I joined Quora early in the new year and, while the hype has in no way abated, I have found that increasingly over the months that I have been visiting, there is less and less reason for me to return on any regular basis. And this puzzles me as I think the concept is, for me, an attractive one.

In the five-odd months I have been visiting the site, I have managed to contribute 58 answers and over one hundred edits (more on that below); this strikes me as a reasonable commitment on my part and one that should return some dividend to me – at least in terms of satisfaction with the site. Alas, no.

And why not? For several reasons. And reasons that have only really become apparent over the last month as I started to spend time on the Unix & Linux Stack Exchange site. The contrast between the experience of the two sites, that are in essence very similar, couldn’t be more striking.

The primary difference between the two is that Unix.SE isn’t a social site: you don’t follow anyone. Which, when you think about Q&A sites, makes a lot more sense than Quora’s model. I still don’t understand the point of following people on Quora. Perhaps I am not following enough people, or the wrong ones, but I have yet to see any benefit in following on the site.

One other area of seeming similarity between the sites is the ability to vote up (or down) questions and answers. On Unix.SE, this is implemented as a real strength as these votes determine contributor’s reputation, described thus:

Reputation is a rough measurement of how much the community trusts you. Reputation is never given, it is earned by convincing fellow users that you know what you’re talking about.

This has a twofold effect: it motivates contributors to provide quality answers (and rewards them when they do), and more importantly, gives you a heuristic for how credible an answer may be. Quora doesn’t have this concept and, for me, is much the poorer for it.

Reputation has consequences on Unix.SE: the more you accrue, the more privileges on the site you are granted. Again, this builds the community’s capability while encouraging individual contributors.

Some of the other areas where Quora is really left wanting in comparison to Unix.SE are around the design of the site. The text box where you enter your answers or edits is a case in point:

On Quora, it is like trying to change a car battery through a letterbox. The single, overriding impression is that entering information into the site is neither a priority nor pleasurable which, for a site aspiring to be a knowledge repository is incredibly perverse.

This is compounded by the fact that the tools for presenting the information that you want to add to the site can charitably be described as rudimentary (one of the reasons that I have 100-odd edits is that it takes at least two to three attempts to present information in a semi-acceptable fashion). Searching for “quora markup tools” is as disappointing as it is frustrating.

Being able to present information appropriately is critical. The StackExchange sites allow all sorts of HTML tags and markdown, without apparently any additional XSS risk¹. The result is questions and answers that are formatted in a way that enhances their legibility, rather than detracting from it.

There are some other areas where Quora is not measuring up for me: opaque editorial and moderation policies and practices and an unwillingness to more actively curate the questions – and no, those two statements are not contradictory.

The overall impression I am left with is that while both sites have similar aims (albeit Unix.SE’s is much more specific in terms of the subject matter), Quora seems to me to be really failing at building a community around it. Unless, of course, you count social media experts…

Notes

1. God knows if that is actually the reason they refuse to allow potential contributors the ability to adequately format their information, but it seems plausible enough.

I have been a solid user of TTYtter for a couple of years now, much preferring the speed and power of a command line Twitter client. And, after having tried–at one time or another–most of the Linux or cross-platform Twitter clients, I can safely say that for me TTYtter is easily the best.

There is, however, a new ncurses client that is showing a lot of promise: Tyrs. Only at the 0.2.1.1 stage, and with a fraction of the functionality that TTYtter has, it is nevertheless a very nice client.

Requiring little in the way of dependencies, only python-twitter and python-oauth2, both available in the AUR or in the debian/Ubuntu repos (if you are using debian/Ubuntu, you will need the newer python-twitter-0.8.2 for it to run).

Tyrs is easy to configure, with good documentation and a simple configuration file, located by default at $XDG_CONFIG_HOME/tyrs/tyrs.cfg. The default keybinds are Vim-centric, naturally…

My config is pretty straightforward at this point:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

[keys]
retweet           = p
retweet_and_edit  = P
mentions          = a
reply             = r

[colors]
highlight         = 1
header            = 6
hashtag           = 4
attag             = 2
text              = 7
warning_msg       = 1
info_msg          = 2

[params]
refresh           = 1
tweet_border      = 0

Written in Python, the code is well commented and relatively easy to find your way around if your are keen on hacking.

The developer, Nicolas Paris, is both approachable and helpful and is actively developing Tyrs. With some more functionality in the pipeline, it looks to be a very nice application.

I posted a couple of weeks ago about how much I was enjoying Surfraw, to the point of writing a couple of my own elvi. As part of the ongoing tweaking of my Surfraw experience, I have been experimenting with different text-based browsers, to see which one best fits my workflow.

I have previously been using both links and it’s homophone, lynx. While these have been perfectly acceptable, they weren’t quite providing the level of functionality —and integration with my vim-oriented workflow— that I was looking for. Enter w3m.

w3m is a text-based web browser as well as a pager like `more’ or `less’. With w3m you can browse web pages through a terminal emulator window (xterm, rxvt or something like that). Moreover, w3m can be used as a text formatting tool which typesets HTML into plain text.

w3m is a step up from the other two for a variety of reasons. It supports tables, frames, SSL connections, color and inline images. It also renders pages in a form as true to their original layout as possible. The killer features on top of that are, for me, the above mentioned rendering of HTML as text (meaning you can save content easily), and the customizable keybinds.

There was however, one significant sticking point. After countless hours fruitlessly searching, I was unable to find a way to yank URLs to the clipboard. As I have been spending a little time of late on Unix & Linux Stack Exchange, I thought I would ask the good folk there…

Less than fourty hours later, I had an answer so simple and elegant, that I was literally grinning when I read it. It deserves quoting in full. It is that ingenious.

Set the “External Browser” option to sh -c 'echo "$0" | xsel'

You can use xsel -b to use the clipboard instead of the primary selection. An alternative to xsel is xclip (xclip for the primary selection, xclip -selection CLIPBOARD for the clipboard).

In ~/.w3m/config, that’s the extbrowser setting. Then press M to copy (yank) the URL of the current page, or ESCM to copy the URL of the link under the cursor.

You can use the second or third external browser for that instead; then the key combination is 2M or 3M or 2ESCM or 3ESCM. Gilles on Unix & Linux Stack Exchange

This simple hack means that it is now trivial to yank either the URL of the current page, or of the link under the cursor to primary selection. It was just a matter then of adding some intuitive keybinds – as the M and EscapeM combination are redolent of the unamed editor:

1
2

keymap y EXTERN
keymap Y EXTERN_LINK

The perfect UNIX solution…

Updated 8/5/11

Gilles has edited and updated his answer with a slight change:

1

sh -c 'printf %s "$0" | xsel'

This prevents a new line being appended.