Invalid byte sequence in US-ASCII

[cpater]

Invalid byte sequence in US-ASCII

[jamesshore]

That's surprising. What happens if you load rabu-sched.rb in a text editor, then save it out again? (And do you see anything odd about that line when you load it in a text editor?) My best guess is that my text editor somehow inserted a non-ASCII code into the file. I switched to a different editor between 0.1.4 and 0.2.

[jamesshore]

I think I found the problem. There's an ë on line 8727, which is part of the es5-shim-1.2.0.js library. That library is new in v0.2. Rabu-sched.rb is showing up as a UTF-8 file on my Mac, but it looks like your computer is interpreting it as US-ASCII.

I'm not sure how to fix it, but at least we know where the problem lies.

(To isolate the problem, I loaded the file in a text editor, resaved it using the ISO-Latin 1 encoding, then did git diff from the command line.)

[cpater]

Hi,
I compared the file from the releasbundle with a copy that I had converted
to utf-8.
The difference was in Line: 8726 : // -- fschaefer Florian Schfer
Copyright (C) 2010 MIT License

After I replaced the with ae (which is a correct replacement for in
german) it works.

2011/5/16 jamesshore <
[email protected]>

I think I found the problem. There's an on line 8727, which is part of
the es5-shim-1.2.0.js library. That library is new in v0.2. Rabu-sched.rb is
showing up as a UTF-8 file on my Mac, but it looks like your computer is
interpreting it as US-ASCII.

I'm not sure how to fix it, but at least we know where the problem lies.

(To find the problem, I loaded the file in a text editor, resaved it using
the ISO-Latin 1 encoding, then did git diff from the command line.)

Reply to this email directly or view it on GitHub:
#7 (comment)

[jamesshore]

Which version of Ruby are you using?

[jamesshore]

After looking into this further, it appears to be caused by an incompatibility between Ruby 1.9 and Ruby 1.8. I've checked in a new version of rabu-sched.rb that may fix the problem. Can you please try it? The updated file is here: https://github.com/teamrabu/rabu-schedule/raw/9a6cfe90685c5156123aefa00eb72304a623c78c/released/rabu-sched.rb

[jamesshore]

If the above fix doesn't do the trick, a proper fix will have to wait for the Java CLI rewrite.

[cpater]

Hi,
my ruby version is: ruby 1.9.2p180 (2011-02-18) [i386-mingw32]

the provided fix produces the following error:

rabu-sched.rb:34:in gsub': incompatible character encodings: UTF-8 and CP850 (E ncoding::CompatibilityError) from rabu-sched.rb:34:inhtml'
from rabu-sched.rb:47:in `

'

The file attatched to this mail works for me.

2011/5/19 jamesshore <
[email protected]>

If the above fix doesn't do the trick, a proper fix will have to wait for
the Java CLI rewrite.

Reply to this email directly or view it on GitHub:
#7 (comment)

[jamesshore]

It looks like GitHub stripped the attachment.

[jamesshore]

Hi Christoph,

I got your email with the attached file--thank you. I'm going to solve the problem by rewriting the CLI in Java.

Warning: Ruby rant follows. There's no need to read this...

The file you sent addresses the problem's symptoms by eliminating non-ASCII characters, but it doesn't address the root cause of the problem, which is that Ruby changed its approach to character encodings Ruby 1.8 and Ruby 1.9. This change introduced a gaping backwards compatibility rift between the two versions. It may be possible to write code that runs on both versions, but if it is, my Google-fu is too weak to find the solution. Apparently, most people are using Ruby for Rails and don't have to worry about any Ruby environment other than their own.

Although I could use your fix to solve the immediate problem, it would recur next time we add a file with non-ASCII data. It's easy for a third-party library to do that without anyone noticing--that's what happened this time--so it would only be a matter of time before this problem cropped up again. That would make me a sad panda.

This experience (along with the build difficulties people have been having) has convinced me that Ruby is a piss-^H^H^H^H poor choice for distributing software. Although I'm sure we could work around Ruby's cross-version and cross-platform incompatibilities, it's not worth it. I was considering rewriting the CLI in Java anyway--now I'm convinced that it's the right thing to do. Java has a larger installed base, JARs are a better distribution mechanism than gems, Java's charset support is robust and well-established, and Java has better backwards compatibility. The CLI is only seven lines of code, so rewriting it will be easy.

#end rant.

[jamesshore]

I think I've fixed this. Please download the new Rabu JAR file from this URL: https://github.com/teamrabu/rabu-schedule/blob/master/released/rabu.jar?raw=true

Run it from the command-line using java -jar rabu.jar < estimates.rabu > projection.html and let me know if it works!

Note: The estimates.rabu file must be saved in UTF-8 format.

[cpater]

Hi,
yes it is fixed. Also the umlaut in the estimates.rabu files are coorect in
the projection.html.

Thanks

2011/5/30 jamesshore <
[email protected]>

I think I've fixed this. Please download the new Rabu JAR file from this
URL:
https://github.com/teamrabu/rabu-schedule/blob/master/released/rabu.jar?raw=true

Run it from the command-line using java -jar rabu.jar < estimates.rabu > projection.html and let me know if it works!

Note: The estimates.rabu file must be saved in UTF-8 format.

Reply to this email directly or view it on GitHub:
#7 (comment)

[jamesshore]

Fantastic--thanks for checking.