From: Ilias Lazaridis <ilias@...>
Date: 2011-05-16T23:55:25+09:00
Subject: Re: BARRIER - ruby gem - code converter not found (UTF-16LE to IBM737)

On 16 ÌÜúïò, 17:31, Luis Lavena <luislav...@gmail.com> wrote:
> On May 16, 8:22 am, Ilias Lazaridis <il...@lazaridis.com> wrote:
>
> > After visiting ruby-lang.org, I decided to get the actual 1.9.2
> > version via
>
> > ERROR:  While executing gem ... (Encoding::ConverterNotFoundError)
> >     code converter not found (UTF-16LE to IBM737)
>
> > A workaround:
>
> > $> chcp 850
> > $> gem install <some gem>
> > $> chcp 737
>
> > I've queried for a solutions, but didn't found something concrete,
> > although this defect seems to be know for half a year.
>
> > Is this an issue with ruby, with "gem" or with the rubyinstaller?
>
> The error is not an bug, there is no way to convert IBM737 encoding to
> UTF16 or UTF8, period.

I cannot verify this, but even if your right:

Then it is most possibly an implementation defect, see below.

> You need to update your console to use a modern set of encoding/
> codepage, also, do yourself a favor and use a TrueType font for it,
> which will solve any possible rendering issue you might experience in
> the future.
>
> See this:https://github.com/cucumber/cucumber/wiki/Troubleshooting
>
> See how to permanently set your console. IBM737 and 850 are DOS-age
> encodings.

I have a simple laptop, with Windows 7 on it, greek version.

DOS-age encodings or not:

I had _never_ with any toolkit any problems, and ruby 1.8 worked fine,
too.

And of course I have as simple expectation: that a native windows
installer will install a ruby 1.9 that works.

I don't like to touch the very basic OS setup, but I've tried:

Your "Troubleshooting" document suggests: chcp 1252 / chcp 65001.

But with those I can't write greek chars.

.

--
http://lazaridis.com