scaling down ruote

3459 views
Skip to first unread message

John Mettraux

unread,
Nov 15, 2013, 5:27:31 PM11/15/13
to ruote

Hello all,

my goal would be to go on with ruote 3.0.

The problem I have: I still have to maintain ruote 2.x, it's OK, these days
it's mostly answering one email per week. And Danny and Torsten are helping
(thanks guys!).

The time I can devote to open source has been reduced to 1 or 2 hours per
day. I know it sounds like a luxury, but it's quickly killed by an email on
the mailing list, and I also have other, smaller open source projects.

There are a few bugs in ruote 2.x that are hard to kill for me. I consider
them a direct result of the design decisions that went into ruote 2.x, hence
my wish to start working on ruote 3.0.

Somehow, I want to dissuade people from using ruote. No new projects. I'll do
my best to help people maintain their existing installs. That's basically
what I did for the whole year 2013.

Ruote 3.0, if I manage to find the time to build it, will probably not be a
Ruby project, but it should be easily accessible, as a service, from a Ruby
application.


Kind regards,

--
John Mettraux - http://lambda.io/jmettraux

Steve Tuckner

unread,
Nov 15, 2013, 6:18:15 PM11/15/13
to openwfe...@googlegroups.com
Are the "bugs you would like to kill" part of the five open issues on the ruote github account? I wouldn't mind looking at some of them if you'd like the help...

Steve
> --
> --
> you received this message because you are subscribed to the "ruote users" group.
> to post : send email to openwfe...@googlegroups.com
> to unsubscribe : send email to openwferu-use...@googlegroups.com
> more options : http://groups.google.com/group/openwferu-users?hl=en
> ---
> You received this message because you are subscribed to the Google Groups "ruote" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to openwferu-use...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

John Mettraux

unread,
Nov 15, 2013, 6:27:42 PM11/15/13
to openwfe...@googlegroups.com

On Fri, Nov 15, 2013 at 05:18:15PM -0600, Steve Tuckner wrote:
>
> Are the "bugs you would like to kill" part of the five open issues on the
> ruote github account?

Hello Steve,

yes.

There is also this one: https://groups.google.com/forum/#!topic/openwferu-users/rd0pnazNxu4

> I wouldn't mind looking at some of them if you'd like
> the help...

Well, most of them require understanding how the thing runs. That'd probably
end up in me having to help you help me.

I think it's better if there are no new projects using ruote.

Mario Camou

unread,
Nov 16, 2013, 1:16:06 AM11/16/13
to ruote

Hi John,


>
> On Nov 15, 2013 11:27 PM, "John Mettraux" <jmet...@gmail.com> wrote:
> > Ruote 3.0, if I manage to find the time to build it, will probably not be a
> > Ruby project, but it should be easily accessible, as a service, from a Ruby
> > application.

That's interesting. Care to share your thoughts on why not Ruby, what language or sort of language you're thinking about and why?

Cheers,
-Mario.

John Mettraux

unread,
Nov 16, 2013, 3:26:07 AM11/16/13
to openwfe...@googlegroups.com
Hello Mario,

I went too wide with ruote 2.x, one "adapter" for any sort of storage out
there. Ruote 3.0 will [perhaps] focus on FS storage. There'll be a set of
"conventional datastructures" manipulated by the tools that'd make up ruote
3.0, whatever the language they're written in.

I would stay away from Ruby to avoid any "hey I want to integrate that into
my Rails project, where is the sample I can copy/paste?" hassle.

I also thought that going for Ruby would allow people helping build the
system, but no, it didn't, so I'm free to build it in any language I like.
What's important (you Madrid guys know it very well), is that other systems
can talk to it.


Best regards,

Mario Camou

unread,
Nov 16, 2013, 6:28:00 AM11/16/13
to openwfe...@googlegroups.com
I completely agree with you. Elixir perhaps? Or Scala or Go :)

I hope you get the time to do this!

Danny Fullerton

unread,
Nov 16, 2013, 2:22:58 PM11/16/13
to openwfe...@googlegroups.com
Hello,

I can't help but wonder how this would change anything. Ruote's users
gravitates around Ruby so I don't see how changing the language of
version 3 would fix this. Better to simply start a whole new project.

Sure scaling down some features, reducing the codebase, etc can help but
I'm really not quite sure about the language thing...

It's true we see a lot of inquiries about "how to integrate with Rails"
but that's probably because there's a need. What if `we` (Ruote's users
- not John) try to fix this: develop a project which helps integrating
Ruote with Rails (not Ruote in Rails) while keeping a clean demarcation
between both of them? It's precisely what I'm working on. Could that be
helpful?

Too little to late?

regards,

--
Danny Fullerton
Founder
Mantor Organization


signature.asc

Damon Torgerson

unread,
Nov 16, 2013, 3:13:16 PM11/16/13
to openwfe...@googlegroups.com
As one of the "how do I integrate this with Rails" folks, I really appreciate all of your efforts John. If there is a way that I can help with taking "examples" off of your plate as per Danny's suggestion, I'd like to help. I really like the idea of treating Ruote as a service rather than "integrating" it directly in the app.

Damon Torgerson

John Mettraux

unread,
Nov 16, 2013, 4:34:58 PM11/16/13
to openwfe...@googlegroups.com

On Sat, Nov 16, 2013 at 02:22:58PM -0500, Danny Fullerton wrote:
>
> Hello,

Hello Danny,

thanks for your help these days.


> I can't help but wonder how this would change anything. Ruote's users
> gravitates around Ruby so I don't see how changing the language of
> version 3 would fix this. Better to simply start a whole new project.

Yes, it would be a new project (flon).


> Sure scaling down some features, reducing the codebase, etc can help but
> I'm really not quite sure about the language thing...

Using a developer friendly language like Ruby didn't help build a dynamic. I
started stupid "adapters" that I had to maintain, side projects were started
but left behind (and I had to take over and maintain).

Every time there is help coming, there is lot of good will, but that lasts 1
or 2 months. After that, people have jobs and families and the project with
ruote is over. Someone else maintains the project (and asks questions on the
mailing list).


> It's true we see a lot of inquiries about "how to integrate with Rails"
> but that's probably because there's a need. What if `we` (Ruote's users
> - not John) try to fix this: develop a project which helps integrating
> Ruote with Rails (not Ruote in Rails) while keeping a clean demarcation
> between both of them? It's precisely what I'm working on. Could that be
> helpful?

Rails is a problem, once again it's my fault. Running ruote along with Rails
in the same Ruby runtime should be avoided. I proposed it in the first place,
my bad.

Speaking of examples, according to http://ruote.rubyforge.org/source.html
there is:

* http://github.com/tosch/ruote-on-rails
* http://github.com/threetee/ruote-rails-example

I can't maintain them.


> Too little to late?

I created this situation and I'm trying to get out of it by a) preventing the
creation of new projects with ruote b) supporting existing users as best as I
can.

This year 2013, almost nothing happened, the project is OK (a few bugs, yes),
it's little known. I think it's time for me to put it to sleep.

This year too we had (You Danny and I) this little chat where you said "let
ruote be known" or something like that. Were it be known now (after 6 years
of obscurity), I could not support/sustain the increased activity.


Kind regards,

John

John Mettraux

unread,
Nov 16, 2013, 4:38:36 PM11/16/13
to openwfe...@googlegroups.com

On Sat, Nov 16, 2013 at 12:13:16PM -0800, Damon Torgerson wrote:
>
> As one of the "how do I integrate this with Rails" folks, I really
> appreciate all of your efforts John. If there is a way that I can help with
> taking "examples" off of your plate as per Danny's suggestion, I'd like to
> help. I really like the idea of treating Ruote as a service rather than
> "integrating" it directly in the app.

Hello Damon,

thanks for your message.

According to http://ruote.rubyforge.org/source.html, there is

* https://github.com/tosch/ruote-on-rails
* https://github.com/threetee/ruote-rails-example

both unmaintained.

But I prefer to dissuade people from using ruote in their new projects.


Kind regards,

John

Farrel Lifson

unread,
Nov 16, 2013, 4:45:55 PM11/16/13
to openwfe...@googlegroups.com
On Saturday, 16 November 2013 10:26:07 UTC+2, John Mettraux wrote:
I also thought that going for Ruby would allow people helping build the
system, but no, it didn't, so I'm free to build it in any language I like.
What's important (you Madrid guys know it very well), is that other systems
can talk to it.

I hope you reconsider, I was first introduced to Ruote by Kenneth Kalmer at Rubyfuza two years ago and have been using it mainly on some fun personal projects. I have been planning to try and get it introduced into my current work projects where a tool like Ruote is sorely needed.

Farrel

John Mettraux

unread,
Nov 16, 2013, 4:59:24 PM11/16/13
to openwfe...@googlegroups.com
Hello Farrel,

sorry I want to move away from Ruby. I don't like the whole "you're doing it
wrong" hammering and/or "let me sell you an ebook" show business thing.

I know nobody will follow me in the depths of the tool, so I don't need to
use Ruby.


Thanks for the message,

John

Gerente de Sistemas

unread,
Nov 16, 2013, 5:28:38 PM11/16/13
to openwfe...@googlegroups.com
unsubscribe

John Mettraux

unread,
Nov 16, 2013, 5:30:50 PM11/16/13
to openwfe...@googlegroups.com

On Sat, Nov 16, 2013 at 08:28:38PM -0200, Gerente de Sistemas wrote:
>
> unsubscribe

Thanks a lot!

John

Hartog De Mik

unread,
Nov 17, 2013, 7:15:12 AM11/17/13
to openwfe...@googlegroups.com
Ai...

I think Ruby and Ruote are a great fit! And that is mostly because I learned to seperate the Ruote from the Rails (as mentioned above). I love Ruote just for what it is; and I have to agree; Ruby is a terrible language, because it is to flexible, it takes very sturdy developers to understand wtf they are actualy doing in ruby. The fact that nothing is documented (Ruote is a positive exception) does not help.

I would be very interested in your designs ideas for "flon" (sound like something that exits nasal cavities ;-p). Is there any way we could see you make the evolution (and ask pesky questions)?

Kindest of regards,
Hartog

Doug Bryant

unread,
Nov 17, 2013, 11:37:20 AM11/17/13
to openwfe...@googlegroups.com
John,

Tough decision but I agree it's probably the right one to make.  As a standalone service, it should reduce the the number of issues due to reduced number of config options (adapters to storage engines, etc).  I would imagine a fair number of users are standing up a service around Ruote anyway and for those who aren't, having Ruote or Flon as a service would lead to better integration practices.

I don't think Ruote being written in Ruby was as much of the problem as Ruote being a library for workflows.  Being a library/gem allowed the most flexibility integrating Ruote but probably caused most of the confusion about how to use properly. Were Ruote a stand-alone service, it really wouldn't matter which language it was written in. 

I would be very interested in watching and participating in the evolution of ruote 3 or flon also.  Let us know how we can help.

Doug

Christophe Malaurie

unread,
Nov 17, 2013, 3:34:31 PM11/17/13
to openwfe...@googlegroups.com
Dear all
First tx for ruote development and maintenance over the years.
I won't debate on is ruby bad or good, bad developers and more importantly bad designers will succeed in project failures even with perfect tools.
I have been using ruote together with rails and rails was the pain, neither ruote nor the integration of the two worlds.
A few comments then (and may be a stream for "flon" could be created):
- ruote concepts are far beyond any average designer/developer understanding because at start they do not exactly get what is a business process. look how many questions you had about participants and what is it...
- ruote is may be too low level: i m pretty sure that if you wrap up your tool in a sexy GUI more people will like it. A GUI with predefined actions would probably restrict the use of ruote but it could gain in accessibility
- ruote admin is probably again too low level although there were very good things available with ruotekit. but i experienced a hand over case and people were scared by it.
Anyway, many thanks again and let us know about your new flon
Christophe

John Mettraux

unread,
Nov 17, 2013, 4:11:32 PM11/17/13
to openwfe...@googlegroups.com

On Sun, Nov 17, 2013 at 12:34:31PM -0800, Christophe Malaurie wrote:
>
> (...)
>
> - ruote concepts are far beyond any average designer/developer
> understanding because at start they do not exactly get what is a business
> process. look how many questions you had about participants and what is
> it...
> - ruote is may be too low level: i m pretty sure that if you wrap up your
> tool in a sexy GUI more people will like it. A GUI with predefined actions
> would probably restrict the use of ruote but it could gain in accessibility
> - ruote admin is probably again too low level although there were very good
> things available with ruotekit. but i experienced a hand over case and
> people were scared by it.

Hello Christophe,

many thanks for that piece of feedback. One of the things I want for flon is
better tools to visualize/decide/act.

I can make well-defined interfaces and people can build clients out of it.
But those clients tend to become abandonware. Perhaps people built better
things behind their company walls, but I don't know of them.

Kind regards,

John

Idan Moyal

unread,
Nov 25, 2013, 3:15:06 AM11/25/13
to openwfe...@googlegroups.com
Hi John,

Good work!

Would really love to see a Python version of Ruote :/

John Mettraux

unread,
Nov 25, 2013, 7:11:04 PM11/25/13
to openwfe...@googlegroups.com

On Mon, Nov 25, 2013 at 12:15:06AM -0800, Idan Moyal wrote:
>
> Good work!
>
> Would really love to see a Python version of Ruote :/

There's a half a dozen Python people on this list. One of you should start
something ;-)


Cheers,

Or Cohen

unread,
Dec 2, 2013, 5:38:40 AM12/2/13
to openwfe...@googlegroups.com

On Monday, November 25, 2013 10:15:06 AM UTC+2, Idan Moyal wrote:

Would really love to see a Python version of Ruote :/

Out of this list: https://wiki.openstack.org/wiki/NovaOrchestration/WorkflowEngines, Spiff workflow seems interesting. 

Shubham Bansal

unread,
Dec 16, 2013, 4:33:49 PM12/16/13
to openwfe...@googlegroups.com
Hi John - Thanks for the awesome work on ruote. I understand that you are trying to scale down Ruote, but can you please move the documentation out of RubyForge onto github. It would be extremely helpful for people who are already using Ruote for existing projects. I would be more than happy to help with this.

John Mettraux

unread,
Dec 16, 2013, 4:54:11 PM12/16/13
to openwfe...@googlegroups.com

On Mon, Dec 16, 2013 at 01:33:49PM -0800, Shubham Bansal wrote:
>
> Hi John - Thanks for the awesome work on ruote. I understand that you are
> trying to scale down Ruote, but can you please move the documentation out
> of RubyForge onto github. It would be extremely helpful for people who are
> already using Ruote for existing projects. I would be more than happy to
> help with this.

Hello Shubham,

it's actually already on GitHub:

https://github.com/jmettraux/ruote_website


Best regards,

John Mettraux

unread,
Dec 16, 2013, 5:21:04 PM12/16/13
to openwfe...@googlegroups.com

On Tue, Dec 17, 2013 at 06:54:11AM +0900, John Mettraux wrote:
>
> On Mon, Dec 16, 2013 at 01:33:49PM -0800, Shubham Bansal wrote:
> >
> > Hi John - Thanks for the awesome work on ruote. I understand that you are
> > trying to scale down Ruote, but can you please move the documentation out
> > of RubyForge onto github. It would be extremely helpful for people who are
> > already using Ruote for existing projects. I would be more than happy to
> > help with this.
>
> Hello Shubham,
>
> it's actually already on GitHub:
>
> https://github.com/jmettraux/ruote_website

And the answer to the next question is:

http://ruote.io/index.html

Shubham Bansal

unread,
Dec 16, 2013, 5:29:33 PM12/16/13
to openwfe...@googlegroups.com
Awesome. Thanks for the quick reply.

Freddie Peña

unread,
Jan 22, 2014, 5:39:00 AM1/22/14
to openwfe...@googlegroups.com
Sorry for replying this late to this, but I just wanted to say thanks to John and everyone else that was involved with Ruote 1.x-2.x. I learned a lot about workflows, coding, and open source projects in general from your support on this mailing list and IRC.

Whatever you choose to do with Ruote 3.0, if it ever comes down to it, I know it will be great :).

Cheers!
Reply all
Reply to author
Forward
0 new messages