<sigh>

Chris,

I spent a good three hours yesterday testing Magma and logging, in 
detail, my experiences doing so.  As it happened, I came across several 
issues that I felt needed to be brought to your attention.  I was 
confident that by doing so I would be helping you to find the bugs, fix 
them, and thus improve Magma.  After a few iterations of this, I 
thought, Magma might be suitable for production use, and we would all 
benefit.

It's very clear from your responses below, however, that this isn't 
something that interests you.  Magma "works for you", and any problems 
any of the rest of us have with it are due to "incorrect expectations", 
bugs in Squeak, user error when setting preferences, or, evidently, are 
simply figments of our imagination.

Ok.  I've clearly been wasting my time.  If it's more important to you 
to defend Magma's reputation by denying that bugs exist, than it is to 
defend Magma's quality by fixing them, then there's no way that I can 
help, and absolutely no chance that I will ever use it.

But in case it helps:
- Debug mode is *not* on in my images.  I just checked.  I still 
experienced locked images while trying to connect, and no commit 
errors.  Perhaps the logic that checks for debug mode is faulty?
- Saving and quitting a running server *did* leave my image in a state 
where it couldn't act as a server.  I'm happy to send you the image.
- Killing the process of a running server *did* leave my database in a 
state where I couldn't access the data.  I'm happy to send you the 
.magma file.
- I'm not making up what I'm calling "the silent failure issue".  It 
may be very specific to my setup: perhaps it only affects those using 
Squeak 3.7b on a Carbon VM running on aluminum Powerbooks in rainy 
climates.  But simply denying that it exists doesn't help, and is 
unlikely to increase anyone's confidence in Magma.

One other note:

>> How easy is it to lose data?
> Actually, not too easy, as long as hardware doesn't fail.

Having a process die suddenly does not require hardware failure.  
Operating systems kill processes all the time.  Not to mention 
sysadmins.  If I can't "killall -9 squeak" without fearing for my data, 
that's a big problem.  Again, I would be much happier if you seemed to 
be acknowledging the problem rather than explaining it away.

Cheers,
Avi

Chris Muller wrote:

> All the tests that you performed either worked for me or your 
> expectations differed from how it works.

> My main purpose in reporting this is to clarify that the current 
> release,
> "1.0gamma7" truly is "gamma" quality IMO, not "alpha" or "beta" 
> quality as has
> been characterized, at least somewhat, by the poor experience you had.

> My guess is, you will not be able to produce a script
> that locks the image, corrupts data, produces inconsistent results, or 
> anything
> else really bad like that unless you go outside the bounds of what 
> Magma
> "supports".

> ... there may be a platform-specific bug in Squeak somewhere.

>> Lesson learned: do not save and quit a running magma server.
> No, this works too.

> There is no no silent-failure business.

> A ha!  You ARE in debug mode...

> The defaults for both of these are false, so I'm not sure what 
> prompted you
> to change them to true.