580 Second Street, Suite 210
Oakland, CA 94607
+1-510-628-3380 x5027

Date: Sun, 26 Nov 2000 20:17:13 -0800

From:   Grant Bowman


Subject:   Links not quotes of messages & subject lines

Hi Eugene,

[Responding to Eugene Kim's letter on November 26, 2000...]

I want to offer an example of my configuration and why linking to messages right now isn't ideal for me yet. I hope this helps to identify and work on the related issues going forward.

Basically, we need to integrate the email browser with a url browser. Logically approaching this, a mail browser is just a specialized browser that handles one type of content. A url browser handles many types of content like http, ftp and others. Why is it we need a special tool for email?

Eugene Kim,, [001126 16:01]:

[...] Third, link to messages rather than quote them. This is one that Doug is constantly promoting, and one that is nontrivial to do with the tools we have at our disposal. It's one of those areas that immediately arises as a necessary tool feature. Not surprisingly, the only two people who seem to practice this -- Doug and Rod -- are people who have tools that support

this feature by providing granular addressability and some linking capability.

You know, I don't know where to link to instead of quoting this... I put the link to your previous message above, but I can't like to a particular paragraph. Should we use the egroups links as a conconical source? I wish I could sort it by author like I can in my email client. Do we need one canonical source? If the canonical source moves, will I lose functionality?

Maybe we could write our own commentary and post it to our own web sites like Rod does? There is certainly the issue of software, but also of the service/implementation not going anywhere. The need for reliability and being able to read email is high.

Email is the Internet's killer app, which an average volume of 500 times that of web traffic. I found that number very interesting from a previous link someone posted. I don't know what the best answer is.

If we all had websites linking to and fro -without- new tools would it be easier or more difficult to follow the conversation compared to quoting in messages?

Without using a mailer that already has fine-grained links, I don't feel it's acceptable. The web/url browser and email client need to be used in a more seamless environment than is typically available.

An example from personal experience: I use the email client mutt running on a seperate machine, and a browser Netscape running locally. I telnet to this seperate machine to read my mail. Due to the way the IS configuration has changed, this is the best way to do it. Going back and forth between url browser (http, ftp, whatever) can be a real pain. I didn't have a way to automate mutt telling netscape (in my case lynx or w3m) to start a browser from within an email. Also any bookmarks set in lynx are lost when I go back to netscape. Next, when I am done browsing and without fine-grained links, it's difficult to take the content from there and save this to an email I want to write since I quit from one program to enter another.

Should I change my tools and habits to evolve? Yes. What do I change to?

Well, I found a tool called urlview for mutt. I just now took the time to download, compile, install and test it. It's still clunky, but it works better than no connection between mutt and a browser at all. It allows URLs within mutt viewed email messages to launch a web browser.

(story) of the install

Unfortunately, this took some persistence. I tried the URL provided in Mutt's help file...

...but the DNS didn't resolve, so the site doesn't seem to exist. I checked's site, bingo,

...contains urlview-0.9.tar.gz. After downloading and unpacking, there is even an install script specifically for SuSE. I was pleasantly surprised. Compiling it for both the machine I read email on and my own machine (since they both use different versions of some libraries), installing it, testing it out, seems to work ok.

Now, would I have cared much if I were a normal user that didn't want to install & configure files and binaries? I doubt it. I didn't want to take the time just now even if I know how to do it, but I am glad I did. It makes for a decent story of what and how long it takes to modify tools to collaborate better.

Eugene Kim,, [001126 16:01]:

[...] Fourth, change the subject lines of e-mail to reflect the content of the message, not the title of the thread. Lack of foresight has rendered the subject header irrelevant. Most subject lines tell me nothing except the subject of the first message in the thread and the fact that a message is part of a thread. This latter feature is mostly unnecessary today, as decent mail clients will use the In-Reply-To header.

I have found it difficult to follow some threads on mail lists because a few people's clients don't use In-Reply-To, causing choppy threads. My solution is to revert back to subject, author and/or date sorting. One of the four usually gets me what I want. Perhaps if I had the time and I felt it important enough I would edit every email that caused me this problem and add in the In-Reply-To header myself.

In the past while (two months or so) I have had trouble keeping up with the latest on OHS lists. Since I have many unread messages, usually I just use a command that gives me any unread messages within the mail list. This doesn't do me much good if I haven't kept up. This is an issue that might be solved with tagging or a command to show just new messages to the list vs. previously unseen/unread messages.

Another annoyance:

I also have some mail lists that always have the same few messages on the bottom. They are always there because the date fields in them are not correct. I guess I should just edit them in the files and put in an extra X-header if I want to save the original data as a comment.

My impression is that while all of these things are possible, the tradeoff between time and the difficulties often weighs on the side of not doing it because support levels and standard expectations are low.

Eugene Kim,, [001126 16:01]:

[...] These are just a few suggestions off the top of my head. I would definitely like to see those interested in practicing good KM attempt to incorporate some of these suggestions. I'd also love to see comments on the above as well as new suggestions for practicing KM using existing tools.

[... signatures omitted ...]

I hope this is along the lines of what you were thinking of.



Grant Bowman
+1-510-628-3380 x5027