A comment, via email, from Andrew Shebanow:

> I liked your article, but noticed that all of the software
> you discuss isn't really interactively collaborative.

> Several companies, including the one I work for
> (PlaceWare), offer browser-based real-time meeting tools
> that would work very nicely for some types of scientific
> collaboration. You can give presentations, annotate on
> slides, take polls, type text, view web pages, or do
> screen sharing so that people can see you run your
> simulations.

> All provide free versions of their service, usually
> limitedto 3-6 participants, for people to try out.
> Here are some URLs to play with:

> http://myplaceware.com

> http://www.webex.com

> http://www.centranow.com

> Hope you find this info useful,

> Andrew Shebanow

Edited by author 08-03-2000 12:35 PM
in reference to: http://software-carpentry.codesourcery.com...are/report.html#225

Great Article!

I'd like to see a sequel that touches on:
WWW Page authoring
www.everything2.com
Internet Relay Chat(irc)
Instant Messaging(especially odigo)
http://webapps.editthispage.com/
thirdvoice
http://www.gooey.com/

jtokash@homestead.com

in reference to: http://software-carpentry.codesourcery.com...ware/report.html#27

Sadly, Timedancer isn't dancing, or has run out of time...

Anyone know the story?

Greg Wilson wrote:

> If XML really is going to be the "universal canvas",
> perhaps we should start thinking about
> how to write programs *on* that canvas, as well as
> *for* it?

No argument from me on that score :-)

FWIW, C# takes a (small) step in that direction:

public class MyClass {
   /// <remarks>MyMethod is a method in the MyClass class.
   /// <para>Here's how you could make a second
   /// paragraph in a description.
   /// <see cref="System.Console.WriteLine"/> for
   /// information about output statements.</para>
   /// <seealso cref="MyClass.Main"/>
   /// </remarks>
   public static void MyMethod(int Int1) {
   }

   public static void Main () {
   }
}

I personally find this horrid. Why not leverage the parse tree that's already going to be built for this, or any other language, and decorate that tree with comments?

Well, because we'd need writing tools to share that parse tree with the compiler at a deep level, and that's not an easy transition to make.

- Jon

Peter Lim wrote:

> people are now looking to add WebDav support into
> AOLserver (or _back_ into AOLserver.)

Interesting. Didn't know that -- thanks!

Back to the future, eh?

- Jon

Edited by author 08-02-2000 11:26 PM
Walter Ludwick wrote:

> I wonder why there is no mention of Pike
> When you speak of "limited support for Wiki-like
> automatic formatting," is this what you mean?

No, I mean things like auto-<p> (converting double newlines into HTML paragraph markers).

Pike's a horse of a different color -- an outliner that manages structured data, and can emit HTML. I consider it to be a bit of a special-purpose tool, though, not as universal as the basic Manila kit. Since (last time I checked) your Manila site has to connect back to Pike via XML-RPC, Pike tends to run into firewall/NAT resistance.

- Jon

Jason Reidy wrote:

> Remember that you're asking people with many other things
> to do to set aside some time and learn something which
> may or may not be around in a few years.

Actually, my wish is not to ask people to replace one obscure and arcane markup discipline with another. My wish is for the industry to deliver writing tools that support various requirements, including math, citations, etc., in ways that are transparently easy to learn -- and therefore irresistible -- but that at the same time yield valid, well-structured, and thus reusable content.

Until the industry delivers such tools, there's no reason to abandon TeX. Conversely, until users demand such tools, there's no reason for the industry to deliver them.

- Jon

in reference to: http://software-carpentry.codesourcery.com...are/report.html#134

I wonder why there is no mention of Pike in this paragraph (http://pikebeta.userland.com/ - a cross-platform outlining tool that updates Manila sites via XML-RPC).

When you speak of "limited support for Wiki-like automatic formatting," is this what you mean? That's a vague description, which doesn't adequately describe what Pike does, IMO.

in reference to: http://software-carpentry.codesourcery.com...are/report.html#168

Was TeX immediately picked up by everyone? Not that I recall (brief encounters while I was in jr. high in the mid-late 80s showed a smaller use than it currently has). Likewise, don't expect an immediate uptake of the latest and greatest.

Remember that you're asking people with many other things to do to set aside some time and learn something which may or may not be around in a few years. TeX had the defense of being completely and totally documented free software. It was also the only game in town for high-quality mathematical publishing.

The free part wasn't so much about liberty as permanence. TeX will be around for as long as the acid-free paper on which it's printed. And documents written in TeX will be decypherable. Start adding in a ton of extra packages and gizmos, and you start adding in problems. However, the wide availability of LaTeX and other macro packages pretty much guarantees their availability for at least a hundred years.

Consider how much _recent_ history has been lost when ISPs go under, projects fail, etc. Our great library of the Internet has a horrible retention policy...

Jason

Edited by author 08-02-2000 08:59 AM
in reference to: http://software-carpentry.codesourcery.com...are/report.html#127

Several years, AOLserver/GNNserver/Naviserver had some of this same capability when used with the AOLpress browser/editor. AOLpress was a just-average browser, but you could edit web pages in the browser itself (not by opening another program, such as Netscape Composer). If you had the correct permissions on the AOLserver web server, you could save your pages or your changes back to the web server.

The current version of AOLserver, v3, has removed the capabilities (unfortunately) because most people didn't use them (as mentioned in your article, most people just look for information instead of contributing). Also, due to lack of interest, AOLpress is no longer being developed or supported, and because of licensing restrictions in some of the source code, AOLpress has not been open-sourced.

AOLserver, however, has been changed from restrictive licensing to Open Source with version 3, and people are now looking to add WebDav support into AOLserver (or _back_ into AOLserver.)

in reference to: http://software-carpentry.codesourcery.com...are/report.html#160

Reading the list of things that can't be done with equations on the web, I'm struck again by how web-hostile programs themselves are. No language I know of allows programmers to mark up source to support semantic search, embed anchors and links, and so on. Javadoc and Python doc strings are a good first step, but are very limited, language-specific, and completely unconstrained syntactically.

If XML really is going to be the "universal canvas", perhaps we should start thinking about how to write programs *on* that canvas, as well as *for* it?

Jon writes:

> The granularity of CritLink is, I think, at the level of
> elements or perhaps phrases? I can't access crit.org at
> the moment to check.

Yep, "phrase" level - you copy an excerpt you wish annotate, and the system brackets it in cute little glyphs (of different styles, conveying the nature of your comment). The glyphs use mouse-over bubbles to show an excerpt from the first comment - offhand, i suspect for scaling it should indicate the number of comments, etc.

Edited by author 07-31-2000 12:40 PM
CritLink, Loquacious, BooHoo

Thanks to both Monty and Ken for reminding me of CritLink, which I ought to have mentioned in the report.

Also noteworthy in this vein are some free services running at greenspun.com:

- Loquacious (http://www.greenspun.com/com/home.html), used to enable readers to add comments to a page linked to from one of your pages.

- Boohoo (http://www.greenspun.com/boohoo/index.tcl), used to enable readers to add "related links" to a page linked to from one of your pages.

The granularity of Loquacious/Boohoo is page-level.

The granularity of CritLink is, I think, at the level of elements or perhaps phrases? I can't access crit.org at the moment to check.

The technique used here works at the element level, which has advantages (can refer precisely to locations within a document) and disadvantages (requires special preparation). I personally think that such preparation shouldn't need to be special, it should just be an automatic consequence of using smarter writing tools that anticipate the possibility of granular interaction with the content when it goes online, and instrument the content accordingly.

- Jon

in reference to: http://software-carpentry.codesourcery.com...are/report.html#222

The computer security community has an open source technology (Ka Ping-Yee's CritLink) featured at crit.org which allows anyone to annotate html documents and start discussions, much like your annotation links on this document.

Monty

in reference to: http://software-carpentry.codesourcery.com...are/report.html#223

I've encountered a problem which may be impeding some people who otherwise would be submitting notes in your document. The popup submission window, at least in IE 5, is a bit too small so that he "post comment" button is scrolled off the screen - and there's no scroll bar to get to that point. Tabbing out of the comment box or mouse-dragging a select region to that point will expose it, but maybe some people give up before trying that...