|
|
Log in / Subscribe / Register

Waiting for Emacs 22 (and looking forward to Emacs 23)

The much-delayed Emacs 22 release has been covered here a couple of times recently. Since the last article, it would appear that the Emacs process has hit its lowest point, and things should be getting better from here. In the long term, though, the Emacs developers may have to take a hard look at their release management process if they want to keep the project healthy.

The low point was probably sometime around when Richard Stallman got tired of people asking when a release might happen:

I have been insulted and abused many times here lately. I did not respond to most of these insults, but I did take offense.

A number of developers responded that they had no intent to insult or abuse, but that they do have real concerns about how the process works. A couple of examples:

The current feature freeze has now lasted for more than 3 years, during which Emacs _development_ has practically been at a stand-still, so it is no wonder your team of _loyal_ developers is getting frustrated and starts to question your principles, and may start looking for other (more productive) projects to work on.

(Kim Storm).

I learned a bit of lisp, applied some basic color scaling theory, and produced a patch which added great new functionality.... That was Summer, 2001. Six years later, and the fruits of my early toil still aren't available in any released version of Emacs. So, while I continue to maintain a personally relevant programming mode, and contribute bug fixes where they impact that mode, I have not taken on any other "feature improvements" to Emacs. To me, the value equation just doesn't compute.

(JD Smith).

Clearly, the extended Emacs development cycle is proving frustrating for developers. The situation with the Linux kernel was once similar; changes merged at the beginning of a development cycle could take years to make it to a stable release. In that case, distributors responded by backporting changes into older releases, but that doesn't happen with Emacs.

The good news is that the biggest blocker - some questions about whether the Python mode code could be distributed by the FSF - appears to have resolved itself in the best possible way: the code has been cleared. Inevitably, there's another bug or two in need of squashing before the release can happen, but the remaining wait should be relatively short. Hopefully.

Some of the Emacs developers are already looking forward to the Emacs 23 development cycle. One of the first things that may go in is multi-tty support, which allows a single emacs instance to drive multiple terminals or X connections. This code apparently still does not work on all architectures, though, meaning it needs some work before it is truly ready. The other big change is a complete rework of character set handling; only Emacs would come with a news item reading "The Emacs character set is now a superset of Unicode. (It has about four times the code space, which should be plenty)." There's a lot of other work waiting to be merged, but getting the unicode-2 branch and multi-tty working together looks like it should be enough to keep the developers busy for a little while. Happily, they are starting to think about this sort of challenge rather than wondering if their previous work will ever be released.


to post comments

What about readable fonts ?

Posted May 17, 2007 5:11 UTC (Thu) by khim (subscriber, #9252) [Link]

I was under impression that this too was Emacs 23.0 material... Or is it included in 22.0 ?

Waiting for Emacs 22 (and looking forward to Emacs 23)

Posted May 17, 2007 8:39 UTC (Thu) by zuki (subscriber, #41808) [Link] (1 responses)

"multi-tty support, which allows a single emacs instance to drive multiple terminals or X connections"

You can already 'drive' multiple X connections - make-frame-on-display opens a window on another display.

Waiting for Emacs 22 (and looking forward to Emacs 23)

Posted May 18, 2007 4:20 UTC (Fri) by dann (guest, #11621) [Link]

multi-tty is about driving multiple X and terminal connections _simultaneously_.
Currently you can't have both an X and a terminal connection.

For the unicode branch another very interesting feature is the ability to use
anti-aliased fonts.

Emacs...

Posted May 17, 2007 14:20 UTC (Thu) by alessandro.russo (guest, #6471) [Link] (4 responses)

I remember very well the first time I realized the power of Emacs. It was in 1992, at Stanford.

Since then I heavily used emacs in the 90's, basically for latex and programming. I also wrote some small lisp code and helped RMS with X widgets.

Now emacs doesn't fit any more on my desktop. The main reason is lack of support for font antialiasing; it has been in CVS for years (unfortunately it doesn't work very well), but as far as I understand it's not going to be in emacs 22.

I use it from time to time for very specific tasks (for instance editing a binary file).

Too bad.

ale

Emacs...

Posted May 17, 2007 16:10 UTC (Thu) by zooko (guest, #2589) [Link] (2 responses)

I'm a long-time user of XEmacs, and I recently installed the new XEmacs unstable release in order to see if the fonts look better. They do! They are very pretty! I have found two new bugs in the XEmacs unstable version that I encounter frequently, but I know how to work-around. Other than those two, everything else works.

So if you like emacs but you can't stand ugly fonts, give XEmacs a try. I ended up compiling 21.5.27 myself from the source tarball on Linux. On Mac I use Andrew Choi's port of XEmacs (also 21.5.27), and on Windows I use cygwin's X11 and cygwin's XEmacs, 21.5.23. That one has uglier fonts than the Linux and Mac versions do.

Regards,

Zooko

Emacs...

Posted May 18, 2007 21:21 UTC (Fri) by hein.zelle (guest, #33324) [Link] (1 responses)

Indeed, unless the XFT support within emacs shapes up rather soon, I'm seriously starting to consider the impossible - switching to another editor. I'll certainly give xemacs a try if that ever happens.

I'm amazed this is going on for so long without even a properly working cvs version: I'm starting to have serious trouble getting emacs to even use a readible font in modern linux distributions, either with dual screen setups or high resolution displays. I really don't care about emacs and the way it works being old-fashioned, but the font issue had better be resolved sometime soon.

The point someone made about fixing the defaults is also a good one - it's rather prehistoric to have an editor which requires configuration to show colors, not switch the backspace and delete keys (does that still happen? not sure) and to open new files in a running editor process. I also blame distributors here: why don't debian/ubuntu/fedora add these things to a default ~/.emacs file and put it in /etc/skel ? Or put it in the system-wide emacs configuration debian already has?

Emacs...

Posted May 18, 2007 23:21 UTC (Fri) by hein.zelle (guest, #33324) [Link]

Well, thankfully Emacs proved me wrong within 2 hours of trying out the latest CVS version of the XFT branch. Apologies to the developers for the previous post - it seems to run fine, and I can even select the fonts from my Xresources file like I used to. And I must say - it looks great with truetype fonts and anti-aliasing.

Emacs...

Posted Jun 5, 2007 0:33 UTC (Tue) by fcsaszar (guest, #45596) [Link]

You can have a very nice looking font without requiring anti aliasing by using the Terminus font (http://fractal.csie.org/~eric/wiki/Terminus_font or apt-get install xfonts-terminus). In fact, I prefer it to any antialiased font.

Waiting for Emacs 22 (and looking forward to Emacs 23)

Posted May 17, 2007 16:11 UTC (Thu) by walters (subscriber, #7396) [Link] (2 responses)

I too am frustrated that the work I did on Emacs in college over 5 years ago still isn't out.

Emacs really needs to modernize in a lot of ways. The community has collectively learned a lot about how free software project management works best. In particular, Release early, release often. A bit of a cliche for sure, but that doesn't make it less true.

Besides that, Emacs really needs to realize that their users are almost universally programmers or system administrators, and that they are likely running Emacs on computers with fast processors and 512MB of RAM or above. The choices for defaults need to be fixed. There's no reason for font-lock to be off by default. Also, Emacs should take a hard look at some of the work people have been doing to make Emacs a more reliable IDE. Just in general, think about how to optimize the software for what normal people are using it for (instead of the people who are Emacs developers who still use it to read mail)

Emacs tuned for fast computers

Posted May 19, 2007 22:07 UTC (Sat) by giraffedata (guest, #1954) [Link]

Emacs really needs to realize that their users are almost universally programmers or system administrators, and that they are likely running Emacs on computers with fast processors and 512MB of RAM or above.

I wonder if programmers and system administrators have more powerful machines than the average computer user. The things I think of that need high power are end-user type things: web browsing, games, watching movies, etc. Emacs is highly useful today on systems that are pretty much worthless for all those things.

I am personally a good counterexample: I am a computer nerd and satisfy most of my requirements with old systems. The main pressure for me to upgrade is to be able to use newer Emacs, because Emacs does in fact assume a fairly modern computer. I got a new computer so I could upgrade from Emacs 19 to 20, and now I'm planning to upgrade so I can move to 21. I won't bother with 22. If I weren't a computer nerd, I could easily trash my current systems and go the store and get faster ones to put in their place. But because of all the customization I've done, it's not that easy.

Waiting for Emacs 22 (and looking forward to Emacs 23)

Posted May 20, 2007 14:28 UTC (Sun) by cyd (guest, #4153) [Link]

> There's no reason for font-lock to be off by default

It is on by default in Emacs 22.

Waiting for Emacs

Posted May 17, 2007 16:14 UTC (Thu) by tjc (guest, #137) [Link]

"Waiting for Emacs" sounds like a movie about a bunch of old kernel hackers sitting around in the day room of a nursing home waiting for a kernel panic.

Would it help if a distro included a CVS snapshot?

Posted May 23, 2007 5:54 UTC (Wed) by richdawe (subscriber, #33805) [Link] (2 responses)

I ran Emacs CVS built against DJGPP CVS under Windows '98 SE for a while, and I had no major problems with it. This was three or four years ago. But I'm not exactly an Emacs power user.

I think it would help to get some exposure in some distro like Fedora -- take a certain CVS snapshot of Emacs 22 and ship that. It seems to me that there is value in shipping CVS to get Emacs 22 more exposure -- more people using it means more people finding bugs, and contributing patches.

Debian distro includes a CVS snapshot

Posted May 23, 2007 20:31 UTC (Wed) by Max.Hyre (subscriber, #1054) [Link]

Check out Debian's emacs-snapshot package. M-x version returns
GNU Emacs 22.0.93.1 (i486-pc-linux-gnu, X toolkit, Xaw3d scroll bars) of 2007-02-17 on pacem, modified by Debian

True, it's in the unstable section, but I've been using it for months (>= 6) with roughly zero problems.

Would it help if a distro included a CVS snapshot?

Posted May 26, 2007 12:17 UTC (Sat) by richdawe (subscriber, #33805) [Link]

Fedora 7 is going to ship with Emacs 22.

Waiting for Emacs 22 (and looking forward to Emacs 23)

Posted Jun 4, 2007 23:52 UTC (Mon) by jdtsmith (guest, #45595) [Link]

As one of the referenced developers (who, as my full post emphasizes, is but a minor contributor to Emacs), I can say that despite the frustration, and whatever the process, the wait was worth it. Emacs 22 is a spectacular, polished editor with unparalleled functionality. Congratulations to Richard and the development team on the release!

JD


Copyright © 2007, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds