Hannes,

Could I see a little more of the stack to get some context?

I should point out that I don't necessarily keep the mcz files up-to-date and that the complete install of FileTree should involve a load from a git repository[1], to get the absolutely latest code.

I'd also be interested to know which version of the MonticelloFileTree-Core.squeak43 package that is loaded ... 

Finally I run tests using travisCI[2] for each push to github, but I test against the image downloaded from [3] (including latest github code for Squeak4.3 branch) and that passing all of the unit tests, with my latest updates...

Dale

[1] https://github.com/dalehenrich/filetree/blob/pharo1.3/doc/SqueakInstall.md
[2] http://travis-ci.org/#!/dalehenrich/filetree/builds/1810028
[3] http://ftp.squeak.org/4.3/Squeak4.3.zip
----- Original Message -----
| From: "H. Hirzel" <hannes.hirzel at gmail.com>
| To: "The general-purpose Squeak developers list" <squeak-dev at lists.squeakfoundation.org>
| Sent: Thursday, July 12, 2012 3:16:13 PM
| Subject: Re: [squeak-dev] Installing Filetree
| 
| Hi Bert
| 
| Installer ss3
|         project: 'FileTree';
|         install: 'MonticelloFileTree-Core.squeak43'.
| 
| worked fine in a fully updated image.
| 
| I then did the following test.
| I created a package 'HHTestingFileTree' in the Monticello Browser and
| added a filetree repo on the local harddisk
| 
| I created a class 'Dummy' in the category 'HHTestingFileTree'
| 
| Then I wanted to save it in the Monticello Browser with 'save it'.
| 
| A debugger came up breaking in MCScanner >> next
| 
| next
| 	| c |
| 	stream skipSeparators.
| 	c := stream peek.
| 	c = $# ifTrue: [c := stream next; peek].
| 	c = $' ifTrue: [^ self nextString].
| 	c = $( ifTrue: [^ self nextArray].
| 	c isAlphaNumeric ifTrue: [^ self nextSymbol].       <<<<<< c is nil
| 	self error: 'Unknown token type'.
| 
| 
| Am I missing something? Does it work for you
| 
| --Hannes
| 
| On 7/5/12, Bert Freudenberg <bert at freudenbergs.de> wrote:
| > On 2012-07-05, at 16:25, Levente Uzonyi wrote:
| >
| >> On Wed, 4 Jul 2012, Bert Freudenberg wrote:
| >>
| >>>  with the fixes I posted today:
| >>
| >> As a side-effect several (~30) Monticello tests fail or raise an
| >> error.
| >
| >
| > Ah. I fixed that problem now. The test was using an unexpectedly
| > short name
| > for a MC version ("MC1").
| >
| > It's back down to 4 failures, no errors.
| >
| > - Bert -
| >
| >
| >
| >
| 
|