Software

Talking Shop: Idea for building the perfect e-mail program

TechRepublic Vice President for Content Bob Artner and Tom Ranalli, director of IT for Lyon & Lyon, an intellectual property law firm in Los Angeles, debate the perfect e-mail program.


TechRepublic Vice President for Content Bob Artner and Tom Ranalli, director of IT for Lyon & Lyon, an intellectual property law firm in Los Angeles, continue their e-mail discussion today with a dialogue about their designs for the perfect e-mail client.
A new feature continues today on TechRepublic called In My Humble Opinion, derived from the common e-mail and chat room abbreviation IMHO. On occasion, we will ask two people to discuss, or sometimes even argue, an interesting point. Let us know if you are interested in participating in a future IMHO article.

Follow these links to catch up on the exchange between Bob Artner and Tom Ranalli:


From: Bob Artner
To: Tom Ranalli
Sent: Mon. 01/08/2001 12:53 PM (EST)
Subject: E-mail Utopia

Tom: Well, at long last we're in agreement on some issues!

We both see the value of Intelligent Folders, as well as the ability to automatically discern whether a person should receive a file, or a pointer to that same file on a network server. In addition, we seem to agree that the OS UI would look different if the OS were really designed around the e-mail client, and not the other way around.

We don't agree about viruses, though (oh well, perfect agreement would have been too much to expect!). To be honest, I think you misunderstood my take on this. I certainly agree with you that integrating the OS and the e-mail client could result in potentially more damaging viruses. My point was that if we really honored e-mail as the Killer App, then the particular vulnerabilities of the Office macro language would never have been allowed to happen. Specifically, the ability to write a macro that would start spamming entries in a client Address Book would never have seen the light of day.

Now before you (or anyone reading) jump on this, let me say that I understand the difficulty. After all, a macro language is designed by definition to automate actions that the user is allowed to do manually. Therefore, it's hard to limit the power of the macro language without also limiting the functionality of the end-user (or the developer who is writing for the end-user).

Difficult, but not impossible. Just off the top of my head, how about this: Require some kind of password authentication for certain types of macro operations. In other words, before certain operations took place, you'd have to check a permissions table or request a user password. Cumbersome, but it might prove effective.

By the way, how useful do you think that macro feature is? Have you ever had an end user come up to you and say, "Tom, what I need is the ability to create a macro that will automatically send files to everyone in my Contacts folder."

Me, either.

While we're on the subject of the Utopian E-mail Application, let me throw out a couple of more dream features. (And, yes, as I said yesterday, I know that some of this is already worked in.)
  • Better Meeting Scheduling: When scheduling a meeting, I'd like to be able to see what my resources are for the specified time just by clicking on an AV Closet icon or something—just show me if the projector is available without making me try to book it as a resource—I know that's a small thing, but it bugs me. I'd also like to be told when I've tried to book too many people into a conference room, and what kind of outlets it has.
  • Unified Messaging: Granted, you can get this now, but most of the solutions I've seen look like the 1.0 versions that they, in fact, are. I'd like one place to go to check my voice mail, e-mail, and PCS messages. I'd also like to be able to use my e-mail client for desktop videoconferencing, since that's just another form of communication.
  • Contextual Searching: Several firms are offering, or will shortly offer, software that scans your spam messages, looking for e-mail that you might actually want to read. Unfortunately, most of those are based on some type of keyword selection scheme and are vulnerable to the same kind of tactics that made HTML metadata keywords worthless. A real spam sifter would look at the kind of messages I read, the contents of my legitimate Inbox e-mail, and learn what kinds of topics I care about. In fact, I'd like to be typing an e-mail to someone, and have my e-mail client notify me that my spam inbox contains an e-mail on this very topic. Granted, I'd probably turn this feature off much of the time, but I'd like to have it.

What about you?
Bob



From: Tom Ranalli
To: Bob Artner
Sent: Mon. 01/08/2001 11:27 PM (PST)
Subject: Re: E-mail Utopia

Bob: I knew you'd be shocked that we're finally in agreement ... but as they say, even a blind squirrel finds a nut every once in a while!

First, a brief word on macros or the possible neutering thereof. Expanding a bit on your concept of permission tables/passwords—after all, this is a Utopian world we're creating—implement public key decryption before any type of code is executed (programs or macros). With the Pentium IV processor-driven PCs entering the mainstream, the execution time spent on decrypting before executing would be relatively trivial. You would exchange keys with the company/individual providing the software; the trust level could be checked before downloading or installing, giving the user a comfort level that they aren't getting something virulent in their system. The publishers would have your key as well and then be assured that you weren't pirating the software and effectively, you are now a registered user before the software is used.

Now, more on the UEA—Utopian E-mail Application (look, Bob—once it becomes an acronym, it's as good as written, right?). As you mentioned, many of these things are available in places, but they all aren't in the same place.
  • The Scheduler: I'd love an intelligent meeting scheduler, because I also find it annoying that something so simple as resource planning—which a computer is BORN for—for some reason cannot be done. I want a graphical representation of the floor the conference room is on, where the outlets are located in the room, the human capacity, the coffee delivery schedule. Why not?
  • Unification: Don't forget inbound faxes, voice recitation of e-mail, automatic foreign language translation to text and/or voice of inbound messages, and universal PDA exchange. I want it so that when I walk out the door, I am completely and totally at one with my handheld PC and no matter where I am, my Mobile UEA helps it appear as though I'm virtually everywhere and anywhere.
  • Is It Spam or Is it Beef?: I'm not sure if your spam sifter would be the answer to the spam problem. I'd be more inclined to have a universal spammer database that would be updated like the virus signature file is today. All the keywords that have been predetermined to signal spam (by some sort of artificial intelligence method and by one's own definitions) would be used to make the choice if it is spam or not. The universal spam database would be updated regularly and transparently from an Internet server to which one is constantly connected.

Shifting gears for a moment, I want to throw in a few things I DON'T want to see in the UEA.
  • Death to the Blue Screen: First, I'll be happy to see stability, plain and simple. The system crash would be a thing in the past—barring a hardware failure or a power outage, I want my new computing appliance to be as bulletproof as Kevlar. The days of the ubiquitous "blue screen of death" and Microsoft's response "wait until the next rev" must end. With a well-designed operating system and tightly integrated code, this side of Utopia is possible.
  • My Client is Better than Yours: This is where I get Fascist on you, Bob. In my version of Utopia where the e-mail is the core app in the environment, I only want to see one e-mail client, not everyone's version of what is best. Now before you or someone screams "what about free enterprise?," ask yourself in a true Utopia, isn't everyone on the same page?
  • The Web is the Best: OK, maybe not the best, but a great idea if you forget your trusty PDA at home when you're stuck in Bayonne in the dead of winter. How about a Web client that has ALL of the functions that the e-mail client has and even the interface is exactly the same? Expanding this theme even further, I posit that not only is e-mail the Killer App for our UEA, but Internet browsing must be a critical piece of our kernel.

Time to come in for a landing, Bob—at least until the next Linus Torval reads this exchange and runs with it.

Tom



From: Bob Artner
To: Tom Ranalli
Sent: Tue. 01/09/2001 3:45 PM (EST)
Subject: Utopian Email Solutions—how we get there

Tom: Well, we started off talking about blind carbon copies, and we're ending up discussing spam sorters of the future.

Unfortunately, I don't think we're likely to see practical unified messaging for a few years yet. Further, I think limiting spam is going to get more difficult, at least for the near-term. As you see continued consolidation in the ISP industry, it won't be practical to maintain lists of "spam-friendly" ISPs, and block them, without also blocking access to millions of legitimate potential customers.

I think intelligent agents at the client end are the best bet.

In fact, I think intelligent agents in general are going to be more important than most people think, over the next five years. As you point out, Intel and the other chip makers are introducing one GHz chips now—that will probably mean four-five GHz chips five years from now. What do you need that much desktop processing power for?

The fashionable answer: broadband media applications. While I would relish the chance to view customized versions of Wallace and Gromit films at work, adding my own digital effects, I'm not sure that's going to prove a compelling argument to my boss for a PC upgrade: buy me a new PC, and I can create my own digital movies! Believe me, I'm going to try it, but I'm not hopeful.

What other uses are there for the new megachips? Well, as an old database guy, I could argue that PCs will need more processing power to handle database queries. However, it seems clear that more and more of that kind of processing is being done on servers: if you're going to connect to a database server with a Web browser, how much client-side processing power do you need?

So what's left (assuming you don't need to do massive 6-dimension Pivot Tables in Excel)? What is the Killer App going to be for the 1 GHz chip?

I think the answer could be intelligent agents. While I could spend a ton of time trying to tame my Inbox, I'd prefer to have my e-mail client take first crack at it. With a 1 GHz processor and 512 Megs of RAM (which will probably be standard in a couple of years), who knows what Outlook 2003 should be able to do for me:
  • Shopping
  • Translation
  • Transcription
  • Time management
  • Meeting preparation
  • Document tracking
  • Project management

Now you could argue whether or not all these functions are really part of the e-mail client, but remember, I was stating that the OS should be built around the e-mail client.

In any event, I remain convinced that e-mail will be one of the great unforeseen stories of the new decade—and you heard it first here at TechRepublic.

Bob

Catching up
Follow these links to catch up on the exchange between Bob Artner and Tom Ranalli:
Come up with your own Utopian E-mail Application and tell us what yours would do. Are there industry-specific things your Killer App would do? Start a discussion below or send us a note.�
Here is the third in a series of e-mail exchanges between Bob Artner and Tom Ranalli:

From: Bob Artner
To: Tom Ranalli
Sent: Sat. 01/06/2001 12:21 PM (EST)
Subject: The Future of E-mail

Tom: Don't get me wrong—I agree that there are valid reasons for limiting archiving of messages in e-mail systems—the way those systems are currently designed!

Let's put aside the legal issues for a moment, because lawyers depress me.

You're absolutely right that mail server performance and storage requirements are legitimate concerns for all organizations. I agree with you—well, I agree with you to the extent that anyone whose Exchange mailbox currently contains 140 megabytes can agree with you, without feeling guilty.

My point is that I'd like to see an operating system organized around an e-mail client, and not the other way around. If everyone agrees that e-mail is the Killer App, why don't we treat it as such?

Remember back a couple of years ago, when everyone was saying that the Web browser was the Killer App, and that Microsoft Office was dead? At the same time, people were predicting that stand-alone e-mail clients would wither and fade away, replaced by a Web browser interface to your e-mail server. Well, you don't hear that talk as much anymore—I guess most folks decided that Hotmail, while useful as a secondary e-mail service, isn't the last word in user friendliness. And yes, I have used Microsoft Web-based Exchange. While it's better than Hotmail and has its uses, I could never use it as a permanent replacement for my Outlook client.

What do I mean when I talk about building the OS around the mail client? Well, here is a short list of some of the things I'd like to see. (I know that some of these are available in bits and pieces, but not as one package):
  • Intelligent folders: I'd like my e-mail application to monitor the kind of messages I keep in my client and automatically deal with them in a way that goes way beyond the auto archive feature of most clients. In other words, perhaps I could configure my client so that it always shows me the last 30 days of messages. If I scroll past that date, it automatically goes to wherever the older stuff is archived and retrieves it, without me having to change folders. When it comes to attachments, I'd like the client (or the server) to be smart enough to realize that I've stored six different files called Monthly Sales Report in my Inbox and to automatically create a subfolder called Monthly Sales Report within my Reports folder and place all the attachments there. Further, after it's done that, it knows to automatically filter for that report and move it when it gets sent to me next month.
  • Attachment vs. Pointer: TechRepublic, like most companies, has people working in multiple offices and lots of folks working remotely out of their homes. Since not everyone has the same ease of access to the corporate network, people here often mail documents to multiple recipients, rather than posting a file in a single location and sending an e-mail pointing to the document's location. The e-mail server should be intelligent enough to know who has access to the internal network and who doesn't, and send only pointers to those who access, while sending the entire document to those who don't.
  • Virus Protection: Granted, most of the really troublesome viruses from the last two years were based upon the macro technologies of Microsoft Office, but that's really my point. If we designed the OS around the e-mail client, a lot of those vulnerabilities would never have been allowed in the first place.
  • New User Interface: If we built the operating system around the e-mail client, I'm not sure what the user interface would look like, but I'm pretty sure that it wouldn't employ the desktop metaphor. It would also give you a lot more capabilities from the client. I'll give you one idea. Consider the From field in your Inbox. When I right-click on the person's name, I should be able to get a lot more than I do from the Properties option (assuming this is a fellow employee). I should be able to get a bio and photo, the location of the person in the corporate org chart, and whatever personal information the person wishes. You should be able to see a network diagram and see where that particular person is set up. If that employee works for you, from the e-mail client you should have one-click access to personnel information such as review date, sick/vacation time, etc.

As I said, I know that some of this is available now, but none of it is seamless, and all of it is built upon rickety infrastructure. The basic question remains: How much better could our e-mail servers and clients be if we really treated them like the Killer Apps they are?
Bob



From: Tom Ranalli
To: Bob Artner
Sent: Sat. 01/06/2001 2:40 PM (PST)
Subject: Re: The Future of E-mail

Bob: You must've had a second cup of coffee this morning, because the features you describe in your "perfect" e-mail world—building the OS around the mail client—are very innovative and insightful. More on that in the moment, though, I have some business to attend to...

One hundred and forty megabytes used on your Exchange server! Your network administrator must be ready to hang you by your thumbs for that infraction! Do you realize that you're wasting valuable resources by consuming so much space? Don't you think about your coworkers? How about the company bottom line—do you want to spend all of your profits on more equipment to satisfy your never-ending hunger for storage?

OK, I got that out of my system (had to do it, Bob ... old habits die hard). Let's get back to your Killer App (because, like in A.A. Milne's “Winnie the Pooh”, this deserves title caps). I don't know about you, but I was severely disappointed that Web-based e-mail never reached the potential that we all imagined. The concept was simple enough—all you need is a browser and an Internet connection and voila—you have your e-mail. You could be at Granny's house in Idaho firing up the new eMachine you gave her for Christmas and taking a little time-out to check your work e-mail...what could be better than not having to schlep a laptop through airports and holiday crowds? Unfortunately, it is severely limited, and I doubt any serious business is using Web-based e-mail as their firm e-mail client.

The MS e-mail Web client had potential, but it never seems to have gotten past the initial "let's see if we can develop something quickly" stage. I remember predicting that THAT would be the start of what would ultimately replace traditional e-mail client software. Unfortunately it didn't—despite the fact that a great Java app could be written to basically duplicate all of the functions of the e-mail clients we know and love. C'est la vie.

Bob, you listed some great features for your Killer App, but I'd like to comment on the underlying reasons why these ideas make a whole lot of sense.

Why Intelligent Folders? For all of the rhetoric we've thrown around about "information overload," we've seen information only increase for each of us by orders of magnitude beyond what we could've imagined. Who has time (as you suggest) to search where to store file attachments? We're lucky we have the time to discern the e-mail wheat from the chaff, let alone mill the wheat and make bread! The more intelligent the folders, the better. Throw in those aspects of artificial intelligence you referred to by having the app learn what to do with the attachments, archived e-mails, what have you, and you'll make a lot of stressed out people VERY happy. (Do you hear this software developers?)
  • Attachments, Pointers, and Storage Management: You knew it was coming, Bob. Not only should the Killer App intelligently discern when to send a pointer versus an attachment to make communication as efficient as possible, but by only sending pointers when it makes sense, this would seriously reduce the amount of wasted space on your e-mail server. Remember your habit of saving e-mails with attached documents in your Exchange folders? Saving a pointer is like saving a penny—eventually they would add up to a million dollars, but who cares?
  • The Virus Threat: Here is where we disagree, Bob (though I don't fault you for wishing it). Knowing the persistence and misdirected intelligence of these virus writers, do you think that even with a new OS—whether or not the architecture is open—the same folks couldn't find a way to create a nasty virus? More likely, because of the tight integration between the app and the OS, viruses would be more dangerous in their effect and more prevalent than they are. Let's just make sure we build in more robust recovery methods.
  • GUI: I agree that the user interface should change, if for no other reason than the fact that the whole paradigm has changed! Your suggestion of putting more "goo" in GUI (that is, graphics) virtually eliminates the isolation one feels when using today's e-mail client interfaces ("What does John Doe look like over in the Poughkeepsie branch—I'm constantly e-mailing the guy?"). This model would naturally help draw people in a large organization closer and that can only be good for business. Not only would we feel closer to our fellow workers but closer to our data—data is so disparate and difficult to access at times that too much time is wasted looking for such things as personnel files and the like.

All in all, Bob, a great start. Can I be a beta site?
Tom

Here is the second in a series of e-mail exchanges between Artner and Ranalli:

From: Bob Artner
To: Tom Ranalli
Sent: Fri. 01/05/2001 11:40 AM (EST)
Subject: How Long Do You Keep E-mail?
Tom: It's interesting—lots of the people who responded to my original column suggested that I didn't think IT managers had to deal with office politics. You make a reference to office politics in your last reply.

It's a little tough for me to understand, since the whole point of my column was my belief that BCC e-mail messages carry the potential for greater political turmoil within an organization. As you note, office politics are part of any large organization.

I can't imagine what it's like in your case—300 lawyers! <g>

Here at TechRepublic, we find that whenever we publish articles about e-mail, people go crazy. I think that's because e-mail remains the Killer App for most IT professionals, and just about everyone is heavily dependent upon it. We all have our own little e-mail tricks—and our own horror stories.

Since we're never going to agree on the usefulness of the BCC option, let me switch topics.

How long do you keep back e-mail? As you know, there are a lot of reasons for not archiving e-mail that is more than a couple of months old. Among them:
  • It takes a lot of disk space
  • It slows down e-mail server performance
  • It opens you to legal vulnerability—witness what happened at the Microsoft trial

I understand all that. And yet, e-mail is just such a useful tool for organizing and distributing information. At TechRepublic, for example, new business ideas often get started by someone sending out an e-mail to a group of folks, outlining a possible project, and asking for comments. The give and take among team members via e-mail is a good way to refine the concept (or to quickly kill it, before we waste too much time on it). I'm reluctant to lose all the insight that went into those messages.

Or take attachments. While I know it ties up server space, I often find myself saving messages with attachments on the e-mail server rather than saving them locally. For example, every week I get an e-mail from Carmen in our Finance group, with a spreadsheet showing the latest traffic trends. I could save all of those locally and put them in a folder on my hard drive, or on my user directory on a network server. The same thing is true with the monthly revenue reports that Shawn sends at the beginning of each month.

However, I spend so much time with Outlook that it's actually easier for me to find those files by sorting either on the sender's name or the subject line. Why build a series of folders to store such files when there is a perfectly useful organizing tool in my e-mail client?

I guess what I would really like is to have my PC's operating system be built around my e-mail client.

So how long DO you keep your old e-mail?
Bob

From: Tom Ranalli
To: Bob Artner
Sent: Fri. 01/05/2001 10:25 PM (PST)
Subject: Re: How Long Do You Keep E-mail?

Bob: If there is someone who thinks that ANY manager can be insulated from the effects of office politics and be successful, they are living in a fantasy world. In fact, the more we IT managers become intricately involved with the business processes of our company, the more impossible will it be for us to remain cocooned in our IT bubble. A direct side effect of being a critical part of the core business is that we are being dragged, sometimes kicking and screaming, into office politics on a grand scale. It is necessary for our career survival (not to mention our sanity) to be as savvy as possible about the direction of the political winds and to use the tools at our disposal—e-mail and all of its variants—for discerning that. Yes, e-mail is the killer app for all of us.

Now, on to your rhetorical question. As an IT director, I certainly am concerned about retention—or should I say hoarding—of e-mail messages. You shouldn't dismiss so lightly those reasons that one shouldn't keep back e-mail. E-mail DOES consume a lot of disk space, and too many amassed e-mails DOES impact e-mail server performance. These aren't trivial concerns as IT is expected to be able to address both effectively without letting users be impacted. As for legal vulnerability, the threat is real and is a constant concern of senior management in many industries, the legal business not withstanding.

Despite these valid concerns, we are constantly fighting objections to running the clean-up agent every week, which cleans out deleted items and old inbox items (which can be moved to a safe folder at any time). Our users complain, saying they don't understand why it isn't normal to have 5,000 e-mails in their deleted items folder! If I hear "suppose I don't want to delete those e-mails after all" one more time as a response, I'll just lose it.

One should be able to retain certain e-mail for as long as reasonably necessary, but must be very selective about what falls into that category. Your example about business idea collaboration is a good one—but why retain it on the server? Why not archive the related folder to your local storage for later retrieval? Some would say just print it for later review, but I recommend saving a tree when I can (yes, Bob, even I fall prey to political correctness at times). Of course, some bright person will point out that such a solution just moves the problem somewhere else (not to mention, any storage medium including paper is subject to discovery in a trial).

Retaining e-mail with attachments? THAT wastes server space unnecessarily. Your reasoning for not doing that isn't on firm ground, Bob. Why not just keep a folder tree that is an analog of what you have as my e-mail tree? Attachments are then easily saved and found later, especially if you're smart enough to save the e-mail body in the folder, too. The problem is that if everyone in your organization does the same thing you do, you may have hundreds of copies of the same large file taking space up on the server. Can your company afford to buy another drive for your server every time it runs out of space?

Somebody has to guard the storage gates from you and your fellow Mongols, Bob.
Tom

Here is the first in the series of e-mail exchanges between Artner and Ranalli:

From: Bob Artner
To: Tom Ranalli
Sent: Wed 01/03/2001 5:53 PM (EST)
Subject: BCC discussion
Tom: Right off the bat, I have to thank you for one thing—I've never been called politically correct before!

As I'm sure you saw, my column advocating eliminating the BCC (blind carbon copy) e-mail option generated a ton of posts, most of them like yours, arguing that I was (to put it mildly) misguided.

I've got to tell you, I'm unrepentant—with one small exception. (The exception? Several folks pointed out that using BCC was a handy way to keep a customized mailing list that allowed the receivers to keep their e-mail addresses private. I don't do that, but I can see the usefulness of the technique.)

That caveat aside, I remain unpersuaded that BCC's utility outweighs the damage it can do to an organization.

Take the example you raised in your original post:

“BCCs are just the lazy man's way of keeping your boss in the loop on things without clueing anyone in about it.
“Why does one do that? Well, some people get intimidated/angry/concerned when the big cheese is explicitly informed of things. However, the boss needs to be aware of things at times by some fashion (verbally or otherwise). I find BCCing him much more convenient and expedient.”

Now I get a lot of this kind of e-mail, and send some of it, too. I prefer to use the FORWARD option, and I'll tell you why.

First, it allows me to send some additional information to the person I'm adding to the e-mail, explaining the situation and giving some context. I can't do that using a BCC.

Second, it reminds the reader that they received an e-mail containing information that might be sensitive. If you get as much e-mail as I do (and I'm sure you do), it's hard to remember what e-mail threads you are supposed to know about, and which ones you're not supposed to know about.

Has this ever happened to you: after you receive a BCC e-mail, you reference the message's contents to someone who wasn't supposed to know you had received it? Not fun.
Bob

From: Tom Ranalli
To: Bob Artner
Sent: Wed 01/03/2001 8:56 PM (PST)
Subject: RE: BCC discussion
Bob: There is a fatal flaw in your argument—aside from your being PLEASED that I suggested you are politically correct. If you like the idea that BCCs are useful for the customized mailing lists, then unless you have the secret for controlling human behavior, you can't have that one use without inviting all the others. What are you going to do, have users sign a declaration stating ''I shall only use the BCC power for good, not for evil"? Puh-leeeeze!

However, this isn't a discussion of behavioral science—or is it? If you forward something sensitive to someone, they are just as liable to forward it to the wrong person as when they are BCCed (and they are just as liable to let it slip in casual conversation!). Shall we discuss the recent ruckus caused by the well-publicized naughty e-mail forwarded to thousands throughout the world?

Let's face it, Bob—there has been both intentional and inadvertent communication of sensitive information since the beginning of time. Heck, whole governments have risen and fallen as a result of such disclosures. E-mail is scary to you because it is so real-time. Frankly, I'd rather see my exact written words than hear someone else's interpretation or misquoting of what I said come back to haunt me. I can't defend spoken words someone false[ly] attributes to me, but I certainly can stand by anything I've truly written.

I maintain using the BCC is not only a practical and expedient way to keep the appropriate people in the loop, but it is critical to one's political survival to do so. It may seem to some to be inherently dishonest to BCC others, but when you're dealing with 100-plus e-mails a day and managing the technology for a 300-person law firm at the same time—it is a lifesaver.

The bottom line is to use the BCC responsibly and judiciously—just like any other tool. BCC is for grown-ups, don't be afraid of it.
Tom

Editor's Picks