?
> 
> (the official link to the latest release is
> https://github.com/neomutt/neomutt/archive/neomutt-20171215.tar.gz)
> 
> I did no made the structure with "x86" as in the sample since the
> structure for both for x86 and x64 with exact the same source code,
> .hint and .cygport file.

Thanks.

Maybe it's not a good idea to describe it as a fork of Mutt, when 
https://www.neomutt.org/about.html says "It's not a fork of Mutt.
It's a large set of feature patches." :)

You can name your .cygport file just "neomutt.cygport"

You need to provide at least --prefix=/usr to ./configure (currently 
everything is being installed into /usr/local, not /usr), and possibly 
others to satisfy the requirements of [1]

[1] https://cygwin.com/packaging-package-files.html#package_contents

[ITP] neomutt

[Federico Kircheis]

On 01/21/2018 06:51 PM, Jon Turney wrote:
> Maybe it's not a good idea to describe it as a fork of Mutt, when 
> https://www.neomutt.org/about.html says "It's not a fork of Mutt.
> It's a large set of feature patches." :)

Well, I copied the text from the homepage, so it seems to be 
inconsistent. AFAIK the code has been forked and enhanced/patched and 
something merged back, but just to be sure I'll ask the authors what 
they would prefer as description.

> You can name your .cygport file just "neomutt.cygport"

Renamed.

> You need to provide at least --prefix=/usr to ./configure (currently 
> everything is being installed into /usr/local, not /usr), and possibly 
> others to satisfy the requirements of [1]
> 
> [1] https://cygwin.com/packaging-package-files.html#package_contents

I updated the neomutt.cygport file and passed all parameters as 
described in [1], I'm unsure if localstatedir and infodir are really 
supported since they do not appear in the "Makefile.autosetup" file (but 
they do appear in the "system.ctl" file...).

I'll need to investigate it, maybe neomutt defaults are the same with 
those required (for example, the default for mandir seems to be the same 
as the required path), or maybe they are not used at all.

Federico

Re: [ITP] neomutt

[Federico Kircheis]

On 01/21/2018 09:18 PM, Federico Kircheis wrote:
> 
> On 01/21/2018 06:51 PM, Jon Turney wrote:
>> Maybe it's not a good idea to describe it as a fork of Mutt, when 
>> https://www.neomutt.org/about.html says "It's not a fork of Mutt.
>> It's a large set of feature patches." :)
> 
> Well, I copied the text from the homepage, so it seems to be 
> inconsistent. AFAIK the code has been forked and enhanced/patched and 
> something merged back, but just to be sure I'll ask the authors what 
> they would prefer as description.
> 

I've asked the author of the project: "NeoMutt is a fork of Mutt".

Neomutt is not a fork of Mutt in the sense that the two projects are 
diverging.
Fixes done 'upstream' are backported to neomutt, meaning that neomutt is 
an extension of mutt.

>> You need to provide at least --prefix=/usr to ./configure (currently 
>> everything is being installed into /usr/local, not /usr), and possibly 
>> others to satisfy the requirements of [1]
>>
>> [1] https://cygwin.com/packaging-package-files.html#package_contents
> 
> I updated the neomutt.cygport file and passed all parameters as 
> described in [1], I'm unsure if localstatedir and infodir are really 
> supported since they do not appear in the "Makefile.autosetup" file (but 
> they do appear in the "system.ctl" file...).
> 
> I'll need to investigate it, maybe neomutt defaults are the same with 
> those required (for example, the default for mandir seems to be the same 
> as the required path), or maybe they are not used at all.
> 

The parameters seems to be part of autosetup, but those directories are 
not used by neomutt.


Is there anything else I can/should do in order to get the package approved?

Federico

Re: [ITP] neomutt

[Jon Turney]

On 24/01/2018 18:07, Federico Kircheis wrote:
> On 01/21/2018 09:18 PM, Federico Kircheis wrote:
>>
>> On 01/21/2018 06:51 PM, Jon Turney wrote:
>>> Maybe it's not a good idea to describe it as a fork of Mutt, when 
>>> https://www.neomutt.org/about.html says "It's not a fork of Mutt.
>>> It's a large set of feature patches." :)
>>
>> Well, I copied the text from the homepage, so it seems to be 
>> inconsistent. AFAIK the code has been forked and enhanced/patched and 
>> something merged back, but just to be sure I'll ask the authors what 
>> they would prefer as description.
>>
> 
> I've asked the author of the project: "NeoMutt is a fork of Mutt".
> 
> Neomutt is not a fork of Mutt in the sense that the two projects are 
> diverging.
> Fixes done 'upstream' are backported to neomutt, meaning that neomutt is 
> an extension of mutt.
> 
>>> You need to provide at least --prefix=/usr to ./configure (currently 
>>> everything is being installed into /usr/local, not /usr), and 
>>> possibly others to satisfy the requirements of [1]
>>>
>>> [1] https://cygwin.com/packaging-package-files.html#package_contents
>>
>> I updated the neomutt.cygport file and passed all parameters as 
>> described in [1], I'm unsure if localstatedir and infodir are really 
>> supported since they do not appear in the "Makefile.autosetup" file 
>> (but they do appear in the "system.ctl" file...).
>>
>> I'll need to investigate it, maybe neomutt defaults are the same with 
>> those required (for example, the default for mandir seems to be the 
>> same as the required path), or maybe they are not used at all.
>>
> 
> The parameters seems to be part of autosetup, but those directories are 
> not used by neomutt.

Ok, no problem.

> Is there anything else I can/should do in order to get the package 
> approved?

I think you should also be using 'make_etc_defaults /etc/neomuttrc' in 
the install step.

Please provide a ssh key as per https://cygwin.com/package-upload.html

Re: [ITP] neomutt

[Federico Kircheis]

Name: Federico Kircheis
Package: neomutt
---- BEGIN SSH2 PUBLIC KEY ----
Comment: "4096-bit RSA, converted from OpenSSH"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---- END SSH2 PUBLIC KEY ----

Re: [ITP] neomutt

[Jon Turney]

On 28/01/2018 11:38, Federico Kircheis wrote:
> Name: Federico Kircheis
> Package: neomutt

Done

Re: [ITP] neomutt

[Federico Kircheis]

On 01/28/2018 03:43 PM, Jon Turney wrote:
> On 28/01/2018 11:38, Federico Kircheis wrote:
>> Name: Federico Kircheis
>> Package: neomutt
> 
> Done

Uploaded.

Re: [ITP] neomutt

[Jon Turney]

On 30/01/2018 05:56, Federico Kircheis wrote:
> On 01/28/2018 03:43 PM, Jon Turney wrote:
>> On 28/01/2018 11:38, Federico Kircheis wrote:
>>> Name: Federico Kircheis
>>> Package: neomutt
>>
>> Done
> 
> Uploaded.

Please upload a x86 package as well.

Re: [ITP] neomutt

[Federico Kircheis]

On 01/31/2018 06:55 PM, Jon Turney wrote:
> On 30/01/2018 05:56, Federico Kircheis wrote:
>> On 01/28/2018 03:43 PM, Jon Turney wrote:
>>> On 28/01/2018 11:38, Federico Kircheis wrote:
>>>> Name: Federico Kircheis
>>>> Package: neomutt
>>>
>>> Done
>>
>> Uploaded.
> 
> Please upload a x86 package as well.
> 

I'm sorry, I uploaded it, too.

I thought that the previous upload would have been valid both for x64 
and x86.

I was unsure how to proceed, so I've downloaded a x86 version of cygwin, 
copied the cygport file in a separate directory, and so on.

Is there a way to update the package for x64 and x86 at once?
Without using ftp directly of course.

`man cygport` mentions the `--32` and `--64` parameters for building the 
package, but not for uploading it and I did not want to try it out 
before messing something up on the server :-)


Federico

Re: [ITP] neomutt

[Jon Turney]

On 31/01/2018 19:21, Federico Kircheis wrote:
> On 01/31/2018 06:55 PM, Jon Turney wrote:
>> On 30/01/2018 05:56, Federico Kircheis wrote:
>>> On 01/28/2018 03:43 PM, Jon Turney wrote:
>>>> On 28/01/2018 11:38, Federico Kircheis wrote:
>>>>> Name: Federico Kircheis
>>>>> Package: neomutt
>>>>
>>>> Done
>>>
>>> Uploaded.
>>
>> Please upload a x86 package as well.
> 
> I'm sorry, I uploaded it, too.
> 
> I thought that the previous upload would have been valid both for x64 
> and x86.
> 
> I was unsure how to proceed, so I've downloaded a x86 version of cygwin, 
> copied the cygport file in a separate directory, and so on.

This is probably the most straightforward way.

> Is there a way to update the package for x64 and x86 at once?

You don't need to do this, since the process which receives the uploads 
can tolerate x86 and x86 arriving separately.

> Without using ftp directly of course.

If you really want to ensure they are processed at once, I guess you'd 
have to use 'cygport stage' to upload the files for each arch, then 
manually upload !ready files (as described in [1])

But, I can't think of any reasons to bother doing that for a single package.

[1] https://cygwin.com/package-upload.html