With inline LaTeX previews, we're already surprisingly close. In fact, I'd say that going all the way would be almost a step back. WYSIWYG is ultimately not an ideal editing paradigm: it wins in the short term, being easy to learn, but drags you down in the long term.
I've recently started using Quora a bit more. Unlike StackOverflow, they use a WYSIWYG editor. I've found this significantly less convenient than StackOverflow's markdown. Similarly, switching from Word to LaTeX was an improvement for most tasks once I got used to it.
Unfortunately, LaTeX has a bunch of its own shortcomings not related to it's non-WYSIWYG nature. For common tasks, I think going from markdown to LaTeX is ideal. Markdown itself is far from perfect, but it's the best compromise I've found especially with Pandoc's extensions.
So here's my idea for a great emacs-based document editor: markdown with inline math previews coupled with a full live preview to the side. All the necessary modes for this already exist (like whizzy TeX and AucTeX's previews), so it should be much easier to put together than a full WYSIWYG editor. More productive, too.
If have come to greatly appreciate Lyx' [1] WYSIWYMean approach. When you are mostly writing code or plaintext markdown is great, but if you have to typeset and edit a lot of formulas using Tex quickly becomes a pain. Lyx allows you to edit them graphically with the standard Tex syntax and displays them while you type which is a huge improvement to just rendering them after you have finished it. Unfortunately the videos I've found online all heavily use the mouse interface to choose subscript, special characters etc. which doesn't do Lyx' real efficiency via keyboard shortcuts justice.
If you want to try it yourself: The most important commands are [ctrl]+[m] to enter math mode (or [ctrl]+[shift]+[m] for display formulas) and [alt]+[m] for a lot of math shortcuts, specifically [alt]+[m] and then any bracket to open matching size adapting brackets of this type. To enter a sum for example press "[ctr]+[shift]+[m]\sum[space]_k=1[space]^\infty[space][space]\frac[space]1[down arrow]k^2[space][space][space]". The shortcut [alt]+[p] and then [space] to choose the paragraphs meaning (section, itemize, standard, …) is also a huge productivity boost.
>With inline LaTeX previews, we're already surprisingly close. In fact, I'd say that going all the way would be almost a step back. WYSIWYG is ultimately not an ideal editing paradigm: it wins in the short term, being easy to learn, but drags you down in the long term.
That --in the extend that it happens-- is a byproduct of the limitations of current WYSIWYG editors, not something inherent in the idea of WYSIWYG editing.
It's not like we had a lot of brainstorming and innovating solutions competing in this area (in fact, there are only 3, all too similar, major products: Word, Pages and Open Office, of which one has 90% of the users).
Second, what's "inconvenient" and "distracting" for one user, is a must and an inspiration for another.
Not to mention there are several different use cases for WYSIWYG editing, and LaTeX style editing doesn't cover them all. For example it's dreadful for quick experimentation with placement and formatting of small, disparate elements (basically, anything not according to the "spec").
>I've recently started using Quora a bit more. Unlike StackOverflow, they use a WYSIWYG editor. I've found this significantly less convenient than StackOverflow's markdown.
WYSIWYG's strong point is not small website comments.
> That --in the extend that it happens-- is a byproduct of the limitations of current WYSIWYG editors, not something inherent in the idea of WYSIWYG editing.
If WYSIWYG editing is not inherently limited, why do pretty much no editors actually support it? For example, spelling/grammar checkers are useful, but when pseudo-WYSIWYG editors implement them, they underline the text, or change its colour. This colouring and formatting doesn't appear when the document is printed.
Hyperlinks are usually hit and miss; sometimes they print as 'blue and underline' (but, of course, are useless), sometimes they don't.
Annotations don't print, but they do show on the screen. Automatically-generated content like page numbers are often shown differently, eg. with a grey background. Again, this doesn't appear when printed.
In short, what you see is certainly not what you get; but removing these non-WYSIWYG features would be a step backwards.
>If WYSIWYG editing is not inherently limited, why do pretty much no editors actually support it? For example, spelling/grammar checkers are useful, but when pseudo-WYSIWYG editors implement them, they underline the text, or change its colour. This colouring and formatting doesn't appear when the document is printed.
That's just an extra convinience feature you can turn on or off, and doesn't have anything to do with whether the editor is WYSIWYG or not.
It's not like the program forces you to see red swiggly lines underneath words when editing...
>Hyperlinks are usually hit and miss; sometimes they print as 'blue and underline' (but, of course, are useless), sometimes they don't.
Which is configurable too in all editors I know of. And the "blue and underline" is not exactly useless -- it conveys the information that this part of the document was a hyperlink initially.
>In short, what you see is certainly not what you get
No, it's 99.9% of what you get, minus additional layers of information that people expect to be there, and you can turn on or off.
I have found WYSIWYM concept (What You See Is What You Mean) and WymEditor[1] in particular a very useful solution, at least for website editing.
Previously I have used Xinha, FSCKEditor and similar, but there was always some problem with the resulting markup. WymEditor copes much better in this regard... (not affiliated, just really like the concept and WymEditor)
> It's not like we had a lot of brainstorming and innovating solutions competing in this area (in fact, there are only 3, all too similar, major products: Word, Pages and Open Office, of which one has 90% of the users).
We also had PageMaker, FrameMaker, Word Perfect, Ami Pro.
You might want to give plain TeX a try--it's surprisingly lightweight (once you learn how to change font sizes and/or decide not to care). A lot of the bloat comes from LaTeX and using 'eplain' will give you most of the crossreferencing and bibliography commands (if you need it).
Running latexmk -pvc in a separate terminal will give you rebuilt dvis on the fly (or pdfs). AucTeX can generate inline previews for plain tex just as well as LaTeX.
Of course, org mode is the standard recommendation for "emacs based markdown replacement with inline math previews"
>So here's my idea for a great emacs-based document editor: markdown with inline math previews coupled with a full live preview to the side. All the necessary modes for this already exist
So org-mode has you covered there, basically. Org isn't markdown, but it's close (and full of tons of note-taking features), and you can do inline latex with live preview in org-mode. I use it for notes all the time.
I thought working in WordPerfect's text mode was an extremely easy and productive way to do document formatting. Then again, I still remember the little keyboard shortcut cheat sheet [1] at the top of our PC at home.
Though I end up just using LaTeX (with memoir class) most of the time, I've always thought LyX hit a sweet spot between strict enforcement of document structure and forcing people to write markup.
For the life of me, I can't figure out why LyX doesn't exist as a collaborative google docs style web app.
Still has the problem of needing to talk collaborators, advisors, and other folk into learning LaTeX. In the alternate reality of my dreams, all people that I work with in an academic environment would write LaTeX and collaborate via GitHub.
I've personally been thinking about some sort of GitHub, LyXified markdown browser plugin, in which collaborators' changes get committed to a separate git branch. Then I die a little inside and soldier on.
> I don't know how to use Org mode, and don't know what it does (it seems to do so many things), but if it displays through Emacs then there are many formatting features that it can't display in a WYSIWYG fashion like Libre Office.
I can't believe Stallman doesn't know how to use Org mode. If he is interested in selling people on Emacs, then Org mode is one of the killer features for the presentation. I don't expect him to know something he has no use for, but he should know the most popular components in the Emacs ecosystem. Org mode is one of the only reasons I started using Emacs.
I completely believe it. First of all, Stallman isn't known for hiding his opinions, or masking his version of the truth.
But beyond that, Stallman has often failed or refused to use technologies that the rest of us take for granted. Years ago, I spoke with him about a proposed bill in the US Congress that would have affected intellectual property laws. (I can't remember the specifics.) At some point, Stallman, who had strongly encouraged people to contact their representatives to oppose the bill, told me that he hadn't actually read it. I told him that it was on the Web, to which he responded, "I don't surf the Web."
Now, I can understand being against certain browsers, servers, and operating systems. But to flat-out refuse to read things on the Web struck me as counterproductive. I don't know if he has changed his attitude toward the Web in the years since, but assume that when he says he doesn't use a certain technology, he means it.
"I generally do not connect to web sites from my own machine, aside from a few sites I have some special relationship with. I fetch web pages from other sites by sending mail to a program (see git://git.gnu.org/womb/hacks.git) that fetches them, much like wget, and then mails them back to me. Then I look at them using a web browser, unless it is easy to see the text in the HTML page directly. I usually try lynx first, then a graphical browser if the page needs it." [1]
Is this purely for anonymity purposes, as he seems to imply? It seems like quite the workflow when a (relatively) modern anonymity tool like Tor would probably be fine.
I wonder if it's deliberately complex to discourage easy, shallow browsing, or if he has some other rationale.
RMS just has a workflow, developed probably sometime in the late 80's and nearly only involving emacs. He also uses a Lemote Yeeloong (which can barely run ancient versions of IceCat), and runs it in a tty most of the time with just emacs.
TOR is bullshit because it's a festering pile of assholes all trying to pretend that no one knows what no one is doing, without knowing who's doing the pretending not to know.
C'mon, guys, I'm serious. Please don't sniff my
exit node. It's not nice, you know! We're all
supposed to be mysterious strangers, and not know
each other, but still implicily trust all participants
even though we'd never do that in real life. And oh, by
the way, here's a browser bundle that still uses
third party cookies, and runs javascript by default,
but fuck the world, because javascript is hella
convenient, and animated web pages are just super cool!
In several instances, Stallman has asked people to complain to their representatives that bills posted online required running non-free Javascript on their computers in order to access them. I'm not sure if the instance you mention was influenced by this issue, but he does avoid reading things that require non-free software.
> But to flat-out refuse to read things on the Web
As you described it, he did not actually flat-out refuse. He simply stated that he hasn't read all the documented published on the world wide web, since he don't "surf" the web.
His approach to the web has been described many times. When someone sends him a link that they want him to read, stallman uses the program wget to download the file and reads it offline.
When I specifically asked him if he had read the bill in question to which he was objecting, which was available on the Web, he said that he had not, because he doesn't surf the Web. He, not I, equated "surfing" with "reading documents."
The impression I got from that conversation was that Stallman makes no effort to go out and find things that would be useful and of interest to him.
I don't call that "surfing," but "research" and "responsible," especially from someone who then calls upon people to contact their public representatives.
I don't care how much he hangs out on Hacker News, or elsewhere. I do care that he clearly indicated, in that conversation, that he will not do the necessary legwork (which doesn't even require using your legs) that can even help to strengthen his arguments.
I also think that it's helpful for everyone to read things from opposing and diverse viewpoints. By only reading those documents that his supporters send him, I worry that Stallman is ignoring arguments that may inform his own, or even change them.
>I worry that Stallman is ignoring arguments that may inform his own, or even change them.
Of course he is! It is called confirmation bias. It is very very strong, especially with people like Stallman who hold strong beliefs. We are all guilty of it to some extent.
Confirmation bias (also called confirmatory bias or myside bias) is a tendency of people to favor information that confirms their beliefs or hypotheses.[Note 1][1] People display this bias when they gather or remember information selectively, or when they interpret it in a biased way. The effect is stronger for emotionally charged issues and for deeply entrenched beliefs. They also tend to interpret ambiguous evidence as supporting their existing position. Biased search, interpretation and memory have been invoked to explain attitude polarization (when a disagreement becomes more extreme even though the different parties are exposed to the same evidence), belief perseverance (when beliefs persist after the evidence for them is shown to be false), the irrational primacy effect (a greater reliance on information encountered early in a series) and illusory correlation (when people falsely perceive an association between two events or situations).
A series of experiments in the 1960s suggested that people are biased toward confirming their existing beliefs. Later work re-interpreted these results as a tendency to test ideas in a one-sided way, focusing on one possibility and ignoring alternatives. In certain situations, this tendency can bias people's conclusions. Explanations for the observed biases include wishful thinking and the limited human capacity to process information. Another explanation is that people show confirmation bias because they are weighing up the costs of being wrong, rather than investigating in a neutral, scientific way.
----
The most interesting is the last link. They did a study about the show The Colbert Report, which is satire, He plays a parody of conservative political pundits. The study looked at the viewer's political beliefs and what they thought of Colbert's.
"More liberal participants believed Colbert was liberal and that the show was satirical. More conservative participants believed Colbert was conservative and genuinely believed his “satirical” arguments. Essentially, viewers of liberal and conservative orientations tended to perceive Colbert as supporting whatever views they personally held."
I actually read a fascinating article about the Colbert Report saying that conservative viewers of the show thought that he was a conservative satirising a liberal satirising a conservative. Fascinating, really.
In general, it is important to do independent and individual research before voicing an opposition.
However, it does depend on context. If this unnamed bill just happens to be SOPA, I am actually happy that we do not put such restrain on opposing voices. At some point, group action is more important than a handful individuals that has the time and energy to research the question independently.
The concerning part with that anecdote is that Stallman dodged the real issue --- reading the bill he was encouraging people to complain about --- by brushing it off because of the medium through which it was available. It's not unreasonable to expect someone to have read, in part at the very least, a bill they're actively campaigning against.
Has he given his reasoning for that? Does it tie into his privacy concerns? I have read a little about how his computer is air-gapped and such, but I don't remember specifics.
Depends on your definition of "very." It was created 10 years ago, this year. In the grand scheme of the emacs history, it's around a 3rd of the life of emacs already
I've seen Programming teachers requiring emacs for a course while not using anything it has to offer, not even as a programmable editor, and even displaying signs of anti-abstraction (no regularity in actions for a sequence of identical text modification tasks).
Some of the people responding are steering the discussion to a layout language with a preview window. I don't know if they are doing it because they prefer to work in such a user-hostile mode (I did this for a book, in Eclipse. Ugh.), or if they think this is a more sane goal.
WYSIWYG has its own issues. most users of word processors have no idea that paragraphs are objects in an object model, but the command structure only becomes clear when you realize that. Most users just hack at a document it until it looks right enough. At the really diabolical end of the spectrum I could show you an Ericsson documentation template that manages to manifest dozens of bugs in Word, laying in wait to eat your previous hour's work. I'm sure you have inherited documents like that.
It's all more or less a kludge, and WYSIWYG never is quite, nor is it real direct manipulation. At best it is something like "moderately friendly visual document CAD, if you get the trick behind the slick appearance."
If not quite the Correct Answer, it might be the Best Available Answer.
My mother used to be a Linotype ninja, blazing thru the sunday paper's ad inserts. The user input language was a page description language, little different from PostScript or HP's PCL (or HPGL/2). While modern systems are more accessible, I've not seen a system since that matched that concision and productivity.
I loved the precise control of WordPerfect (reveal codes) and FrameMaker (parametric styles). I can't comment on modern InDesign (stopped around version 2), but early on it was not considered a feature complete replacement for FrameMaker.
People speak highly of LaTex. I've not used it for real work.
I've done my share of UI, scenegraph, layout engines, plotter/print drivers, prepress / imposition, etc.
I gave up trying to integrate general purpose constraints systems into layouts. Too many edge cases, hard to debug. I now believe layouts heuristics should be implemented imperatively.
So I created DesignGridLayout, which captures grid based rules for UIs in a "fluent" API (aka method chaining). https://designgridlayout.java.net/examples.html (The awesome Jean-Francois Poilpret has maintained, extended the project for years.)
Though the box model sucks, I haven't yet given up on some mythical, simple scenegraphs + typography + content hybrid. I still don't know what that'll look like.
I would not say it is naive. I wonder if it is an itch Stallman has had for 25 years, but it is certainly an itch I have. Not for me, I would do perfectly well without word processed documents, but for customers and coworkers, they like PDF if not powerpoint or MSWord documents.
For now, I have edited and generated them with reStructured Text, but the workflow is not entirely satisfactory (and I don't feel like dipping into python to improve things there, I would rather write rst document preprocessors in lisp when possible).
Another aspect is that if we improve the graphic abilities of emacs, it may be easier and faster to implement chrome like lighttable to attract newbies. Sublime Text probably has capted a percentage of users who would have chosen emacs, if it could have competed in the esthetic plane. Also, it is a little regretable to be limited to ascii art for the few diagrams that are sometimes needed in the documentations we write as programmers. If we could edit vectorial graphics in emacs, it would be great (we can already render svg, but perhaps thru an external program).
So I would not say naive, but on the contrary, it may condensate a critical mass of itches to motivate some scratches.
Quite simply, explicit markup makes it very easy to see what formatting will be applied to what text.
WYSIWYG only shows you the end result, with no clean way to see how you got there. Was this font introduced because of some theme? Was it applied because of some toolbar button? Is it the result of some template? Was it copied from somewhere else, thereby baking someone else's theming into the copied text?
These are the questions that make WYSIWYG so confusing. These are the things that make explicit markup so straight forward. I don't think you can have WYSIWYG without the confusion or while maintaining the power of explicit markup.
If anything, Markdown or RST or Org provide a compelling middle ground: Markup is still explicit but minimal, and styling tries to come as close as possible to WYSIWYG without sacrificing control.
This, I think, is a far more compelling route to take than WYSIWYG or LaTeX-style explicit markup.
i have read literally hundreds of threads discussing
this general issue over the years, and i can say that
this comment right here has come as close to the heart
of the problem as any other one i've read in that time.
congratulations, derbasti, on your observational oxyopia.
I prefer the paradigm where you edit in something like markdown but the live preview is available on the right hand side or down below the text. All it needs is a catchy acronym.
The controls for the text should be immediate in the text, that makes sense and is the most powerful implementation. But you have to see what it all means and be able to identify problems along the way when you make a mistake & know what your markdown is creating. Otherwise you type it out and notice that you forgot to force 15 line breaks.
I think this is usually called What You See Is What You Mean, or WYSIWYM.
However, there could be a performance problem. Font setting and layout in Latex is markedly superior to, say, web browsers or Word. But Latex often takes several seconds to lay stuff out. Then again, Indesign can do it, so it should be possible in general.
markdown has several such solutions, online and off.
the offline apps are typically mac, the leaders being
"marked" (marked2.com) which works with any text-editor,
and "multimarkdown composer" (multimarkdown.com), which
wraps a dedicated-editor around the conversion-routine.
i will soon be introducing my own light-markup system
-- z.m.l. (zen markup language) -- and will offer apps
which are cross-platform offline, as well as web-apps,
including one with an a.p.i. that can be used by anyone.
send it a light-markup .zml file; it sends back .html.
and this is where stallman's request has gone awry...
namely, you don't need to wrap the conversion-routines
into the app, or change the interface of the app at all.
instead, simply route your light-markup plain-text file
to a converter, and show the output in a preview window.
and yes, it should be side-by-side with your editor and
show the changes in real-time as you make (or save) them.
my light-markup system differs from markdown in that it:
1) is targeted at long-form documents, such as books,
2) avoids the problems which plague markdown, 3) focuses
on lightness as an asset that _eases_editing_, rather than
as something that _fosters_readability_of_the_raw_format_
(as there's no reason to read the file in that raw format).
speaking of readability, as well as my focus on long-form,
.html is not the only output format supported by my system;
we also need .epub, .mobi, .pdf, and still-to-come formats.
the other important thing about z.m.l. is that its focus
on _books_ shines a different focus on needed functionality,
compared to that provided by generation of a mere web-page.
all this and even more, coming before thanksgiving day...
This sounds exciting! (Though I must say, I am all but married to Org at this point.)
On a different note though, I feel that Microsoft Word, and web browsers, really do a terrible job at rendering text. Kerning is often bad, there are frequent widows and orphans, pagination is often a mess, hyphenation is laughable. These are real issues and they are not solved by a different markup language.
So far, TeX and Indesign are the only software I know of that really manage to solve this problem. I would love to see a new export format that can deal with this.
Maybe what we really need is a new middle ground: A kind of LLVM for text processing. An intermidiary machine format that serves as a target for all different kinds of markup languages that can be further compiled to beautifully rendered text in whatever format you desire.
needed to clear my plate / get that off my chest...
*
if you'd be so kind as to point to either of those,
here at h.n. or elsewhere, i'd consider it an honor.
(i have never been able to get any traction here;
i guess i need to go see what the reddit kids say.)
*
i too am totally dissatisfied with web typography.
it's atrocious. but a lot of the problems you note
_can_ be corrected, such as widows, and pagination,
if you adopt a mindset that considers it important.
and i won't stop until i've created e-books that
are _both_ highly functional _and_ very beautiful.
anyway, thanks again.
-bowerbird
p.s. but i don't think we even need hyphenation now.
Visualize RMS as like the patron saint of a religion, which is full of schism and sub-sect.
He's using his position to influence mindshare, and he doesn't think he's got a handle on the details. He just wants some of his devotees to start kicking the problem around.
Stallman lacks a coherent vision. He has an end goal but he really doesn't have a great plan to get there. It's really frustrating. Emacs could be a lot better. For instance, it has taken forever to get a high-performance Lisp working inside of Emacs. I think Guile is partly there?
Anyway, since we'll all be long dead before his plan starts to work, I think the better solution is to support inexpensive software. For example, I pay for Sublime Text. Recently I bought PixelMator and Sketch, and I'm planning on learning how to use them soon. :-)
Sure it would be great if Free Software ruled but faster change comes with a paid ecosystem. The real problem was that software was expensive. If it's simply inexpensive, we'll get most of what we need.
You misunderstand the point of the FSF. It's free as in freedom, not free as in no money. The whole thing started because rms wanted to modify a printer driver to give functionality their last printer had. He wasn't being stingy, he just wanted his workflow back. The FSF has no problems with charging for software, they have no objection to Red Hat, for instance, charging for the GPL Linux kernel (although they have other problems with Red Hat). The FSF is concerned with users having access to the source, not with everyone being users or non-users having access to the source.
The real problem was that software was expensive??
Also, I don't know how much of the success/failure of FOSS can be pinned on RMS. He's just one man. The plan was never for the entire movement to be dependent on him.
Moreover, the amount of work that he has done towards this end is incredible. His productivity (at least in the early days of GNU) was astounding and inspiring. Perhaps one RMS didn't get us there, but two? Or five? It'd be a treat to know what that would have looked like.
Is it that Stallman lacks a manager's mindset of breaking the vision into manageable pieces, and then delegating them out? Or is it that this is so against the ethos, that he couldn't make it happen with a volunteer workforce?
Well, I guess if you consider his other interests, like the GNU Userland (which is most of what users recognise as linux), the GPL licsense (and generally 'Free Software'), and his part in the opening up of the software/hardware market...
I think this was a joke from Stallman. Your comment reads like we can conclude from EMACS going off track that he sucks as a leader. You and I might (quick strongly in my case) dislike Free Software, but I think we can recognise that Stallman gets things done.
If you want to point at a place where Stallman didn't succeed, I'd probably look more at Herd. At least people use Emacs.
Fair enough. I consider Emacs a success by almost any measure. The question is, "If he wanted it for 25 years, why hasn't it happened?" It could be time and attention. It could be the ethos of working with volunteers.
This isn't coming from a dislike of Free Software per say, more just an appreciation of the challenges of herding cats (volunteers).
This is the epitome of the challenge of open source.
RMS whines : "25 years ago I hoped we would extend Emacs to do WYSIWG word processing. That is why we added text properties and variable width fonts. However, more features are still needed to achieve this.
Could people please start working on the features that are needed?"
And he's 100% accurate, it has been 25 years, and there is an open source WYSIWYG word processor, called Libre Office these days, but that isn't what RMS wants. He wants someone to do the work to make his tool of choice into something which can do what he wants to do in it.
A lot of people go this way, and we see several tools that all do variations on the same thing in their own peculiar way (Vive du choix!) but that means it is really really hard to figure out how to get somethings done when each set of tools rely on their own set of other tools.
The nice thing about Cathedrals is that you know what is expected of you :-)
> called Libre Office these days, but that isn't what RMS wants.
If you bothered to read the mail thread, you would see that RMS answer that specific question. The answer he gave was that Libre Officer do not have the multi-buffer system of emacs, and the core design is not made to be a lisp interpreter. As such, adding emacs like features to libre office would unlikely work effectively in the libre office project.
It would simply be easier and less work to modify emacs to do WYSIWYG, then adding emacs work flow and emacs features to libre office.
I don't see how what you just quoted contradicts Chuck's "[Libre Office] isn't what RMS wants" post. Libre Office is a free software WYSIWYG word processor with an extremely powerful feature set. But it isn't what RMS wants, because it doesn't have some things Emacs does.
It would be easier and less work to modify Emacs to do WYSIWYG than to add Emacs work flow and features to LibreOffice.
I'd suggest that you're vastly underestimating what would be involved in "doing WYSIWYG" to a sufficient degree that Emacs would actually be useful as a word processor to people who are not already dedicated Emacs users. I did read the mail thread, and it's pretty clear to me that a lot of the folks there -- and I do not mean to take away from their brilliance by saying this -- don't have a blinking clue what a full-featured word processor is capable of.
The reason that this hasn't been done, I suspect, is that the intersection of users for whom WYSIWYG is important and for whom having Emacs-ish features is important is virtually nil.
"The reason that this hasn't been done, I suspect, is that the intersection of users for whom WYSIWYG is important and for whom having Emacs-ish features is important is virtually nil."
We know from inspection the value is at least 1.0 (RMS).
> And he's 100% accurate, it has been 25 years, and there is an open source WYSIWYG word processor, called Libre Office these days, but that isn't what RMS wants. He wants someone to do the work to make his tool of choice into something which can do what he wants to do in it.
It is not that RMS simply want his tool of choice, but rather tool that has the work flow and features he needs. Its like someone comming from git, and wanting that workflow in cvs. Its not the name or code base of the program that matter. Chuck's post however implies NIH and a stubbornness in using other peoples tools, rather than a need for a specific work flow and feature sets.
I don't see why this is specific to Open Source. Android does all the same stuff iOS does. WebKit was a senseless clone of stuff already handled well by Gecko. How many Javascript interprers do we need in this life? What's the point of clang again when we have gcc and icc and msvc?
People work on what they want to work on, based on the products that they want (or that they think someone else might want to buy). Sometimes they're right. Sometimes they're wrong.
Sometimes they troll. RMS isn't known for that (at all), but... if you did want to troll the free software nuts, WYSIWYG emacs certainly has to be up there on the list of flameworthy titles...
It is specific actually to all volunteer organizations. Such organizations have a hard time getting things done for which the population of the organization that would benefit is 'small' relative to the capable volunteers. I often use KiCAD as an example, its an excellent EDA program for Linux but is nearly as capable as early Windows offerings from commercial outfits were 10 years ago [1]. The challenge was that people who wanted to use EDA tools couldn't write one, people who could write EDA tools didn't need one. So until Jean-Pierre came along there were no EDA tools for Linux.
Basically in a 'free software' context, there was no way for people who wanted an EDA tool to fund the development of that tool spontaneously. Jean-Pierre got it to the point where is was kinda sorta useful and it has since gathered enough momentum to get to be very useful.
So RMS wants emacs to have a WYSIWYG mode, and has wanted it for 25 years, but there is no way for him and say the other 150 people [2] in the world that want that to express some sort of financial interest so that someone who could do it would be willing to sit down and spend a year and do it.
At a company they have this revenue stream and someone says "We need a new product" and they pick one and set some developers on the path of making it real. But in the open source world we don't have that (either the unified direction or the funding to push for it).
I've long felt that we could perhaps create a 'prize' system ala the X-prize where people could donate to a 'prize fund' if they wanted something, like "I'll donate $10 to a prize fund for an awesome CAD tool that runs on Linux." and that prize fund would grow as people donated to it, and anyone who wanted to claim the prize could do so by shipping / releasing a product that met the requirements of the people who had donated to the prize fund. Once a prize fund got to $100K or so I'm sure you would find a couple of programmers who would take the chance to sit down and write it to claim the prize. This would satisfy RMS' philosophy that you pay for the creation of code, not the use or redistribution of it.
[1] This is not a disparagement of KiCAD, it started much later and it is going through much the same evolution of other tools that started decades before it.
[2] This number is pulled out of the air because it is also a problem of identifying how many people would like this feature, could be 10 could be 10,000.
Richard Stallman, the founder of the Free Software Foundation in the field of Information Technology, has been awarded nine honorary doctorates from various international educational institutions from 1996 through 2011 including the North American Lakehead University in 2009, and now refers to himself as "Dr. Richard Stallman" in speeches, talks, videos and email.
It's a pain to see "Dr" next to bullshit like this:
> "Skype: No way! That's nonfree (freedom-denying) software."
You don't like skype, fine. I am prety sure what it enables millions of people to do: freely communicate no matter the distance, but I don't really get what freedoms it denies. Do I lose som freedom just but installing it, or must I use it for that to take an effect?
Freedom, in Stallmans words, is not meant to be 'free as in beer' (like free skype calls). Instead he refers to the freedom to read, modify, understand, and ultimately choose everything about the software that you use. I.e. be able to answer these questions (say about Skype):
- Is any communication not properly encrypted
- Is your communication being spied on / being recorded
- Is there a way to add feature "x" to the application
- Is there a way to remove feature "x" from the application
- Is there a way to use the application on your specific hard / software
With Skype, none of that is possible. You can't see what's happening with your Skype calls internally, you can't see if there's somebody else listening. If Skype comes out with a new interface, and you really don't like it, you can't change it, if Skype removes support for older operating systems that you just happen to have to use because your IT department still wants to run Windows XP, then you can't support, change or fix any of it.
Of course, Skype is a free product, so you may say the proverbial never look a gift horse in the mouth, but Emacs is also a free product, and provides all these benefits I listed above.
You have to understand Stallman as someone who is concerned about the future, not so much about the now. Right now using Skype is all fine, but imagine the world in 20 years time, when everything is connected to software, and every movement you make is tracked. Wouldn't it be nice if all the software then would be fully transparent, allowing you to see every detail that is happening, and change it all to your liking, without being afraid of what happens if you do this or that.
Sounds like an utopist dream right? And Stallman is the guy who tries to fight for it, however unwinnable the fight may seem.
Let's go over your list for an ordinary phone call. Here the 'client app' is your phone terminal, and the 'server' is the phone operator.
> - Is any communication not properly encrypted
Yes, it's not encrypted at all.
> - Is your communication being spied on / being recorded
I would say yes, at a similar level of probability as a skype call.
> - Is there a way to add feature "x" to the application
> - Is there a way to remove feature "x" from the application
The specs for the phone terminal are not open-source, so you can only do that with some hacking. The same is true for skype, even though there it might be more complicated technically.
> - Is there a way to use the application on your specific hard / software
Irrelevant for a physical phone.
The main difference between the skype client and a phone, is that a phone follows an open protocol, and therefore you can build your own. This is the only reason I can think of why using a phone is 'free' while using skype is not. But I think that if the skype protocol was made open, Stallman would still argue that it is not free and therefore should not be used, and therefore I do not understand his argument.
Right, but at least you can tell. No-one knows whether the encryption Skype claims to be applying actually works. And you can integrate an encryption system into a phone if you want to, whereas if you try and do that with skype they'll change the client so that your thing doesn't work any more.
> The specs for the phone terminal are not open-source, so you can only do that with some hacking.
Speak for yourself. There are open source phones available if you want them. There are no open-source skype clients.
> Irrelevant for a physical phone.
Sure. But I don't have to use a physical phone. I can run a virtual phone on my freebsd box, whereas I can't run skype.
> The main difference between the skype client and a phone, is that a phone follows an open protocol, and therefore you can build your own. This is the only reason I can think of why using a phone is 'free' while using skype is not.
That's not a small difference; I'd say it's the most important difference.
Stallman (and FSF) uses Asterisk, and assumingly a free phone stack when using their physical phones.
Stallman has always refused to use unfree software in devices he own or are directly using. Skype is within that criteria, while an Asterisk powered free software device is not.
"Free software is a matter of liberty, not price." - Richard Stallman
So to understand the word freedom in the context of free software, I tend to lean back towards liberty as a philosophy question. What freedom means, is a question famous people have pondered over for ages. So to ask a single simple question if Skype gives freedom, here is a quote from John Locke:
"Persons have a right or liberty to not be subject to the inconstant, uncertain, unknown, and arbitrary wills of others."
Since the developers of skype can subject their users to any number of inconstant, uncertain, unknown, and arbitrary decisions, like including the NSA as secret listeners to private communications, Skype can not be defined as providing liberty.
It's not a software problem. Or at least, it's not purely a software problem. For the vast majority that freedom would make no difference. They don't know how to change things anyway. I remember a few months back on here when we were going 'Why don't people use PGP more? Oh yeah, the interface sucks.' And did anyone change the interface? Nope, not to my knowledge at least.
Did any of us make better encryption tools, despite there being available encryption libraries? Not to my knowledge.
That sort of freedom is tied to your capabilities as much as it is your tools. If your understanding of a computer goes something along the line of 'that magical glowing box under the desk.' it makes no difference either way.
To get to real freedom, in the sense that Stallman seems to mean, requires more than a change to open source software, it requires a system be built from the ground up to not just enable but also encourage co-evolutionary mastery. And it's not clear how that would work. Having the source code of the system available is probably a part of it, but you also need people to be able to start off with simple changes using the same language that the system is described in - preferably they'd use that language at least to some degree to instruct the system to do basic operations.
A lot of that probably has to do with how modular you can make a program - when you make changes how much of the program do you have to hold in your head to understand their effects? And I think that's probably, ultimately, a processor architecture level decision if you want it to be how most things on the system work.
Looking at it from the software side: at the moment, even with open source software, sure you can theoretically go in and learn about the program and make alterations and ... but that is not a trivial thing to do. Even for people with deep knowledge of computing it's not a trivial thing to go and learn about how the program works, and if they don't already know the language it's even worse. There are loads of open source projects out there at the moment where what needs to be done is essentially obvious but there's such a lack of people with the abilities and interest to do it.
"I remember a few months back on here when we were going 'Why don't people use PGP more? Oh yeah, the interface sucks.' And did anyone change the interface? Nope, not to my knowledge at least."
Some people do use PGP, but they pay for it or it's an internal component in a larger system. GPG, the OSS version does have a sucky interface, but that's not the real issue people don't use it, there's three others:
1) Most people don't give a shit about encryption
2) Most people are too fucking lazy to understand/use encryption (Not being cynical, this is reality.)
3) Writing good encryption software is thankless and discouraging. There's a ton of commercial companies that use SSH and GPG that haven't donated squat to the developers.
"Did any of us make better encryption tools, despite there being available encryption libraries? Not to my knowledge."
You don't work where I do. I wrap GPG in another layer of software to make it more usable to the above average folks at my job. It's mainly limited to automated processes.
That's just stupid dream not utopist.
Having access to source code would me absolutely nothing, unless you are willing to check the actual code installed on the each system involved in communication chain.
Back to the point: Skype does not give me some freedom, fine. But it denies zero freedoms too.
By their philosophy, it does. If that feels hand-wavey to you I would agree; I find the whole "define 'freedom' as 'things I want'" thing more than a little sneaky--but it is consistent with the rest of their philosophy.
Would you be against defining freedom as synonym with liberty, and thus see the FSF philosophy as one that adapts the philosophy of liberty of old to the digital age?
Your post doesn't quite parse for me; while I am fairly conversant in political philosophy I'm having trouble connecting "philosophy of liberty" to anything the FSF does.
My annoyance is simpler, though: I dislike propaganda. That the FSF seeks to re-define the word to the exclusion of all else is bothersome, much as, for another example, the American Republican Party seeks to redefine "socialist" to mean "center-right". It's the poor debate tactic of the extremist who can't accept the validity of opposing positions.
You can choose to be denied a freedom, quite obviously. If you choose to use Skype you are granted the ability to make free Skype calls. You are denied the freedom to:
* Freely distribute
* Modify
* Read the source code.
Among others, so it's a trade off. One that Stallman disagrees with and believes denies you too many freedoms.
Stallman is a weird guy, and hardly a pragmatist. But I don't think the things he says can be described as Bullshit, they are largely internally consistant. And to be frank, given the NSA revelations, I think we need rather more weird idealists than less.
You don't, mostly[1], lose freedom by installing Skype. You lose freedom by becoming reliant on Skype, and by installing (and using) Skype you are putting yourself at significant risk of that.
[1] You also may lose some control over your system, which may limit your freedom. At the extreme, consider malware. Recent, relevant Dilbert comic that I assume is an exaggeration: http://www.dilbert.com/strips/comic/2013-11-06
It's extremely hard to agree with Stallman, but you have to admire the fact that he lives by his principles.
He truly believes the world would be a better place if all software was free, and he lives his life in support of that. People rarely devote their lives to a single principle like that.
Skype might be convenient and wonderful in the short term, Stallman believes that supporting it is not the right step for society in the long term.
It made me wonder what's free-er about an ordinary phone call.
You can use any phone that you like to place a call. You can even build your own phone. (Depending on your choice of phone you might have to rely on non-free parts like GSM voice codecs.)
You have the freedom of choosing between multiple service providers, all of which are interoperable. If you want, you can even get access to the underlying SS7 phone network and be your own provider.
Until something like Ekiga does multipart conference calls, a minimum of 10 nodes and more voice only, then it is hard for me to switch from something like Skype.
Me and my buddies recently switched from Skype to Mumble+IRC. Works great, I like this setup way better. The big downside is of course that it requires a server.
there is no relation between "Dr" and his rant about Skype.
You would not believe the number of PHD who are not that intelligent (to be nice). Also Stallman is a master in its profession and "Dr", who is just acquired by doing a Phd, seems to me like it's not doing him enough justice.
It denies freedom to the software more than freedom to the person. For copyleft licences the software is most important; for permissive licences people are most important.
I've always felt it was this kind of thinking that puts Emacs at odds with the UNIX philosophy of do one thing well. Unless your one thing is everything-you-can-do-with-text.
Some projects are like that, because the synergy from a core functionality works well with added features. Its like how a window manager has menus and keyboard shortcuts even if its main purpose is to draw windows, or session managers are normally full working desktops.
Emacs do exactly one thing well (editing text), and a bunch of other things sort of OK. The sort of OK stuff is there because their core, they are about editing text. Like with an IDE, if the editing is crap, the ide is crap. As such, all IDE's require a good text editor to start with before they can do anything well.
Emacs is a conscious computer virus with the aim of hollowing Unix out function by function, eventually to overthrow it. It's like a vampire: a shadow of what made it beautiful in life (elisp vs the old lisp machines), sustained mostly by blood (of its new users) and its long-term goal of revenge.
Yes I think Emacs' philosophy is at odds with UNIX philosophy; I also think it's /superior to/ UNIX philosophy. It sounds like you take UNIX philosophy to be gospel, and deviation from that to be incorrect. I'm not trying now to convert you to Emacs' philosophy. Rather I think there are a lot of people who only know UNIX philosophy, when in fact there are alternatives.
emacs is entirely and wholy at odds with the UNIX philosophy, indeed. emacs is a Lisp Machine, running in a VM on unix. Read the Unix Hater Handbook for more info.
Could people please start working on the features that are needed?
As someone aware of gnu, but not an active participant, how effective are requests like this? Does "Can people start working on this?" actually get results? I'm curious as this gets to the heart of why they may have trouble finishing things. (You can't toss money at someone to do the dirty work)
I'm coming with an open mind, and would like to hear either side of this.
I love emacs, and although I have long tried switching to other editors (I am fairly determined to use web based applications only, brackets is getting close) I havent been able to replace it yet.
However its the only application I didnt know how to copy and paste in when I started using it, its still the only application I use that I dont know how to resize the text in.
It would be kinda nice to see people work on those type of things.
Because I likely dont think the things you relate to web software being unpleasant are inherent to the actual platform and can be fixed, both by 3rd parties producing better web based software and by the platforms themselves being fixed.
That combined with the advantages of having a single, cross platform runtime platform that is shared as opposed to owned as being too important to ignore, I think its silly that people have to consider writing the same app 3 times at a minimum to reach a reasonable portion of the audience.
My biggest issue with web apps is that I don't have control of the software -- someone else can unilaterally decide to change the interface or functionality from day to day. (GMail is an example of a case that regularly irritates me by doing this.)
While I appreciate what you're saying about a common, open runtime being valuable -- I've considered using a self-contained, special purpose browser as the GUI for a project before myself -- I feel like web apps give users less freedom ultimately. Even in the worst case proprietary desktop app with nasty DRM, I could resort to reverse engineering the executable on my machine to find out what it does if I really had to; as soon as you move the computation to a remote server, you're totally at the whim of the operator for the continued ability to even access the program.
Thats exactly what I meant by something that is not inherent to the platform, The example I used (brackets) is an application built primarily using web technologies that isnt served from a website / you have complete control over.
If you just want a common runtime, what's wrong with java apps
Oracle v. Google. By CLR I assume you mean the CLI, but only a fool would trust a "standard" controlled by MS and driven by a proprietary implementation, not to mention that the latest version doesn't even have a patent promise from them.
And all they had to do was pay Keker & Van Nest for two years, without knowing whether they'd win or not - and Google didn't even make a VM that can run Java apps!
Not to mention that Oracle has appealed, so we still don't know if Google is clear or not.
If you want to risk getting yourself or your company involved in that quagmire, by all means do so.
Heh C-+ / C-- actually work, at least in my emacsforosx, have no idea when that started working, but nice, one complaint off my list, but its not hard to imagine many more to make emacs more user friendly
> However its the only application I didnt know how to copy and paste in when I started using it, its still the only application I use that I dont know how to resize the text in.
It's got a very nice way of handling text size and other text "attributes". All the various kinds ("faces") of text are arranged in an inheritance hierarchy, so a change in the size of the text in the base of the hierarchy propagates to all of the text in the application (minus the size of the text in the title bar, which is of course controlled by the operating system).
So for example, to make all Emacs text (except the title-bar text) pretty darn big, you just eval the following:
The thing about "What You See Is What You Get" is that you have to define 'where you get it'. You have to define 'it'.
The implementation will look vastly different if you define 'where' as a printer, desktop, tablet, mobile phone, wearable device, etc. It sounds like RMS means his desktop/laptop computer. On the upside, you have a user archetype: Richard Stallman. A good product manager would start building up a list of user stories:
- As Richard Stallman, I want *goal* so that *benefit*.
>The implementation will look vastly different if you define 'where' as a printer, desktop, tablet, mobile phone, wearable device, etc.
Not since more than a decade ago, when we started using the same rendering infrastructure (e.g Display Postscript, Quartz, Cairo, what Windows uses etc) for both print and display.
In fact nowadays, with hidpi (retina) displays, you even get similar resolution between the average print and monitor.
HTML and CSS are very, very explicitly not WYSIWYG. Their specifications are crystal clear on this point. The same HTML document can and should be displayed differently depending on device.
Your ebooks are also not WYSIWYG. They're also specifically designed to display differently on different devices, and according to user preferences for things like font, text size, brightness, background, line spacing, etc..
WYSIWYG is exactly the opposite of these formats, and does not depend on the device. Postscript and PDF are WYSIWYG.
It's because you're not looking at the data in a WYSIWYG environment. If I take an Acrobat document or MS Word document, print it, and look it next to my screen, I'd expect the layout and 'look' to be perfectly identical (barring blurriness due to one-or-others resolution limitations)
That "all three use Display Postscript, CoreGraphics and Quartz" doesn't mean they use the respective drawing context as it's intended for screen/print work.
Editing LaTeX with Vim on a Mac also uses Quartz to render the Vim window, that doesn't mean that you get what I talked about.
With respect to your examples now, HTML and ePub/Mobi don't guarantee they are WYSIWYG and stable between different viewing devices and contexts. On the contrary, for example even something as basic as the line width can change by making your browser or eBook reader wider.
That said, they are not the standard output of WYSIWYG editors we were talking about either (despite some WYSIWYG editors being able to save in those formats too).
Word, Pages, Open Office et al have print view modes that look identical to what will be printed. As do programs such as InDesign, Quark XPress, etc.
Neat. Really. I really want to use emacs but I find it hard to justify learning all it's intricacies when I'm only going to use it at certain times while programming.
This kills two birds with one stone, gets me off MS Office and into keybind heaven
For some people Emacs acts like a maelstrom for everything you do on a computer. In the beginning you use it to code in one language, then in another language, then you use it for (La)TeX, then you use it as an organizer, then you use it to read your mail, then your RSS feeds, then you embed it into text-fields in your web browser...
At some point everything that does not act like Emacs feels annoying and uncooperative. That justifies the learning curve easily.
Amen. Think of 20 years using the grossly identical UI for all those disparate tasks... How much time do you think you lose learning the little quirks of a new app just because you're typing things inside a differently labelled box?
And then you can script or macro all of those processes using the same set of concepts...
I like enriched-mode for simple markup purposes, like bolding headers. It's nothing spectacular, but makes an on-screen document that much more readable, and it doesn't add that much bulk to a text file, just a few extra markup directives.
It's easy to use, too. Select the text, ALT-o b
= bold, and so forth.
Still, a true WYSIWYG editing mode would be cool once in a while. Although, it's not that much trouble to select text and paste into a nearby LibreOffice window for true formatting.
but we have many great wysiwig editors... isn't emacs specifically for all the people who want a highly configurable weird and wonderful dev tool and care little about wysiwig because they are writing code etc.?
i don't agree with what those kinds want.. but they should be allowed to have it. :)
That may very well be the problem, especially since on shutdown/logout Windows 7 always shows some items not closing, even if none are left. But emacs is about 50-50 on closing while other things always go away.
GNU Emacs 24.3.1 (i386-mingw-nt6.1.7601) of 2013-03-17 on MARVIN
No thank you; I like Emacs as a text editor, and if I need a word processor, there are alternatives out there. No need to make Emacs even more complex.
I've recently started using Quora a bit more. Unlike StackOverflow, they use a WYSIWYG editor. I've found this significantly less convenient than StackOverflow's markdown. Similarly, switching from Word to LaTeX was an improvement for most tasks once I got used to it.
Unfortunately, LaTeX has a bunch of its own shortcomings not related to it's non-WYSIWYG nature. For common tasks, I think going from markdown to LaTeX is ideal. Markdown itself is far from perfect, but it's the best compromise I've found especially with Pandoc's extensions.
So here's my idea for a great emacs-based document editor: markdown with inline math previews coupled with a full live preview to the side. All the necessary modes for this already exist (like whizzy TeX and AucTeX's previews), so it should be much easier to put together than a full WYSIWYG editor. More productive, too.