Emacs Dev Inefficiency and Emacs Web 2.0? (2010)

By Xah Lee. Date: . Last updated: .

Xah Lee wrote:

Starting Emacs Community? http://xahlee.blogspot.com/2010/09/starting-emacs-community.html

Right now xahlee.org is a static site. No forum or any type of interaction. Am not sure what technology i can use with my hosted server to make it Web 2.0. Also, i wish to keep the HTML valid, but i think that's hopeless. Possibly i could start a forum using WordPress.

Jonas Stein [n…@jonasstein.de] wrote:

I think the community is very spread. There are many pages and most need more manpower.

I would rather suggest to fusion some communication channels. One newsgroup with mirror to Mail (perhaps gmane?) would be fine. One wiki with easy public editing.

All services should not be bound to one individual. It should be possible to pass the admin-job to someone else, if the maintainer has no more spare time for the project.

I like your website and i am glad about your contributions, but i think you should not create another micro-forum.

Perhaps you can cooperate with the maintainer of the emacswiki? Or we could try to start a project in the wikimedia world.

But we need the support of the community.

Thanks a lot for the input.

I also like the idea.

I think that it is difficult to get these all together, precisely because one might say selfish motives, including myself. FSF has its own agenda, emacswiki too. FSF is kinda haughty in that it wants to be stand-alone and does not want to take in any other's work except subsumed into FSF/GNU. XEmacs's years of superior tech going basically away now is a good example. [see GNU Emacs and XEmacs Schism, by Ben Wing] FSF could use emacswiki too, but it just doesn't. For example, many of its FAQ or pages and document could link to emacswiki. e.g. Emacs Menu Usability Problem. (See also: My Experience of GNU Emacs vs XEmacs (2007).)

If FSF wants, FSF people could easily initiate the talk to Alex of emacswiki, so that emacswiki can function and provide many services emacs users need such as a coherent library depository, coherent online emacs interactive help center, or even a coherent emacs Web 2.0 social networking for emacs users … so many potentials.

Emacswiki has its own agenda. For example, Alex Schroeder prefers to use his own wiki software the oddmuse [http://www.oddmuse.org/cgi-bin/oddmuse] . He doesn't want to switch to the much widely used and far more powerful MediaWiki [http://www.mediawiki.org/wiki/MediaWiki] used by Wikipedia. The links on the emacswiki seems to want to stick to sister sites such as meatballwiki.org instead of the much better quality Wikipedia. [see Problems of Emacswiki (2008)]

FSF does not want to bundle Visual Basic mode into emacs, even though it is top 5 most used lang. Similar for PHP mode, or code from Lennart's EmacsW32, or the incredibly non-trivial js2-mode [http://steve-yegge.blogspot.com/2008/03/js2-mode-new-javascript-mode-for-emacs.html] by Steve Yegge, and many other modes written by others … POV-Ray mode was discussed and i think paper signed perhaps 2 years ago but still not in GNU Emacs and i don't think it's even in the agenda …

It just take years for a mode to get into GNU emacs even when the mode author is fully enthusiastic about it and willing to sign the paper and FSF willing to include it. There are so much obstacles and frustrations. The primary communication for GNU emacs dev is its dev mailing list http://lists.gnu.org/archive/html/emacs-devel/ . It is very messy and inefficient. In today's twitter, instant messaging, and voice and video chat days ( Skype, gtalk, msn, yahoo, aol …) that just about any joe uses to communicate across the globe, but GNU emacs dev stuck with the ancient mailing list software the GNU Mailman GNU Mailman (1999), then the web version is done by using the 1999's software MHonArc [http://www.mhonarc.org/] , a Perl script that converts email texts to plain HTML the upload periodically to web server as static HTML pages. It doesn't have a modern web interface. No forum, no admin/management structure, no user profile, no image/file archive, no sitefeed updates, no integration with bug database, no integration with source code revision system, no who is online, etc. [see “Free” Software Morality, Richard Stallman, Paperwork Bureaucracy]

In the mailing list, there's huge amounts of opinion throwing. Some appear to be from just Emacs users who haven't coded elisp. People don't agree on things. Evaluation of opinions are not based on a person's existing programing work and knowledge, or established factors such as user base of his emacs mode or distro, or any scientific basis such as social research (e.g. polls) or systematic analysis. Many recognized experts such as Steve Yegge [Steve Yegge] , Ilya Zakharevich [http://search.cpan.org/~ilyaz/] , Randal L Schwartz … or Aquamacs's David Reitter [http://www.david-reitter.com/] and EmacsW32's Lennart Borgman [http://www.emacswiki.org/emacs/LennartBorgman] in the list all seem to have to behave very sheepishly to conform to the culture there. Constantly half of the discussion is about FSF's concept of freedom … For example: in recent discussion about incorporating some emacs mode that hookup google services break out into tens or over a hundred messages about the cons/pro of freedom and legality issues (and that's just a recent example). Any discussion on User Interface is a immediate bomb shell, often spanning into hundreds of messages and debates. (and in the end, typically nothing is done or changed, after years, and then the very same issue crops up for a repeat of discussion with new comers.)

Many bugs, suggestions, even if accepted as better, but it is tremendously difficult to get code in or contribute. The culture of the list just kills it. The FSF emacs list has a attitude … for example, say someone made a suggestion and actually submitted a patch. Instead of giving warmest welcome and spend time to help the coder to get his patch in, typically the coder has to do all the work, learn all the systems used by FSF, learn all the ways of FSF's philosophies, learn and follow the full culture of FSF's workings. Then, almost completely on his own without much verbal encouragement, the coder has to make the code work well with the whole elisp code base. The whole experience as i've observed, seems to be like this: “you want your code in GNU emacs? go spend a year learning about FSF, then if your patch passes all checks and we like it, then sign paper to give us the copyright, then it might be in, but we might change many things of your code even rewrite or not even include it if we so happens to forget or end up not using it. You'll get a ‘thank you’ mention in our acknowledgment page that contains hundred other names. Thank you for your efforts and hail to the great freedom commune. Happy hacking!”.

There's also emacser.com, a emacs wiki for serving Chinese emacs users in Asia. Although we all try to do best for the community, but there's just little chance for these perhaps largest ones to merge in some way. Similarly, there's quite a few emacs blogs…

Similarly, GNU emacs, Aquamacs Emacs, Carbon Emacs [http://th.nao.ac.jp/MEMBER/zenitani/emacs-e.html] , http://ourcomments.org/Emacs/EmacsW32Util.html and ErgoEmacs, XEmacs, SXemacs http://www.sxemacs.org/ … almost all are started or maintained by one single person. Merge between any two of them is just impossible for social reasons.

I really wished emacswiki could be like wikipedia so that the articles there are coherent and comprehensive but for various reasons Alex has his own approaches and ideas.

I think the practical solution is just for individuals to make things happen, especially with commercial vision, as in Stack Overflow site. Though, it takes tremendous time to make it happen.

If i am to start a emacs web 2.0 site, i'll need to register a domain name first, then figure out a hosting package that allows any of PHP, Ruby, Python, and arbitrary libraries that i can actually install, then research into software for web 2.0, features to consider (do i want it to be a lib repository like CPAN, or social network like Facebook with chats, or more as comprehensive expositions like Wikipedia, or more a efficient Questions and Answers place like Stack Overflow, or all of the above?). Then, time will be spent to admin and managing the software, and maybe coding and modifying too. Then gather community, advertise, writing tutorial … in the past 3 to 5 years am basically spending on average perhaps 3 hours a day on just writing my emacs tutorial and ErgoEmacs … and the little donation and advertisement i make out of this effort is perhaps $5 a month.

Thinking about your suggestion… if i start a site, make it full open source, no commercial backing, call for collaboration with open source spirit… it might end up just another blog/emacswiki without being high quality as it could be… when i'm out of time, some contributor become maintainers on their spare time… and it's just another wiki that lingers among the thousands. Unless me or someone have sufficient charisma to get all others to unite into a vision on a single goal, such as Richard Stallman with his GPL or Linus Torvalds with Linux. Or, with lots of money to burn as Ubuntu Linux by millionaire Mark Shuttleworth.

Humm… this post started as just few word of thanks but typing and typing turned into a rant. ☺ Hope it didn't offend anyone. I hope emacs the best.

For Alex Schroeder's reply and the original thread at comp.emacs, see: http://groups.google.com/group/comp.emacs/msg/367cee373202358f.

If you have some ideas about improving emacswiki, Alex has opened a discussion at http://www.emacswiki.org/emacs-en/2010-10-02. (thanks Alex)

See also: GNU Emacs Development Inefficiency (2010).