Re: Problem "Stable" is not stable

[Christof Thalhofer <chrisml@xxxxxxxxxxx>]

Am 28.08.24 um 12:29 schrieb Bruce Steers:

> I understand the premise but I do not think it would work because not
>  enough people would be testing.

I have to test each new minor version before I can install it in my
company. I work with it and find and report the bugs. Usually there are
so many new bugs that I have to wait about three or four revisions until
my software works as expected and I can start to deploy the new minor
version of Gambas.

> Most people are loathed to use anything other than stable branch.

And exactly these people are now being disappointed.

> I think a testing branch may not get enough usage to find the bugs.
>
> Most likely testing will get considered stable, added to stable, only
>  then those unforseen bugs get reported, then fixed. Kinda just as it
> is now.

There are far too few days at the moment. Benoit says “please test, I
will release soon” and after five days he releases the new "stable"
version which is in reality full of new bugs. This is definitely not
enough time to find them.

> I think if Gambas had many more users/testers a testing branch would
>  work well.
>
> I could be wrong though, maybe a pre-stable branch would get enough 
> testers to make it viable.

In Debian "testing" is the dev version (which is an equivalent to the
"master" branch in Gambas) so this name would be not the best.

If we could call it "pre-stable" that would be much better. Or let's
call it "new"? So we can reach those who want to have the latest thing.

Alles Gute

Christof Thalhofer

--
Dies ist keine Signatur