[Vm-dev] Questions, communication and process

[stephane ducasse]

Ok thanks
I forwarded that to pharo and we will do a web summary for people.

I will start to read vm code and the material related such as the newblue book chapter too.

Stef



On Apr 9, 2010, at 12:35 PM, Ian Piumarta wrote:

> Hi Stef,
> 
> On Apr 9, 2010, at 2:23 AM, stephane ducasse wrote:
> 
>> 	- How do people report problems? Just sending an email in the mailing-list is enough?
> 
> Here are the options, necessarily biased towards my preference which is to try to avoid having to poll for outstanding issues.
> 
> If it's a general problem of installation or bizarre behaviour that differs between platforms, ask for help on the Squeak VM mailing list.
> 
> http://lists.squeakfoundation.org/mailman/listinfo/vm-dev
> 
> Pharo and Squeak users are EQUALLY WELCOME to report VM issues on that list.  Squeak has several people who regularly take the time to forward things from squeak-dev to vm-dev when appropriate.  You might want to encourage the same thing on the Pharo list.  I know for a fact that there are several vm-dev subscribers that monitor the Pharo list.  (You might want to make it clear that Pharo uses the Squeak VM too.  This does not necessarily appear to be universally understood.)
> 
> If it's a Unix-specific problem that you think might be a bug, report it to the above mailing list AND send me email.  Fixes are best submitted as complete modified files (not 'diffs').  (There might well be a way to tell ediff to use diff output directly, but that does not guarantee that your 'original' and my 'original' are in any way related.  Sending whole files is MUCH more robust.)
> 
> If it's Windows-specific, try the above list and Andreas Raab.  For Mac-specific, the above list and John McIntosh.  If it's common code shared by all platforms, especially code generated by VMMaker, send to Dave Lewis and the above list.
> 
> VM bugs can fed to mantis here: bugs.squeak.org
> 
> The people mentioned above probably look at mantis on a regular basis.  Anything with more than four legs gives me the heebie-jeebies, which isn't why I don't regularly trawl mantis but it'll do.
> 
>> 	- How a fix in one OS is propagated to the others?
> 
> Fixes that affect more than one OS are very often in the common code that is generated automatically and shared between all the platforms.  There is no explicit manual propagation required.  (That being said, the people mentioned above are often in contact on a daily basis.  When co-ordination is required, it will happen.)
> 
> The exception is that some code is shared between Unix and Mac.  John is good at spotting commits from me that affect him, and is good at bugging me when necessary into fixing anything Unixy that might affect him.
> 
>> 	- How the user can know that?
> 
> Watch vm-dev for discussions or commit reports that say the issue was fixed.  If it is in mantis, it might even get closed too.
> 
> The above reflects how I like to work, and I can only speak for myself.  MPG from the others mentioned above might not be comparable.
> 
> Cheers,
> Ian
> 
> (Feel free to forward, re-post, etc., the above as you see fit.)
>