Request new Ruby release

Request new Ruby release

[Steven Penny]

Please release new Cygwin Ruby. Current version is 2.3.6 (Dec 2017), and since
then 2 versions have dropped [1]:

- 2.4.3 (Dec 2017)
- 2.5.0 (Dec 2017)

2.4.0 introduced Enumerable#sum [2], would be nice to have.

[1] http://github.com/ruby/ruby/tags
[2] http://stackoverflow.com/a/41235616


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Steven Penny]

On Sat, 24 Feb 2018 12:51:36, Steven Penny wrote:
> Please release new Cygwin Ruby. Current version is 2.3.6 (Dec 2017), and since
> then 2 versions have dropped [1]:
>
> - 2.4.3 (Dec 2017)
> - 2.5.0 (Dec 2017)
>
> 2.4.0 introduced Enumerable#sum [2], would be nice to have.
>
> [1] http://github.com/ruby/ruby/tags
> [2] http://stackoverflow.com/a/41235616

http://cygwin.com/ml/cygwin/2018-02/msg00262.html

bump. yaakov said he would look into it after 2.5.1 came out, and its out:

- http://github.com/ruby/ruby/releases/tag/v2_5_1
- http://github.com/cygwinports/ruby/issues/1


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Yaakov Selkowitz]

On 2018-05-03 18:16, Steven Penny wrote:
> On Sat, 24 Feb 2018 12:51:36, Steven Penny wrote:
>> Please release new Cygwin Ruby. Current version is 2.3.6 (Dec 2017),
>> and since
>> then 2 versions have dropped [1]:
>>
>> - 2.4.3 (Dec 2017)
>> - 2.5.0 (Dec 2017)
>>
>> 2.4.0 introduced Enumerable#sum [2], would be nice to have.
>>
>> [1] http://github.com/ruby/ruby/tags
>> [2] http://stackoverflow.com/a/41235616
> 
> http://cygwin.com/ml/cygwin/2018-02/msg00262.html
> 
> bump. yaakov said he would look into it after 2.5.1 came out, and its out:
> 
> - http://github.com/ruby/ruby/releases/tag/v2_5_1
> - http://github.com/cygwinports/ruby/issues/1

I'm well aware, and telling me repeatedly won't make it happen any faster.

-- 
Yaakov

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Brian Inglis]

On 2018-05-03 17:16, Steven Penny wrote:
> On Sat, 24 Feb 2018 12:51:36, Steven Penny wrote:
>> Please release new Cygwin Ruby. Current version is 2.3.6 (Dec 2017), and since
>> then 2 versions have dropped [1]:
>> - 2.4.3 (Dec 2017)
>> - 2.5.0 (Dec 2017)
>> 2.4.0 introduced Enumerable#sum [2], would be nice to have.
>> [1] http://github.com/ruby/ruby/tags
>> [2] http://stackoverflow.com/a/41235616
> http://cygwin.com/ml/cygwin/2018-02/msg00262.html
> bump. yaakov said he would look into it after 2.5.1 came out, and its out:
> - http://github.com/ruby/ruby/releases/tag/v2_5_1
> - http://github.com/cygwinports/ruby/issues/1

So something would be nice for you to have, but you have no real issue.

Ruby is not just a language package, it comes with its own ecosystem, like gcc,
perl, python, or TeX: there's 100+ directly dependent packages that have to be
rebuilt, retested, possibly diagnosed then fixed; then the rest of the packages
that have some ruby dependency; and there's also the other 2500+ packages Yaakov
also maintains and updates, often a bunch every week; about 25% of Cygwin
packages in total; all in whatever he can spare of his time to volunteer to Cygwin.
Some related packages have other maintainers, who have to coordinate and
schedule their upgrading, building, testing, and release efforts; also in
whatever they can spare of their time.

For an understaffed, all-volunteer effort, Cygwin does a tremendous job
providing us with a compatible, reliable, stable subsystem working environment,
running in a less stable environment.
Cygwin seems to keep up to date with stable releases of important packages
better than distros who have full time staff, many more volunteers, and don't
also have to deal with an alien environment.
It can continue that way if we can all be patient, supportive, and appreciative
of the volunteer maintainers, *especially* the main contributors, on whom much
depends.

-- 
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Steven Penny]

On Fri, 4 May 2018 00:05:23, Brian Inglis wrote:
> For an understaffed, all-volunteer effort, Cygwin does a tremendous job
> providing us with a compatible, reliable, stable subsystem working environment,
> running in a less stable environment.
> Cygwin seems to keep up to date with stable releases of important packages
> better than distros who have full time staff, many more volunteers, and don't
> also have to deal with an alien environment.

Cygwin is a fantastic project. and yaakov handles a massive amount of packages,
and i commend him for that. but lets not kid ourselves, ok? cygwin *does* keep
up to date with *some* important packages as you say, but not other important
packages. for example, GCC, is over a year old:

http://gnu.mirrors.hoobly.com/gcc

Ruby is over 2 years old:

http://wikipedia.org/wiki/Ruby_%28programming_language%29#Ruby_2.3

Cmake is over a year old:

http://cygwin.com/ml/cygwin/2018-01/msg00190.html

and several packages dont have a static library:

- http://github.com/cygwinports/mingw64-x86_64-expat/issues/1
- http://github.com/cygwinports/mingw64-x86_64-fdk-aac/issues/1
- http://github.com/cygwinports/mingw64-x86_64-gnutls/issues/1
- http://github.com/cygwinports/mingw64-x86_64-lua/issues/1
- http://github.com/cygwinports/mingw64-x86_64-nghttp2/issues/1
- http://github.com/cygwinports/mingw64-x86_64-pcre/issues/1

compare this to debian which provide both shared and static libraries:

http://packages.debian.org/buster/amd64/libnghttp2-dev/filelist

and MSYS2, which provide both up to date versions of everything ive mentioned,
as well as shared and static libraries:

http://repo.msys2.org/mingw/x86_64


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Marco Atzeri]

On 5/4/2018 2:09 PM, Steven Penny wrote:
> On Fri, 4 May 2018 00:05:23, Brian Inglis wrote:
>> For an understaffed, all-volunteer effort, Cygwin does a tremendous job
>> providing us with a compatible, reliable, stable subsystem working 
>> environment,
>> running in a less stable environment.
>> Cygwin seems to keep up to date with stable releases of important 
>> packages
>> better than distros who have full time staff, many more volunteers, 
>> and don't
>> also have to deal with an alien environment.
> 
> Cygwin is a fantastic project. and yaakov handles a massive amount of 
> packages,
> and i commend him for that. but lets not kid ourselves, ok? cygwin 
> *does* keep
> up to date with *some* important packages as you say, but not other 
> important
> packages. for example, GCC, is over a year old:

Steven,
I do not see you doing any package activity release,
so you should refrain to comment on how we (package maintainers)
use our own spare time for this project.

Please note that sometime, things are broken upstream
and it takes time and it could require not available capability to
solve the issues.
https://gcc.gnu.org/ml/fortran/2018-03/msg00014.html

If you want to help for a new package release that you care,
try to build it, check that every thing is in place,
provide if needed patches and guidance and I am almost sure
the relative maintainer will appreciate your effort.
I surely did
https://sourceware.org/ml/cygwin/2018-02/msg00142.html

Regards
Marco





--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Steven Penny]

On Fri, 4 May 2018 15:16:49, Marco Atzeri wrote:
> I do not see you doing any package activity release, so you should refrain to
> comment on how we (package maintainers) use our own spare time for this
> project.

such comments are perfectly acceptable if maintainers are acting in bad faith
with regard to packages. no one is perfect, certainly not myself. and if the
person in question is a maintainer of an important package, for example GCC or
Ruby, it is not acceptable for them to let a year or 2 go by without an update.

if they are worried about bugs, cygwin has a system in place for many years now
to deal with that. you make a package "[test]", so that people can test it. if
a maintainer cannot faithfully keep up with a package at least once a year,
that person should step aside as im sure other would be willing (myself
included) to take up maintainership of said packages.

> Please note that sometime, things are broken upstream
> and it takes time and it could require not available capability to
> solve the issues.
> https://gcc.gnu.org/ml/fortran/2018-03/msg00014.html

sorry, did you really just invoke fortran as a serious argument? fortran is
arguably the oldest programming language still in use, if you can even call it
that. you cant even do HTTP with it:

http://rosettacode.org/wiki/HTTPS#Fortran

youll notice that ADA already got the axe:

http://github.com/cygwinports/cygwin64-gcc/blob/master/cygwin64-gcc.cygport#L52

this could happen to fortran as well.

> If you want to help for a new package release that you care,
> try to build it, check that every thing is in place,
> provide if needed patches and guidance and I am almost sure
> the relative maintainer will appreciate your effort.
> I surely did
> https://sourceware.org/ml/cygwin/2018-02/msg00142.html

so did i:

http://github.com/svnpenn/glade/tree/master/mingw-w64-x86-64/gcc

my GCC build is done, as soon as I finish my G++ build, I had planned to publish
my recipe, but i guess now is as good time as any. i dont have the patch files
as ive seen with other build, so im sure i am missing something.


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Brian Inglis]

On 2018-05-04 18:56, Steven Penny wrote:
> sorry, did you really just invoke fortran as a serious argument? fortran is
> arguably the oldest programming language still in use, if you can even call it
> that. you can't even do HTTP with it:
> http://rosettacode.org/wiki/HTTPS#Fortran

That's what web servers are for, written in more suitable languages, but search
for fortran+fastcgi+nginx, or have a look at https://github.com/rlcarino/heeds
for a production GPL3 web based Fortran app.

> youll notice that ADA already got the axe:
> http://github.com/cygwinports/cygwin64-gcc/blob/master/cygwin64-gcc.cygport#L52

On Cygwin, but GNAT is still part of GCC 9.
ADA is heavily used in safety critical, RT, embedded avionics and ATC where the
focus is on decades long, exhaustive documentation archival, maintenance,
support, and testing of almost static libraries and tools after release,
requirements for which GNU projects are ill-suited, but in which aerospace and
gov/mil project companies and their suppliers have long experience.

Cobol has similar requirements for bread and butter enterprise and government
financial systems, with high requirements for maintainability to handle
legislative and business changes, and long term platform compatibility and support.

> this could happen to fortran as well.

I thought I'd have seen that by now, but there's too much carefully validated
(parallel) HPC (especially modelling) and scientific code, libraries, apps, and
systems used by various government agencies, institutions, and departments for
heavy or important computation in many fields, that it doesn't look like it'll
ever be allowed to die ;^>

-- 
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Marco Atzeri]

On 5/5/2018 2:56 AM, Steven Penny wrote:
> On Fri, 4 May 2018 15:16:49, Marco Atzeri wrote:
>> I do not see you doing any package activity release, so you should 
>> refrain to
>> comment on how we (package maintainers) use our own spare time for this
>> project.
> 
> such comments are perfectly acceptable if maintainers are acting in bad 
> faith
> with regard to packages. no one is perfect, certainly not myself. and if 
> the
> person in question is a maintainer of an important package, for example 
> GCC or
> Ruby, it is not acceptable for them to let a year or 2 go by without an 
> update.

I am still waiting that you show your code.
Jon Yong is doing an hell of job taming a monster program,
and your comments are NOT appropriated.

> if they are worried about bugs, cygwin has a system in place for many 
> years now
> to deal with that. you make a package "[test]", so that people can test 
> it. if
> a maintainer cannot faithfully keep up with a package at least once a year,
> that person should step aside as im sure other would be willing (myself
> included) to take up maintainership of said packages.

I do not see a huge queue of volunteers, and I do not see your
maintaining efforts.

>> Please note that sometime, things are broken upstream
>> and it takes time and it could require not available capability to
>> solve the issues.
>> https://gcc.gnu.org/ml/fortran/2018-03/msg00014.html
> 
> sorry, did you really just invoke fortran as a serious argument? fortran is
> arguably the oldest programming language still in use, if you can even 
> call it

This only show how ignorant you are about programming language and
their usage. Clearly you are not an engineer.

Marco


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Steven Penny]

On Sat, 5 May 2018 07:27:22, Marco Atzeri wrote:
> I am still waiting that you show your code.

i did, here, i can do it again:

http://github.com/svnpenn

> Jon Yong is doing an hell of job taming a monster program,

no one is arguing that, i agree its a tough package to maintain

> and your comments are NOT appropriated.

certainly not by you - and probably others on this list. but my arguments do
have merit. GCC as an example is a fast updating package. and as "politically
incorrect" as it might be, jon could be doing a better job in regards to
velocity of release - at a minimum we should have test versions for GCC 7 and
or 8 already for ALL arches and ALL targets.

> I do not see a huge queue of volunteers,

this is a false assumption - no one volunteers because a maintainer is already
in place - thats like applying for a job that is filled - but if he vacated and
corinna or whatever posted an "opening" - i am confident it would be filled
quickly

> and I do not see your maintaining efforts.

not sure what you mean by this - no i havent maintained an official cygwin
package - but i maintain several public facing cygwin packages:

http://github.com/svnpenn/glade

along with a Cygwin package manager:

http://github.com/svnpenn/sage

> This only show how ignorant you are about programming language and
> their usage. Clearly you are not an engineer.

id say the same to you - as you seen keen on throwing insults rather than links
and examples - cheers.


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Stephen John Smoogen]

On 5 May 2018 at 02:56, Steven Penny <svnpenn@gmail.com> wrote:
> On Sat, 5 May 2018 07:27:22, Marco Atzeri wrote:
>>
>> I am still waiting that you show your code.
>
>
> i did, here, i can do it again:
>
> http://github.com/svnpenn
>
>> Jon Yong is doing an hell of job taming a monster program,
>
>
> no one is arguing that, i agree its a tough package to maintain
>
>> and your comments are NOT appropriated.
>
>
> certainly not by you - and probably others on this list. but my arguments do
> have merit. GCC as an example is a fast updating package. and as
> "politically
> incorrect" as it might be, jon could be doing a better job in regards to
> velocity of release - at a minimum we should have test versions for GCC 7
> and
> or 8 already for ALL arches and ALL targets.
>
>> I do not see a huge queue of volunteers,
>
>
> this is a false assumption - no one volunteers because a maintainer is
> already
> in place - thats like applying for a job that is filled - but if he vacated
> and
> corinna or whatever posted an "opening" - i am confident it would be filled
> quickly
>

That is where this is falling down. A lot of maintainers are not going
to step aside because they have had too many experiences of where
someone has said they will help out/take over etc and then disappear
for a million legitimate reasons. Most maintainers in open source
software will step aside when someone has shown they are going to not
just poke and prod but actually are doing things like

'Hey I made a set of builds with gcc8 and rebuilt all this to see what
broke.. can someone else look at things to see how this is going?'

and then after that person has shown that they have kept up with
things for multiple releases, helped out and been helping them and
others, the responsibilities start getting moved over. This isn't
always the case, if there are various leaf packages but if there are
ones with hundreds of dependencies like gcc or ruby.. it isn't
advertised of 'oh I am looking for a maintainer..' because you get
burnt too many times with the people who want to make a name for
themselves but don't have the skill, the people who are looking to
push an agenda (everything from 'I put in this patch which rootkits
everyone' to 'I think all developers should default to -Werror only
ever'), people who would be good but have no clue who to contact in
other groups when problems occur, and maybe someone who does fit into
this.

Which is why most developers end up looking for the people who show
they want to do something by doing it first, asking for help instead
of demanding it, and start showing up in upstream areas looking to fix
things. When I started into this nearly 30 years ago, I thought that
was too slow, too asinine, and too feudal and needed to be shook up.
In the end I learned that the system is the way it is because it works
in good days and bad ones. It isn't ideal but for small volunteer
projects it seems to be the one long term ones go to.

That said, it isn't the only one and if you want to show everyone how
to do it differently.. you are free to set up your own project which
does things the way you need it to. That is how innovation gets
kickstarted when people get too complacent.


-- 
Stephen J Smoogen.

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Yaakov Selkowitz]

On 2018-05-05 01:56, Steven Penny wrote:
> On Sat, 5 May 2018 07:27:22, Marco Atzeri wrote:
>> Jon Yong is doing an hell of job taming a monster program,
>> and your comments are NOT appropriated.
> 
> certainly not by you - and probably others on this list.

The question is, if you actually understand your comments are not
appreciated, why do you insist on making them anyway?

> GCC as an example is a fast updating package.

No, not really.

> and as "politically incorrect" as it might be, jon could be doing a
> better job in regards to velocity of release 

Jon is doing just fine, and has our full confidence and backing.  We
simply don't have the manpower of large distributions to deal with the
consequences of potential compiler bugs, so we tend to aim for X.3.0.

> at a minimum we should have test versions for GCC 7 and or 8 already 

https://cygwin.com/ml/cygwin-announce/2018-04/msg00034.html
https://cygwin.com/ml/cygwin-announce/2018-05/msg00002.html

> for ALL arches and ALL targets.

Jon does not maintain all of the cross-compilers, and as he already
stated earlier, the native GCC 7 is a prerequisite to the MinGW GCC 7
because GNAT is written in Ada.

As for the aforementioned cygwin*-gcc, those are 32<->64 bit
cross-compilers.  I dropped Ada from those because there really isn't a
use for it since the original 64-bit bootstrapping.

>> and I do not see your maintaining efforts.
> 
> not sure what you mean by this - no i havent maintained an official cygwin
> package

Exactly.  Griping (or worse) at those who are actually doing the work
while you contribute nothing doesn't get you very far.  Perhaps the
following will help you gain some perspective:

Almost 10 years ago, I was the one being similarly targeted, and an
"older" (at least wrt Cygwin), now former, contributor to Cygwin
rightfully came to my defence in epic fashion:

https://cygwin.com/ml/cygwin-xfree/2008-11/msg00242.html

I have never forgotten that, and hence am prepared to similarly defend
my fellow contributors (albeit probably not as eloquently).

>> This only show how ignorant you are about programming language and
>> their usage. Clearly you are not an engineer.
> 
> id say the same to you - as you seen keen on throwing insults rather
> than links and examples - cheers.

You are the one who insists on throwing insults time and again, and I
have warned you about your tone before.  I suggest you take heed.

-- 
Yaakov

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Steven Penny]

On Sun, 6 May 2018 00:54:23, Yaakov Selkowitz wrote:
> The question is, if you actually understand your comments are not
> appreciated, why do you insist on making them anyway?

because they have merit? i said that already.

>> GCC as an example is a fast updating package.
>
> No, not really.

its fast in comparison to cygwin releases.

> Jon does not maintain all of the cross-compilers,

yes he does?

http://cygwin.com/cygwin-pkg-maint

> Exactly.  Griping (or worse) at those who are actually doing the work
> while you contribute nothing doesn't get you very far.

this is just patently false. while my builds may not be "blessed" by cygwin,
they are available for anyone to use, and have been for several years.

> I have never forgotten that, and hence am prepared to similarly defend
> my fellow contributors (albeit probably not as eloquently).

thats nice, but i think your "defense" is unwarranted. all i am asking here is
for at least yearly updates, even if in the form of "[test]" packages, of
important packages. for example, with these:

1. gcc-core
2. gcc-g++
3. mingw64-x86_64-gcc-core
4. mingw64-x86_64-gcc-g++

only 2 of the 4 would meet that criteria. as i was tired of waiting, i built 3
and 4 myself:

http://github.com/svnpenn/glade/tree/master/mingw-w64-x86-64/gcc

and to my surprise with the right options a build only takes about 15 minutes.
so its doesnt seem like a crazy request to me, to have a test build uploaded in
this case.

> You are the one who insists on throwing insults time and again, and I
> have warned you about your tone before.  I suggest you take heed.

do i though? seriously i would like to know. i went through my posts on this
thread:

- http://cygwin.com/ml/cygwin/2018-05/msg00086.html
- http://cygwin.com/ml/cygwin/2018-05/msg00082.html
- http://cygwin.com/ml/cygwin/2018-05/msg00076.html
- http://cygwin.com/ml/cygwin/2018-05/msg00059.html

and i dont see a single insult that i have made. i might be critical of the
current state of certain packages, but i am not seeing insults here from my end.
take care.


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[cyg Simple]

On 5/6/2018 10:08 AM, Steven Penny wrote:
> On Sun, 6 May 2018 00:54:23, Yaakov Selkowitz wrote:
>> The question is, if you actually understand your comments are not
>> appreciated, why do you insist on making them anyway?
> 
> because they have merit? i said that already.
> 

Since you stated in the form of a question, I can say for me, they do
not and based on the conversation of others, not for anyone but you.

>>> GCC as an example is a fast updating package.
>>
>> No, not really.
> 
> its fast in comparison to cygwin releases.
> 

A base package such as GCC requires time to release to an OS and Cygwin
is a emulation of an OS.  Releasing just because its fresh off the press
isn't going to happen.  You have the opportunity to build it for
yourself if you need it sooner but then you are on your own.

>> Jon does not maintain all of the cross-compilers,
> 
> yes he does?
> 

In the form of a question again, he definitely does a great job.

> http://cygwin.com/cygwin-pkg-maint
> 
>> Exactly.  Griping (or worse) at those who are actually doing the work
>> while you contribute nothing doesn't get you very far.
> 
> this is just patently false. while my builds may not be "blessed" by
> cygwin,
> they are available for anyone to use, and have been for several years.
> 

Why are you not "blessed" with your work?  If you provide a service to
Cygwin then why aren't you in the cygwin-pkg-maint list?  It's because
you haven't requested to maintain a package.

>> I have never forgotten that, and hence am prepared to similarly defend
>> my fellow contributors (albeit probably not as eloquently).
> 
> thats nice, but i think your "defense" is unwarranted. all i am asking
> here is
> for at least yearly updates, even if in the form of "[test]" packages, of
> important packages. for example, with these:
> 
> 1. gcc-core
> 2. gcc-g++
> 3. mingw64-x86_64-gcc-core
> 4. mingw64-x86_64-gcc-g++
> 

And then you offend the maintainers who provide their time to provide
you a service of distribution.  If you want to maintain or co-maintain a
package then ask to see if the maintainer needs help.  You haven't done
that, you DEMAND that a release be made.  You might not see it as a
DEMAND but it is one.

> only 2 of the 4 would meet that criteria. as i was tired of waiting, i
> built 3
> and 4 myself:
> 

And you a free to do so.  MinGW isn't GCC and those providing a package
based on it do so in their free time as a gift to you and me.

> http://github.com/svnpenn/glade/tree/master/mingw-w64-x86-64/gcc
> 
> and to my surprise with the right options a build only takes about 15
> minutes.
> so its doesnt seem like a crazy request to me, to have a test build
> uploaded in
> this case.
> 

Maybe to you it doesn't seem like a crazy DEMAND but perhaps there are
reasons you're unaware of.  Ask to help rather than DEMANDing.

>> You are the one who insists on throwing insults time and again, and I
>> have warned you about your tone before.  I suggest you take heed.
> 
> do i though? seriously i would like to know. i went through my posts on
> this
> thread:
> 
> - http://cygwin.com/ml/cygwin/2018-05/msg00086.html
> - http://cygwin.com/ml/cygwin/2018-05/msg00082.html
> - http://cygwin.com/ml/cygwin/2018-05/msg00076.html
> - http://cygwin.com/ml/cygwin/2018-05/msg00059.html
> 
> and i dont see a single insult that i have made. i might be critical of the
> current state of certain packages, but i am not seeing insults here from
> my end.
> take care.

While your intention isn't one of insults those who maintain the
packages read them as such because you DEMAND more of their time with no
effort toward progressing Cygwin from yourself.  That is the insult.

-- 
cyg Simple

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Steven Penny]

On Mon, 14 May 2018 10:31:18, cyg Simple wrote:
>> because they have merit? i said that already.
>> 
>
> Since you stated in the form of a question, I can say for me, they do
> not and based on the conversation of others, not for anyone but you.

let me rephrase: they have merit, full stop. Example 1, quoting myself:

> yaakov said he would look into it after 2.5.1 came out

fact. here is the proof:

http://github.com/cygwinports/ruby/issues/1

Example 2, quoting myself again:

> and its out:

fact. here is the proof:

http://github.com/ruby/ruby/releases/tag/v2_5_1

Example 3, quoting myself again:

> and yaakov handles a massive amount of packages

fact. here is the proof:

http://cygwin.com/cygwin-pkg-maint

Example 4, quoting myself again:

> cygwin *does* keep up to date with *some* important packages

fact. example is the Git package, which as of this writing is totally up to
date:

- http://cygwin.mirrors.hoobly.com/x86_64/release/git
- http://github.com/git/git/releases

Example 5, quoting myself again:

> but not other important packages

fact. Current Cygwin Ruby is 2.3, which was released Dec 2015:

- http://cygwin.mirrors.hoobly.com/x86_64/release/ruby
- http://wikipedia.org/wiki/Ruby_%28programming_language%29#Ruby_2.3

> A base package such as GCC requires time to release to an OS and Cygwin
> is a emulation of an OS.

thats what test package is for

> Releasing just because its fresh off the press isn't going to happen.

like me, i dont see you on this list:

http://cygwin.com/cygwin-pkg-maint

so i'd say you arent in a position to say that. test packages allow this to
happen if the maintainer chooses.

> You have the opportunity to build it for yourself if you need it sooner but
> then you are on your own.

I have mentioned twice already that i have done this:

- http://cygwin.com/ml/cygwin/2018-05/msg00099.html
- http://cygwin.com/ml/cygwin/2018-05/msg00082.html

>>> Jon does not maintain all of the cross-compilers,
>> 
>> yes he does?
>> 
>
> In the form of a question again, he definitely does a great job.

what criteria are you basing this comment on? i am not arguing with you on this
point per se. i have given concrete arguments here where release velocity could
be improved:

- http://cygwin.com/ml/cygwin/2018-05/msg00099.html
- http://cygwin.com/ml/cygwin/2018-05/msg00086.html
- http://cygwin.com/ml/cygwin/2018-05/msg00082.html
- http://cygwin.com/ml/cygwin/2018-05/msg00076.html

> Why are you not "blessed" with your work?  If you provide a service to
> Cygwin then why aren't you in the cygwin-pkg-maint list?  It's because
> you haven't requested to maintain a package.

perhaps you should read the entire thread - i see you missed my other posts,
but it seems you missed Smoogen as well:

http://cygwin.com/ml/cygwin/2018-05/msg00087.html

> And then you offend the maintainers who provide their time to provide
> you a service of distribution.

thats your opinion, and its not germane to this topic. the original topic is
requesting a ruby release, as it is out of date.

> If you want to maintain or co-maintain a package then ask to see if the
> maintainer needs help.

i dont want to do that, but i will volunteer if a spot opens.

> You haven't done that, you DEMAND that a release be made. You might not see it
> as a DEMAND but it is one.

sry, nop.

> And you a free to do so.  MinGW isn't GCC

yes it is. when you compile GCC, as i have done:

http://github.com/svnpenn/glade/blob/master/mingw-w64-x86-64/gcc

you choose a target. normally for this community that is "x86_64-pc-cygwin", but
in my case it is "x86_64-w64-mingw32". but you dont use some magical "MinGW"
repo, its the same GCC. granted, you do need to also install the target
"binutils", "headers" and "runtime", but the same source is used to build GCC
itself.

> Maybe to you it doesn't seem like a crazy DEMAND but perhaps there are
> reasons you're unaware of.  Ask to help rather than DEMANDing.

again not a demand.

> While your intention isn't one of insults those who maintain the
> packages read them as such because you DEMAND more of their time with no
> effort toward progressing Cygwin from yourself.  That is the insult.

i am not demanding anything. i am stating what i think are reasonable
expectations for any software community. if they want to ignore them, thats
their business. ive already moved on, i build my own GCC, which i will use until
the official one drops. however note this: GCC 7 was released over a year ago:

http://gnu.mirrors.hoobly.com/gcc/gcc-7.1.0

if a test package was dropped at that time, we could have tested for 6 months,
then released a final build, and still been 6 months ahead of the current
schedule.


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: What is Cygwin and MinGW [WAS: Request new Ruby release]

[cyg Simple]

I'm moving this to the Cygwin-Talk list.

On 5/14/2018 7:48 PM, Steven Penny wrote:
> On Mon, 14 May 2018 10:31:18, cyg Simple wrote:

>> And you a free to do so.  MinGW isn't GCC
> 
> yes it is. when you compile GCC, as i have done:
> 
> http://github.com/svnpenn/glade/blob/master/mingw-w64-x86-64/gcc
> 
> you choose a target. normally for this community that is
> "x86_64-pc-cygwin", but
> in my case it is "x86_64-w64-mingw32". but you dont use some magical
> "MinGW"
> repo, its the same GCC. granted, you do need to also install the target
> "binutils", "headers" and "runtime", but the same source is used to
> build GCC
> itself.
> 

And given this you think Cygwin is also GCC.  WRONG, MinGW and Cygwin
are just the runtime libraries required to build GCC and other packages.
 GCC is a separate entity that has its own process of accepting patches
for using runtime libraries.  Binutils also is a separate entity that
uses even different methods of patching requirements as both Cygwin and
GCC.  Anyone who states that the GCC product is an entity other than GCC
isn't thinking clearly about the various pieces and parts.

Now to make Cygwin or MinGW useful you need the other pieces and parts
but those pieces and parts do not become the entity of Cygwin or MinGW.
Cygwin and MinGW remain a separate entity providing the runtime
libraries to use to build other pieces and parts.

-- 
cyg Simple

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: Request new Ruby release

[Adam Dinwoodie]

On Tue, 15 May 2018 at 00:49, Steven Penny wrote:
> fact. example is the Git package, which as of this writing is totally up
to
> date:

> - http://cygwin.mirrors.hoobly.com/x86_64/release/git
> - http://github.com/git/git/releases

I've been avoiding this thread as I haven't had anything productive to add.
However a package I maintain has now been invoked as an example, and I feel
compelled to respond; I am not comfortable with my work being held up to
shame other volunteer maintainers.

I entirely understand the desire for up-to-date packages to be available,
and clearly there are vast swathes of our packages that aren't kept at the
bleeding edge of the upstream release tracks. However I don't think this
email trail is a good way to encourage maintainers to put in the effort
required to keep things up to date. Maintainers, at least for the most
part, are going to be well aware when their packages aren't the latest
code, and equally be aware of the possibilities of releasing test packages.
Pointing these things out, particularly repeatedly, is unlikely to be well
received.

For anyone who wants to see Cygwin packages updated, a single prompt to the
maintainer is occasionally useful, but further prompts are likely to be
taken as hassling, not as helpful. Beyond that, I expect most maintainers
will have an idea of what other people could do to help if someone offered;
I know for myself I can point at specific things other people could offer
to help for each of my packages, and I imagine other maintainers would also
be able to answer a question of "I'd like to see this updated, what can I
do to help?", even if the answer is sometimes going to be "nothing".

As others have pointed out, the Cygwin package maintainers are volunteers,
using their own time and resources. I don't think any of them will
appreciate being described as acting in bad faith, myself included (and
while Git is up-to-date, other packages I maintain aren't, for a variety of
reasons, so I very much take the description as being levelled at me as
much as anyone else).

If someone genuinely thinks a package maintainer is acting in bad faith, or
otherwise not keeping up with their responsibilities as a maintainer, *and*
they're able and willing to put in the effort to take over the
maintainership, that's a discussion we can have. However, certainly in the
cases of Ruby and GCC and their related packages, as Yaakov has said, I
don't think there's any question from the other maintainers or the folk who
run the Cygwin project as a whole, that they are working in good faith and
keeping up with their responsibilities to the project, even if individuals
would prefer different behaviour.

This thread is not going to get GCC, Ruby, or any other package updated
more quickly than it otherwise would. I'd recommend folk who want to see
what they can do to get things updated ask what they can do to help, and
accept that the answer may be "nothing".

Adam

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple