Forum: Ruby-core [ruby-trunk - Bug #7088][Open] Documentation of Timeout::timeout is wrong about which thread execute

Posted by Robert Klemme (robert_k78)
on 2012-09-29 18:01
(Received via mailing list)
Issue #7088 has been reported by rklemme (Robert Klemme).

----------------------------------------
Bug #7088: Documentation of Timeout::timeout is wrong about which thread 
executes the block
https://bugs.ruby-lang.org/issues/7088

Author: rklemme (Robert Klemme)
Status: Open
Priority: Normal
Assignee:
Category:
Target version:
ruby -v: ruby 1.9.3p194 (2012-04-20) [i686-linux]


See http://rubydoc.info/stdlib/timeout/Timeout#timeout... - it 
reads "The block will be executed on another thread and will be given 
one argument: sec."  If you make the source code visible or look into 
the file timeout.rb in the library you'll see that the block is invoked 
from the current (calling) thread.  The comment reads as if the author 
had wanted to do an implementation like this:

def timeout_1(timeout, &code)
  raise ArgumentError, "Invalid timeout: %p" % [timeout] unless timeout 
> 0
  raise ArgumentError, "No code to execute" if code.nil?

  worker = Thread.new(&code)
  worker.join(timeout) and worker.value
end

Note: this avoids bug #4285 at the expense of a potentially longer 
running thread.  Killing the background thread would suffer the same 
issues.
Posted by drbrain (Eric Hodel) (Guest)
on 2012-09-29 18:33
(Received via mailing list)
Issue #7088 has been updated by drbrain (Eric Hodel).

Category set to DOC
Target version set to 2.0.0


----------------------------------------
Bug #7088: Documentation of Timeout::timeout is wrong about which thread 
executes the block
https://bugs.ruby-lang.org/issues/7088#change-29795

Author: rklemme (Robert Klemme)
Status: Open
Priority: Normal
Assignee:
Category: DOC
Target version: 2.0.0
ruby -v: ruby 1.9.3p194 (2012-04-20) [i686-linux]


See http://rubydoc.info/stdlib/timeout/Timeout#timeout... - it 
reads "The block will be executed on another thread and will be given 
one argument: sec."  If you make the source code visible or look into 
the file timeout.rb in the library you'll see that the block is invoked 
from the current (calling) thread.  The comment reads as if the author 
had wanted to do an implementation like this:

def timeout_1(timeout, &code)
  raise ArgumentError, "Invalid timeout: %p" % [timeout] unless timeout 
> 0
  raise ArgumentError, "No code to execute" if code.nil?

  worker = Thread.new(&code)
  worker.join(timeout) and worker.value
end

Note: this avoids bug #4285 at the expense of a potentially longer 
running thread.  Killing the background thread would suffer the same 
issues.
Posted by zzak (Zachary Scott) (Guest)
on 2012-10-06 21:41
(Received via mailing list)
Issue #7088 has been updated by zzak (Zachary Scott).

Assignee set to zzak (Zachary Scott)


----------------------------------------
Bug #7088: Documentation of Timeout::timeout is wrong about which thread 
executes the block
https://bugs.ruby-lang.org/issues/7088#change-30127

Author: rklemme (Robert Klemme)
Status: Open
Priority: Normal
Assignee: zzak (Zachary Scott)
Category: DOC
Target version: 2.0.0
ruby -v: ruby 1.9.3p194 (2012-04-20) [i686-linux]


See http://rubydoc.info/stdlib/timeout/Timeout#timeout... - it 
reads "The block will be executed on another thread and will be given 
one argument: sec."  If you make the source code visible or look into 
the file timeout.rb in the library you'll see that the block is invoked 
from the current (calling) thread.  The comment reads as if the author 
had wanted to do an implementation like this:

def timeout_1(timeout, &code)
  raise ArgumentError, "Invalid timeout: %p" % [timeout] unless timeout 
> 0
  raise ArgumentError, "No code to execute" if code.nil?

  worker = Thread.new(&code)
  worker.join(timeout) and worker.value
end

Note: this avoids bug #4285 at the expense of a potentially longer 
running thread.  Killing the background thread would suffer the same 
issues.
Posted by zzak (Zachary Scott) (Guest)
on 2012-10-11 20:29
(Received via mailing list)
Issue #7088 has been updated by zzak (Zachary Scott).


How about we change that to: "The block will be executed on the current 
thread and will be given one argument: sec. While, the timer for sec 
number of seconds will run in another thread".
----------------------------------------
Bug #7088: Documentation of Timeout::timeout is wrong about which thread 
executes the block
https://bugs.ruby-lang.org/issues/7088#change-30327

Author: rklemme (Robert Klemme)
Status: Open
Priority: Normal
Assignee: zzak (Zachary Scott)
Category: DOC
Target version: 2.0.0
ruby -v: ruby 1.9.3p194 (2012-04-20) [i686-linux]


See http://rubydoc.info/stdlib/timeout/Timeout#timeout... - it 
reads "The block will be executed on another thread and will be given 
one argument: sec."  If you make the source code visible or look into 
the file timeout.rb in the library you'll see that the block is invoked 
from the current (calling) thread.  The comment reads as if the author 
had wanted to do an implementation like this:

def timeout_1(timeout, &code)
  raise ArgumentError, "Invalid timeout: %p" % [timeout] unless timeout 
> 0
  raise ArgumentError, "No code to execute" if code.nil?

  worker = Thread.new(&code)
  worker.join(timeout) and worker.value
end

Note: this avoids bug #4285 at the expense of a potentially longer 
running thread.  Killing the background thread would suffer the same 
issues.
Posted by drbrain (Eric Hodel) (Guest)
on 2012-10-11 22:06
(Received via mailing list)
Issue #7088 has been updated by drbrain (Eric Hodel).


I think deleting the paragraph would be sufficient.  We don't need to 
reveal the implementation details of timeout.
----------------------------------------
Bug #7088: Documentation of Timeout::timeout is wrong about which thread 
executes the block
https://bugs.ruby-lang.org/issues/7088#change-30332

Author: rklemme (Robert Klemme)
Status: Open
Priority: Normal
Assignee: zzak (Zachary Scott)
Category: DOC
Target version: 2.0.0
ruby -v: ruby 1.9.3p194 (2012-04-20) [i686-linux]


See http://rubydoc.info/stdlib/timeout/Timeout#timeout... - it 
reads "The block will be executed on another thread and will be given 
one argument: sec."  If you make the source code visible or look into 
the file timeout.rb in the library you'll see that the block is invoked 
from the current (calling) thread.  The comment reads as if the author 
had wanted to do an implementation like this:

def timeout_1(timeout, &code)
  raise ArgumentError, "Invalid timeout: %p" % [timeout] unless timeout 
> 0
  raise ArgumentError, "No code to execute" if code.nil?

  worker = Thread.new(&code)
  worker.join(timeout) and worker.value
end

Note: this avoids bug #4285 at the expense of a potentially longer 
running thread.  Killing the background thread would suffer the same 
issues.
Please log in before posting. Registration is free and takes only a minute.
Existing account (Switch to SSL-encrypted connection)
NEW: Do you have a Google/GoogleMail or Yahoo account? No registration required!
Log in with Google account | Log in with Yahoo account
No account? Register here.