Which versions of Ruby should Cucumber support?

[Mike Sassak]

Hello Cucumber Ruby Users,

A question about the state of adoption of Ruby 1.9.x: what would you
say if Cucumber dropped support for Ruby versions < 1.8.7? What about
dropping support for 1.8.x entirely? How many test suites would burst
into flames if that happened? Rails doesn't support 1.8.6 anymore, but
I don't think many of the big Ruby projects are dropping 1.8.7 just
yet, but one day it will happen. Please note that if anything happens
to Cucumber in this regard, it will happen slowly and deliberately.
You won't wake up one morning to find that all your 1.8.x suites broke
with a bump of the tiny version number.

What are your thoughts about this?
Mike

[aslak hellesoy]

On Wed, Aug 31, 2011 at 4:49 PM, Mike Sassak <msa...@gmail.com> wrote:

Hello Cucumber Ruby Users,

A question about the state of adoption of Ruby 1.9.x: what would you
say if Cucumber dropped support for Ruby versions < 1.8.7? What about
dropping support for 1.8.x entirely? How many test suites would burst
into flames if that happened?

None. Projects who are not on 1.9.x wouldn't upgrade to a 1.9.x-only Cucumber.

Rails doesn't support 1.8.6 anymore, but
I don't think many of the big Ruby projects are dropping 1.8.7 just
yet, but one day it will happen. Please note that if anything happens
to Cucumber in this regard, it will happen slowly and deliberately.
You won't wake up one morning to find that all your 1.8.x suites broke
with a bump of the tiny version number.

What are your thoughts about this?

I don't see a problem supporting 1.8.7 unless it requires a lot of extra effort. TBH I haven't been testing it extensively on 1.8.7 lately so it might already have 1.8.7-related bugs.

Aslak

 

Mike

--
You received this message because you are subscribed to the Google Groups "Cukes" group.
To post to this group, send email to cu...@googlegroups.com.
To unsubscribe from this group, send email to cukes+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/cukes?hl=en.

[David Kowis]

If something as prevalent as Cucumber were to start an end of life
process it might help further getting people onto 1.9.x.

However, many of the enterprise ruby stuff still runs 1.8.x. That
being said, they can continue to use older cucumber that doesn't break
support for it.

Also, are there any advantages to dropping 1.8.x support? Does it save
the core team time? Does it make it perform better?

David

[Eumir Gaspar]

On Thu, Sep 1, 2011 at 4:38 AM, David Kowis <dko...@gmail.com> wrote:

On Wed, Aug 31, 2011 at 10:49 AM, Mike Sassak <msa...@gmail.com> wrote:
> Hello Cucumber Ruby Users,
>
> A question about the state of adoption of Ruby 1.9.x: what would you
> say if Cucumber dropped support for Ruby versions < 1.8.7? What about
> dropping support for 1.8.x entirely? How many test suites would burst
> into flames if that happened? Rails doesn't support 1.8.6 anymore, but
> I don't think many of the big Ruby projects are dropping 1.8.7 just
> yet, but one day it will happen. Please note that if anything happens
> to Cucumber in this regard, it will happen slowly and deliberately.
> You won't wake up one morning to find that all your 1.8.x suites broke
> with a bump of the tiny version number.
>
> What are your thoughts about this?

By dropping support, what do you mean? Like just ignore new bugs that are discovered? Or will this be more of a we all have to switch to 1.9 because cucumber will no longer run in 1.8? If it is the former, then I think that is a good thing, much like Rails did with 1.8.6. With the latter, it will be more troublesome with legacy apps that rely on this. 

I know that upgrading is, of course, for the better, but what about companies that don't have the budget for that? You can't just say "Sorry, but we have to charge you more since we'll have to migrate our test suite to 1.9...unless you don't want tests" or something to clients, right? 

In any case, I think that you just mean not thinking about 1.8 when creating new features and ignoring any new bugs so I guess that's all right. Just my two cents :)

--
Eumir Gaspar
Ruby on Rails Developer/Rails UI Specialist

github: http://github.com/corroded
tech blog: http://aelogica.com/author/eumir/

odesk: http://www.odesk.com/users/~~8951a3b425021bda

stackoverflow: http://stackoverflow.com/users/334545/corroded

working with rails: http://www.workingwithrails.com/person/19554-eumir-gaspar

linked in: http://ph.linkedin.com/pub/eumir-gaspar/7/969/98a

[Andrew Premdas]

> --
> You received this message because you are subscribed to the Google Groups "Cukes" group.
> To post to this group, send email to cu...@googlegroups.com.
> To unsubscribe from this group, send email to cukes+un...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/cukes?hl=en.
>
>

I think a widely used testing tool like Cucumber should be expected to
work on all currently supported versions of Ruby, and I would define
"supported versions" to mean every ruby that is supported by ruby.lang
and currently to mean between now and x years ago. So cucumber drops
support later than ruby.lang. In other words it should ideally be very
conservative about dropping language support.

Assuming that this would kill development then we need a good way to
gently drop versions and leave people with a good outlet if they have
to continue using a really old cucumber. With something as serious is
dropping support for a major ruby version would forking the project
and releasing a legacy gem be a viable solution

e.g. release cucumber1.8.7 gem or a cucumber1.8.x gem

I have no idea if this is practical or sensible, but is does seem to
me that Cucumber will be very useful for working with legacy ruby and
rails projects for many years to come, and that just specifying an
older version of the gem to use is not enough for something so major
as say dropping 1.8.x support.

All best

Andrew

--
------------------------
Andrew Premdas
blog.andrew.premdas.org

[Mike Sassak]

Hi Eumir,

To clarify: there is no possibility I can see of Cucumber dropping
support for 1.8.x in the 1.x series. The code has already been
written, after all, and it would be silly to remove it. Neither do I
think anyone would just start ignoring patches for 1.x, or fail to fix
serious problems or regressions in its behavior. That doesn't serve
anyone well. For new major versions with new code and features,
however, the situation is potentially different.

Mike

> --
> Eumir Gaspar
> Ruby on Rails Developer/Rails UI Specialist
> github: http://github.com/corroded
> tech blog: http://aelogica.com/author/eumir/
> odesk: http://www.odesk.com/users/~~8951a3b425021bda
> stackoverflow: http://stackoverflow.com/users/334545/corroded
> working with
> rails: http://www.workingwithrails.com/person/19554-eumir-gaspar
> linked in: http://ph.linkedin.com/pub/eumir-gaspar/7/969/98a
>

[Deryl R. Doucette]

You are familiar with the fact that Rails 4.0 is already slated, and its been announced, to drop all 1.8.7 support right? I would say to follow that path alongside Rails. The Ruby group, iirc, is slated to EOL 1.8.x as well beyond the requisite security patch backporting. Least I do believe I remember reading that on one of the related MLs.

[George Dinwiddie]

On 8/31/11 12:06 PM, aslak hellesoy wrote:
>
>
> On Wed, Aug 31, 2011 at 4:49 PM, Mike Sassak <msa...@gmail.com
> <mailto:msa...@gmail.com>> wrote:
>
> Hello Cucumber Ruby Users,
>
> A question about the state of adoption of Ruby 1.9.x: what would you
> say if Cucumber dropped support for Ruby versions < 1.8.7? What about
> dropping support for 1.8.x entirely? How many test suites would burst
> into flames if that happened?
>
>
> None. Projects who are not on 1.9.x wouldn't upgrade to a 1.9.x-only
> Cucumber.
>
> Rails doesn't support 1.8.6 anymore, but
> I don't think many of the big Ruby projects are dropping 1.8.7 just
> yet, but one day it will happen. Please note that if anything happens
> to Cucumber in this regard, it will happen slowly and deliberately.
> You won't wake up one morning to find that all your 1.8.x suites broke
> with a bump of the tiny version number.
>
> What are your thoughts about this?
>
>
> I don't see a problem supporting 1.8.7 unless it requires a lot of extra
> effort. TBH I haven't been testing it extensively on 1.8.7 lately so it
> might already have 1.8.7-related bugs.

I'm using jruby-1.6.2, which I believe defaults to 1.8.7 compatibility.
As far as I know, the 1.9 compatibility still lags behind, a little.

I would be wary about dropping support for old versions too quickly.
Larger companies are slow to upgrade, and would drop Cucumber for less
capable tools if the tests required too much baby-sitting to keep them
running.

- George

--
----------------------------------------------------------------------
* George Dinwiddie * http://blog.gdinwiddie.com
Software Development http://www.idiacomputing.com
Consultant and Coach http://www.agilemaryland.org
----------------------------------------------------------------------

[Everton Moreth]

Sorry for the bump, but this decision is quite important for me and all the Company.

We are doing a full reorganization, and we need to know if we'll need to put efforts on upgrading all applications to 1.9.2.
We have currently 18 large applications running on top of 1.8.7, and a full upgrade will not be easy.

Is there a way we can get usage statistics ? Or even a poll ?? Or anything like that ?

EMoreth

--

You received this message because you are subscribed to the Google Groups "Cukes" group.
To post to this group, send email to cu...@googlegroups.com.

To unsubscribe from this group, send email to cukes+unsubscribe@googlegroups.com.

[aslak hellesoy]

On Thu, Sep 8, 2011 at 3:44 PM, Everton Moreth <everton...@gmail.com> wrote:

Sorry for the bump, but this decision is quite important for me and all the Company.

We are doing a full reorganization, and we need to know if we'll need to put efforts on upgrading all applications to 1.9.2.
We have currently 18 large applications running on top of 1.8.7, and a full upgrade will not be easy.

Let's say future versions of Cucumber use 1.9-only features. How would that affect you? Couldn't you just use an older version? 

Aslak

 

To unsubscribe from this group, send email to cukes+un...@googlegroups.com.

[Everton Moreth]

Our development team is pretty straight forward, we started working with Rails 3 and 3.1 (even the RCs) as soon as they came out.
Unfortunately our server side team is pretty lazy... And they always have some excuse to not upgrade ruby.

Keeping 1.8.7 and being unable to follow cucumber evolution would not be good. We are even starting to study the new Cucumber JVM to start using it on our Java projects...

Coincidently we are changing lots of stuff here, and upgrading ruby was not (again) a priority. But if cucumber drops support, it probably will.

EMoreth

[Aslak Hellesøy]

On Sep 8, 2011, at 16:57, Everton Moreth <everton...@gmail.com> wrote:

Our development team is pretty straight forward, we started working with Rails 3 and 3.1 (even the RCs) as soon as they came out.
Unfortunately our server side team is pretty lazy... And they always have some excuse to not upgrade ruby.

Keeping 1.8.7 and being unable to follow cucumber evolution would not be good. We are even starting to study the new Cucumber JVM to start using it on our Java projects...

Coincidently we are changing lots of stuff here, and upgrading ruby was not (again) a priority. But if cucumber drops support, it probably will.

The 1.x series will most likely continue to support 1.8.7. There is no obvious benefit/reason to drop support for it.

However, if a 2.x version (which would probably be a rewrite to get rid of old cruft) sees the light one day, it might be easier to drop 1.8.7 support. But maybe not.

It's too early to say if or when this will happen.

Aslak