Re: [ jEdit-devel ] Fwd: Re: jEdit 4.4pre1 is out!

[Vampire <Vampire@jEdit.org>]

Kazutoshi Satoda schrieb:
> Vampire wrote:
>> Kazutoshi Satoda schrieb:
>>>  If subscription is really required, I beg you to forward the
>>>  post.
>> Ok
>
> Thank you.
>
>>>  I thought "devel" means jedit-devel folder on SF.net files, which
>>>  includes pre-releases. It seems correct to mark a bug which happens in
>>>  4.4.x but doesn't happen in 4.3.x as "regressive".
>> But then almost EVERY bug is regressive, as there is always a previous
>> version. My understanding of regressive was "new in trunk since last
>> release", but to be honest I didn't do anything with regressive yet, so
>> this may be my misunderstanding.
>
> Sorry, let me rephrase more specifically:
> It seems correct to mark a bug about a behavior which has worked in an
> old version but is broken in a new version. In other words, bugs about
> new features are not regressive.
Ah, ok, now I see more clearly the meaning of regressive.
>>>  I know there might be some regressions on 4.3.x or trunk. But I think
>>>  it's not so problematic since the reports should include a clear 
>>> version
>>>  information in it to be marked as regressive.
>> Sure, but if someone opens that page and sees a long list, he will not
>> open all entries to see for what version it applies but maybe don't
>> install / update jEdit. This is also especially true for the severe list
>> which contains quite some entries which might alienate potential users.
>
> Filtering such users, who wouldn't adopt the known problems, out of
> testing is a part of intention of my post.

I'm concerned that you also filter users from using the stable releases, 
not only the pre releases.

> The right way not to alienate them is reducing the list.

Of course this would be better.

> I believe it's possible, don't you?

It would if people would concentrating on fixing bugs. But as you can 
see in that list, there are severe bugs reported 2005. So obviously it 
is not so easily possible. :-(