Apologies if this comes through as a repeat. What is so mysterious/non-responsive about telling you (read the initial email) that I was re-writing an existing file, nothing happened, no errors were reported, and that repeating it with an explicit truncation of the file after opening gave the expected behavior? What I find ironic about your objections is that *my* problem is long-since fixed. I was asking a question the answer to which might have shed some light on how to construct an example that would be useful in analyzing what happened. In short, I was trying to help you and others. Sorry, I don't see how giving me "talking Elisa" in response to that is in any way helpful or justified. Examples are good, but the system has to be trustworthy to run in unknown situations. Robust error reporting is critical to that goal. Sadly Squeak has a long tradition of silent failure and a culture that thinks it is perfectly ok. I find it dangerous. Bill ________________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of csrabak@bol.com.br [csrabak@bol.com.br] Sent: Sunday, November 14, 2010 10:44 PM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] Unix thing? Possible silent failure writing file. Wilhelm, This comment, which we know is coming from a SW developer, looks like an automatic "Elisa" like IA bot replying syntactically correct English but not adding exchange in the dialogue. Your description of the unexpected (for you, that is) behavior doesn't allow us to understand minimally if the 'failure' is: a) a mismatch between your expectation and actual, albeit correct behavior; b) an intermittent occurrence or a systematic one, i.e., reproducible one; c) _what_ exactly was you doing in your code and in which context. Also, since you mentioned Windows to Unix transition, is the 'silent' failure in Unix 'not silent' if you run the same code? I hope that by now you can understand why we need a minimum piece of code in order to envision the tackling of an "ultimately fix". Regards, -- Cesar Rabak Em 14/11/2010 19:26, Schwab,Wilhelm K < bschwab@anest.ufl.edu > escreveu: If that's how you want it. My concern is not over the failure; it's over its silence. I fear that saying nothing only reduces our collective opportunity to ultimately fix it. ________________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of Stéphane Ducasse [stephane.ducasse@inria.fr] Sent: Sunday, November 14, 2010 4:21 PM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] Unix thing? Possible silent failure writing file. send a code snippet else it does not help to send the mail. On Nov 14, 2010, at 10:19 PM, Schwab,Wilhelm K wrote:

Surely I'm not the only person to make the Windows to Unix move. It is either something that happens, or another case of the Squeak primitives failing without saying that they did, let alone why. What you call a crystal ball, I call words from the wise who might have seen something similar before.

________________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of csrabak@bol.com.br [csrabak@bol.com.br] Sent: Sunday, November 14, 2010 3:43 PM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] Unix thing? Possible silent failure writing file.

Without a minimum snippet of the code that purportedly got into this incident, only a crystal ball can help here. . .

You have the impression that I am asking you to do my work. Not the case; I asked you whether a sequence of events was familiar - that's it. The involvement of R and "adding a column" would mean a lot to those who use it. I am not using any of the variations you list below, but I will look at the code behind them; it might offer some clue about any preparation that the newly opened file might need. ________________________________________ From: pharo-project-bounces@lists.gforge.inria.fr [pharo-project-bounces@lists.gforge.inria.fr] On Behalf Of Sven Van Caekenberghe [sven@beta9.be] Sent: Monday, November 15, 2010 8:06 AM To: Pharo-project@lists.gforge.inria.fr Subject: Re: [Pharo-project] Unix thing? Possible silent failure writing file. Bill, On 15 Nov 2010, at 13:35, Schwab,Wilhelm K wrote:

Apologies if this comes through as a repeat.

What is so mysterious/non-responsive about telling you (read the initial email) that I was re-writing an existing file, nothing happened, no errors were reported, and that repeating it with an explicit truncation of the file after opening gave the expected behavior?

What others on this list are saying is that you have to provide _Concrete Code Examples_ independent from all decoration (it doesn't matter that you are using R and doing whatever with it). Making a good, short example that others can run will help you and others. You have to make at least the isolation effort first. Even better, you should look as deep into it as you can. Otherwise you will be ignored because you are asking others to do your work. So, you should have written something like (but maybe this is not what you meant or were using/doing): Shouldn't FileStream>>#fileNamed:do: raise an error when overriding ? For example: FileStream fileNamed: '/tmp/foo.txt' do: [ :stream | stream nextPutAll: 'First'; crlf ]. "works" FileStream fileNamed: '/tmp/foo.txt' do: [ :stream | stream nextPutAll: 'Second'; crlf ]. "overwrites - but should fail IMHO" FileStream oldFileNamed: '/tmp/foo.txt' do: [ :stream | stream nextPutAll: 'Third'; crlf ]. "overwrites" FileStream newFileNamed: '/tmp/foo.txt' do: [ :stream | stream nextPutAll: 'Fourth'; crlf ]. "raises error" HTH, Sven PS: And I also think you should make an Google Code Issue when you are sure there is a bug ;-)