screen-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [screen-devel] screen maintainer


From: Amadeusz Sławiński
Subject: Re: [screen-devel] screen maintainer
Date: Mon, 7 Apr 2014 20:30:14 +0200

On Mon, 7 Apr 2014 19:32:10 +0200
Juergen Weigert <address@hidden> wrote:

> Ahoj Amadeusz!

Hello

> Welcome aboard. Do you have an account at savannah? If so, I am happy
> to grant you admin privileges on screen. 

Just made one, also send request for inclusion so I'm easier to locate.

> Others: I trust you figure it out how to prevent Amadeusz from
> removing all the features that you love. He appears to be willing to
> discuss things. I believe someone who is 'brave enough' to do
> releases is very important. Just don't try to make a perfect release.

I certainly do not plan removing used features, but as some mails shown
some of them have better replacements and are probably not up to modern
technology. I think it's important that this discussion has taken place.
It certainly shows that there are people using screen which I would
like to think makes it even more important to be further enhanced.

> Maintaining a stable 4.x branch while collecting for a 5.0 is nice to
> have. But it is a maintenance overhead.  Maybe it is needed these
> days. When I did screen releases, a dead simple 'release early -
> release often' strategy was just fine.

I think it is important to have branch for trivial bug fixes, even if
some say that 4 is tested well enough.

> I'd actually suggest, Amadeusz does a 4.1.90 beta release directly
> from his branch as is, to show the world, that he is serious about
> doing releases, and to invite the critics. (From *not looking* at the
> code for years, I am unable to tell, which changes would hurt.)
> Having your version released as beta - just for the reviews - might
> help to learn what is important.
> 
> I'd directly bump the version to 4.2 -- several 4.1.0 version are
> floating around. Try google them.

This certainly sounds like a good idea.

> I would not remove the 'ancient code #ifdefs' without being prepared
> to move them back in, as soon as complaints come up. The old dynix,
> sunos or hpux systems that we used 20 years ago ago as our main
> development platforms may be still around somewhere. You never
> know :-)

Not sure how much use some of those platforms will have for some newer
features ;) But I say let's first see if people scream in some
reasonable quantity. Another alternative could be saying that it's not
supported from version 5, which could give another reason for
maintaining v.4 branch.

> I love the nethack feature. It reminds me of playing nethack back at
> the good old days. Well, so much for sentiments. Get over it. Kill
> that feature. Keep the nethack option as a documented noop for
> compatibility in the config file for a while, but who would
> understand the messages these days? It hurts a little, as it degrades
> readability somewhat. A proper implementation would first add locale
> support, then add a nethack-locale.

Who didn't play it at least once? ;)
Yes, I was thinking something similar with adding translations, but I
haven't looked if it's possible to add 'fake' locales.

> Speaking of readability: Having the curly braces opening at the
> far end of an if statement is K&R style for sure. Personally I'd say
> the GNU style, where one can rely on the opening and closing braces 
> to be in the same column is more readable. But I don't really mind, 
> I can read both.
> Whover is editing the files most can ask others to accept his coding
> style. That is a fair request.

I don't have 30 inch monitor, and on smaller ones, one can see more
code when they are on same lines, this one also seems optional ;).
I will certainly look into following the mandatory ones, ie. how
structs and functions should be defined.

> Someone complaind that he did not get access to the repo on savannah. 
> That was not on purpose. It just slipped me. Sorry. 
> 
> I wasn't aware that nobody else would exercise his admin rights.
> This thread did much better, obviously. :-)
> 
>         cheers,
>                 JW-


Amadeusz




reply via email to

[Prev in Thread] Current Thread [Next in Thread]