Ruby 1.9.1 Preview 1 is out

The path to a stable Ruby 1.9 has been long. Almost a year ago, 1.9.0 was released. This turned out to have been less than stable. The core library has grown by many hundreds of methods since then. And, probably more significantly, people have started using 1.9 for real, and they've discovered rough edges that needed to be fixed. In particular, the multinationalization support, which is incredibly ambitious, turned out to be hard to use for library writers. James Edward Gray II was probably the first to bump into this as he updated his FasterCSV library (now part of core Ruby) to handle any encoding supported by Ruby. His problems lead to a lot of discussion, and eventually to an entirely new concept inside the interpreter. And as a result, the core team have decided to slip the release of 1.9.1 by at least a month while they investigate other encoding-related issues lurking in the libraries.

In the meantime, a 1.9.1 preview has been released. Details are here.

If you are the maintainer for any publicly available Ruby code (be it a Gem, an application, or whatever) I strongly urge you to download this preview. You'll be doing the community a great service in two ways. First, the various incompatibilities between 1.9 and 1.8 mean that there's a chance that your code may not work without some tweaks. Making those changes now will help others using your code. As importantly, by using 1.9.1 for real, with real code, you'll potentially discover other rough edges. Reporting these back through ruby-core or RedMine will help the Ruby developers further hone the interpreter.

Now, all this delay leaves me with a problem. The new 1.9-specific PickAxe is now content complete. Some people have been waiting for it for 10 months. And I have to decide: should I hold on for the official 1.9.1 release (which will be, at the earliest, at the end of January 2009) or should I send it to the printers? Let me know what you think.

Comments

[Paul Doerwald]

Definitely wait. I want an authoritative, reliable reference that will last a long, long time.

[Marcus Brito]

I'd rather wait until everything is stable. But then, I bought the PDF version, and have been following all updates for a while now.

People who bought only the dead tree version and haven't seen anything yet might be a little more antsy.

[ryan]

+1 for waiting.

[jonny_noog]

+ 1 for waiting.

Speaking as a relatively new Ruby user, I would greatly appreciate a book that is as accurate and definitive as possible. If this means waiting a little longer, then that is no problem at all for me.

[Hans Torm]

Wait. I have enough books with unreliable information be course they where released before the actual product.

"Don't Assume It - Prove It"

- which is hard to do when it ain't final... :p

[Arthur Axel]

I personally would rather wait for the book to be updated. But this is really what happens with dead tree programming books all the time anyway, which is why now that I know ruby from the first book, I doubt I'll buy another one.

[Alan Gardner]

I agree. Waiting is the best option.

I also think it is a good sign that everyone agrees on this, as it shows a bias towards quality within the community!

[Mike]

Please wait.

[btoone]

Wait to print. PDF is great in the meantime

[SansS]

Quality is not date driven. And you have a good bridge in the pdf solution.

Package the 2.0 paper and the 3.0 pdf for people who need paper and are anxious.

[Martin]

Please wait. I like paper documentation to be correct a lot more than I want it ASAP...

[Dolaur Liberté Crozon Cazin]

Hello,
I think you should wait. If anything needs updates, and quick fixes, the PDF comes handy to support that. We've wait one year already, and January 2009 is not that far anyway. If anything have to go live on paper, it should be as authoritative and "Official" as possible.
So.. Wait.

[Dolaur Liberté Crozon Cazin]

Hello,
I think you should wait. If anything needs updates, and quick fixes, the PDF comes handy to support that. We've wait one year already, and January 2009 is not that far anyway. If anything have to go live on paper, it should be as authoritative and "Official" as possible.
So.. Wait.

[Daniel Waite]

Wait until Ruby 1.9.1 is released for production use. Even if the launch date for 1.9.1 slips another month, or another 6, wait. Please. :)

[anjan bacchu]

hi dave,

as a reader of many of your group's books, I think that you should wait UNLESS you're ready to update the PDF book as soon as 1.9.1 is published and publish an errata for the print book.

BR,
~A

[George Lippert]

Yes, wait. The community knows that 1.9.1 is the first real 1.9 release. Being able to state in your book that, e.g., all examples were tested against 1.9.1 is *much* better than referencing 1.9.0, which would mark the book as instantly obsolete in my eyes.

[Frank]

My wife likes the printed book!
I did not have the Opportunity to check it out, but i will try!
Thanks for keeping this Great Social networking blog!
Cool man!