The Merb framework for Ruby has reached the big 1.0 with an impressive list of features. If you’ve never heard of it, consider Merb a second generation Ruby framework that takes many of the (sometime painful) lessons learned in Rails and applies them to a blank canvas.
I’ve predicted in Gartner research that the Ruby developer community will reach 4 million programmers by 2013. The framework that captures and retains the bulk of these developers will be a big deal. Until now Rails has been the only game in town. That’s about to change.
<back story>
Everyone in the Ruby developer community is familiar with the growing pains of Rails. Scalability questions continue to run rampant as both the framework and the collective RoR community mature over time in real world test cases. Moreover, there are some who just aren’t happy with the close knit community of developers who are evolving the core framework.
This is nothing new. Nobody gets it perfect the first time. We should expect RoR to go through the standard maturity curve and be suspicious if it didn’t. Many argue that Rails will navigate this trial by fire and emerge as a strong and solid foundation for the long term efforts of the larger Ruby developer base.
Others however insist that Rails will be burned to a crisp. They argue that Rails represents is a genetic dead end (Cro-magnon Ruby?). This camp insists that the design defects in Rails are so innate that, while it looks really nice on the surface, the framework lacks the fundamental DNA to stand up to real world heavy lifting.
I’m not expounding either position here. I’ve worked with Rails. I like Rails very much. But then again. I depend heavily on discussions with real world mainstream IT shops to determine when and if a technology has completed its trial by fire.
Toward this end, virtually all the mainstream (Gartner calls them type B) IT shops that have dabbled in RoR so far have built relatively small to medium sized “line of business” solutions and prototypes. They like Rails as well. But all are concerned with safe investments as well. In other words, as an IT manager what happens if my only RoR expert gets hit by a bus on the way to work?
If you develop RoR applications I’d love to here from you. Post a comment here or email me directly at mark.driver@gartner.com
I believe the final chapter on RoR is far from complete. It is clear however that Rails has a long way to go before we can consider it mature for what Gartner calls “systematic” development efforts (large, mission critical, long life cycle, bet your business) beyond the bleeding edge dot com types.
</back story>
So now we have Merb as well. There are other contending frameworks but Merb has a pedigree and some very solid design elements (thread safe yeah!) behind it that might actually be able to give Rails a run for its money.
Personally I’m torn. On one hand Ruby needs consistency in order to reach a real critical mass. One framework to rule them all has its advantages. But choice is good as well.
IF we are gong to see Merb make a real run at Rails then we need…
1) some high profile early adopters to show that it works in the real world
2) solid code examples and documentation for configurations, etc.
3) lots of cool plug-ins
4) good support from tools. Ideally some nice Eclipse plug-ins but also direct support from both 3rd Rail and NetBeans (both solid Ruby IDEs)
5) good books from good publishers; I’m talking to you Addison Wesley and O’Reilly.
6) some buzz from the trenches. Have you switched from Rails to Merb? I’d love to hear from you (see above).
7) inally given the significant head start that Rails enjoys, Merb must prove itself significantly superior to Rails in order to close the gap and take the lead. “As good” wont do it. It’s far too late for that. Merb must demonstrate clear and irrefutable advantages to create enough of a gravity well to draw away the hordes of developers who have committed considerable blood to the Rails cause to-date.
Bottom line:
Are you a RoR developer or the manager of a RoR developer? Then you owe it to yourself to check out Merb and keep an eye on it. You might like what you see enough to make a change. You probably won’t today but you may change your mind over the course of the next 18 months.
In any event, don’t assume the considerable early momentum that Rails has enjoyed as the first “killer app” for Ruby will translate into long term success. Take a good look at both frameworks and expect continued changes as they both evolve over the next couple of years.
Category: Uncategorized Tags: appdev
13 responses so far ↓
1 Warren Henning October 10, 2008 at 2:16 pm
You may be interested to watch the webcasts from MerbCamp that will be going on this weekend. http://www.merbcamp.com
2 Leah Silber October 10, 2008 at 10:05 pm
Also, two Merb books in the works:
(1) Merb in Action by Manning Publications
http://www.manning.com/ivey
(2) The Merb Way by Addison-Wesley
http://blog.obiefernandez.com/content/2008/08/announcing-the-merb-way-by-foy-savas.html
3 Chekke October 11, 2008 at 2:05 pm
So already the Rails and Ruby community understand that Rails sucks after all the hype and Ruby is slow as hell language.
4 Rails and Ruby community October 12, 2008 at 12:02 pm
No Chekke. Having built the best framework (Rails) ever in the best language ever (Ruby) we had to ask ourselves where do we go from here ?. Naturally we said ‘lets build something even better’. With no peers we unfortunately have we have to resort to competing with outselves.
5 Dev Blog AF83 » Blog Archive » Veille technologique : Merb, Vim, geode, iPhone, etc. October 13, 2008 at 1:52 pm
[...] une nouvelle série de liens en vrac : * http://blogs.gartner.com/mark_driver/2008/10/10/merb-10-for-ruby-emerges-real-competition-for-rails/ : une étude de Gartner confirme que la compétition entre les frameworks Ruby on Rails et Merb va [...]
6 Zandao October 14, 2008 at 4:55 pm
No “Rails and Ruby community”. You have seen Seaside and realized you have to do better if you want to reach real critical mass.
7 Dean Cruse October 14, 2008 at 6:40 pm
We’re fans, Mark. It’s definitely a small community right now (think Rails 2005), but an awesome group of developers and a great addition to the Ruby eco-system. Our customers tell us they expect Merb to be both an alternative and complement to Rails on many projects.
In terms of the framework needing tools, we (of course) agree
, and have released our TuneUp (http://tuneup.fiveruns.com) open source app profiling tool for Merb into the community.
8 Mark Driver October 16, 2008 at 10:35 am
All,
I’ve purposely refrained from auditing most of the comments on my blog. I’ll take your comments good or bad but not ugly.
I’ve deleted the first one today. Suffice to say that comments like…
“You are very ignorant. Get your facts straight.” from rabid Rails fanboys arent going to make the comments section of this blog.
If you think my facts are not accurate then say so but also say why.
The alternative is not clever. Its just lazy.
m
9 Jeremy Nicoll October 20, 2008 at 6:41 pm
I love Merb! It’s not much to look at yet, but I’m already running one client’s site on it: http://myinboxnews.com. I admit that I’m a bit of a control freak, and Rails was always getting in my way. Now that Merb is near the 1.0 mark (it’s a pre-release atm, mind you), I’ve yet had to hack the innermost workings of Merb just to do what I want. I can’t say the same thing for Rails. Merb is helpful, but stays out of my way when I need it.
10 Markus Jais October 22, 2008 at 3:34 am
A 3rd Merb book can be found at amazon (to be published 2/2009):
Beginning Merb: Replacing Rails with Merb Framework, Datamapper and HAML by Brian Bommarito
http://www.amazon.com/Beginning-Merb-Replacing-Framework-Datamapper/dp/1430218231/ref=sr_1_2?ie=UTF8&s=books&qid=1224660777&sr=1-2
11 James Stewart October 22, 2008 at 9:50 am
I’m primarily a rails developer but have been watching merb for some time and am currently working on my first serious (albeit relatively simple) project with it.
As noted above there are quite a few books on their way, but until they hit documentation is scarce and patchy. Getting an app up and running requires quite a bit of time spent between google, the api docs and the merb source, which is likely to put off most potential users. Even having tracked merb casually for a while it’s taking me quite a while longer to get this project up and running than my first rails app (just after Agile Web Development with Ruby on Rails).
However it turns out, it’s about time Rails had serious competition in the ruby framework world. It’s interesting that it arrives around the same time that Rails seems to be getting serious about thread-safety and Ruby 1.9. Hopefully the next few months will see us with multiple good options for significantly faster ruby web apps.
12 Gandhi October 28, 2008 at 9:08 am
Merb : Copyright (c) 2008 Engine Yard Inc.
Rails : Copyright (c) 2004-2008 David Heinemeier Hansson
I wonder where “close knit community of developers who are evolving the core framework” applies more.
Rails has about 1400 contributors so far : http://groups.google.com/group/rubyonrails-core/browse_thread/thread/7554c8c5a25b4765
Comparing Merb with Rails is apples and oranges. Rails is a full stack. Merb is a dependency jungle ( it’s an actionpack clone ). There is nothing called ‘Merb way’.
13 Weby September 9, 2009 at 4:35 pm
Hi,
Guess what, I have more people asking me of RoR than MerB. That is not to say the features of merb at 1.0 itself are not fantastic. Actually to a developer it is awesome cool.
So ,power to Ruby developers Merb or RoR.
Regards
Weby