emacs-devel
[Top][All Lists]
Advanced

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

Re: Android port of Emacs


From: Dmitry Gutov
Subject: Re: Android port of Emacs
Date: Mon, 19 Jun 2023 01:26:53 +0300
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.11.0

On 18/06/2023 08:31, Eli Zaretskii wrote:
Date: Sun, 18 Jun 2023 04:10:35 +0300
Cc: emacs-devel@gnu.org
From: Dmitry Gutov <dmitry@gutov.dev>

We shouldn't be afraid to deprecate and remove unmaintained ports. And
we shouldn't (like Po already said) delay releases over them.

If we had a consensus on that

This is easier said than done.  In practice, we don't like to release
knowing that one of the platforms we care about is broken.  In fact, I
don't think it ever happened, not even with NS (MSDOS is not a
platform we care about, so it is not relevant.)  We might, of course,
release a broken port if we are unaware that it's broken, but that's
not the same thing.

There might be a certain loss of face involved, but it really is better to have an Android port of at least some versions of Emacs than not have it at all, isn't it?

And if it's broken, we could as well announce that the Android release is "not ready" and will probably arrive at some later time [when/if it's fixed].

FWIW, I think the important part in this discussion is you and Po arrive at some more or less amenable conclusion, and whether the port is developed externally or internally is secondary.

It's fairly plausible that Android will become a desktop platform too not too far off in the future. Or desktop computing will migrate even more toward portable devices. We'll be more prepared either way.

I guess the main concern that would remain is whether the Android
port complicates the common code enough for it to be noticeable.

I invite you to review the results of

   git diff ...origin/feature/android

and make up your own mind about that.

I took a brief look, but couldn't come to a particular conclusion.

It's a large diff, and not all of it is "common" code. TBF the autotools/make additions look the scariest to me, but I'm not sure whether that's an objective opinion.

As far as the number/volume of Java files go, I'd recommend not paying much attention. Java *is* a verbose language, and at any point in the future we'll almost certainly find someone familiar with it (I have 2 years with it under my belt myself, even if it was a decade ago).



reply via email to

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