Sunday, December 22, 2002

Hmm.  This makes me wish my e-mail system was a self-organizing outliner.  The biggest problem with Outlook is that it makes it difficult to categorize/archive mail and make rules.  Here's some of the attributes I would like to see: Top level nodes would develop around people or type (mail-list).  Contact info for people available via v-card would be located in an icon next to their name.  This eliminates the need for a separate contacts list. Nodes would combine inbound and outbound e-mail to those people. Dragging an e-mail into an outline would co-mingle multiple addresses with a single individual.  Inbox flow would remain the same but would provide indication of whether the e-mail was already categorized to an outline node.  That would allow me to read it and discard it (knowing that it being kept in archive). I could drag people into meta top level nodes to group them by company, organization, or mail list.   Meta nodes would aggregate contact information for individuals attached to the group (this would make it easy to click one button to e-mail the entire group). E-mails that contain multiple individuals would be included in the nodes for each of the individuals listed. A list of new people would be included in my outline automatically.  Spam would work like this in the inbox (I am more than a little wary of spam filters since they can screen out real mail by accident -- so a rules based approach that I could actively manage would be much better): All likely spam based on key word combos would be flagged as potential spam.  Two spam deletes are possible.  One for a soft delete (all e-mails from this specific address are to be blocked in the future) and a hard delete (all e-mails from this domain are to be blocked in the future).  If I delete a person as a spammer, their node in the outline goes away. I could click to view the inbox as a list of people that are known to me.  This would allow me to focus on the good stuff first. I could click to view the inbox by top level nodes (ie. in my case it would include: customers, partners, friends, UserLand, mail-list, etc.) I could quickly delete all e-mails in my inbox that have been categorized with a single click (deleting an e-mail from a node in the outline would delete it from the archive). E-mail views would work like this: I could view a thread of interactions with a person as a fleshed out outline. I could view a thread of interactions with a person as a weblog.  I could view a thread of  interactions with a meta node as either an outline or a weblog. Duplicate e-mails (like those sent to a group), would be filtered out of the weblog view of a meta node. E-mail publishing would work like this: All e-mails or e-mail threads could be be published to a weblog as either a story or as a annotated post via a one-click icon next to each node. If published as a story (basically a new page), the page would be created automatically and a link inserted into the edit box for annotation. I could publish a complete thread to my weblog as a category weblog. I could send an e-mail that contained an OPML file that contained interactions with an individual or group by clicking on a send icon next to the top level node.  This allows me to share the effort I put into categorization of my e-mail. I could make certain nodes and items private so that when I published a top level node, every section except the private areas would be published. Search would work like this: A Google keyword search with Google style output (like Find and Zoe). Results would include a link to the OPML thread for that e-mail. Attachments like PDFs would be converted into HTML for easy viewing (as an option). For all the people that have 10's of thousands of e-mails in their inbox (I know lots of people that do this), self-categorization would be a God-send.  For those that spend the time to prune/organize their categories and build rules, this would make it simple.  The old e-mail paradigm was OK when I got only 25-50 e-mails a day.  Now that I get 400 or more a day, I need some help staying organized. [John Robb's Radio Weblog] 7:22:58 PM    comment []

One last feature item.  If we had a system like the one below, why not include the option to author e-mails as an outline?  It would make it easier to organize thoughts and annotate replies to e-mails received in outline form.   Imagine a lawyer or doctor working with a tool like this.  Drag and drop some outlined case law or medical text info, annotate it, and send it as OPML.  It also offers up some interesting possibilities in automated mergers of annotations from several sources into a single outline (for an RFC, class notes, etc.). [John Robb's Radio Weblog] 7:20:52 PM    comment []

Top 30 IC makers for 2002 Silicon Strategies provides the lay of the land in the semiconductor market. A lot of good background info in this article. Here's your list of players: Top 30 Chip Suppliers in 2002 2002 rank 2001 rank Supplier 2002 sales 2001 sales % change 1 1 Intel $23.47 billion $23.54 billion -0.3% 2 4 Samsung $9.18 billion $6.14 billion 49.5% 3 3 STMicro $6.31 billion $6.36 billion -0.9% 4 5 TI $6.20 billion $6.05 billion 2.5% 5 2 Toshiba $6.19 billion $6.54 billion -5.5% 6 8 Infineon $5.36 billion $4.56 billion 17.5% 7 6 NEC $5.26 billion $5.30 billion -0.8% 8 7 Motorola $4.73 billion $4.83 billion -2.0% 9 9 Philips $4.36 billion $4.41 billion -1.1% 10 10 Hitachi $4.05 billion $4.24 billion -4.6% 11 11 Mitsubishi $3.62 billion $3.87 billion -6.4% 12 14 IBM $3.39 billion $3.56 billion -4.7% 13 16 Matsushita $3.28 billion $3.01 billion 9.1% 14 13 Fujitsu $3.24 billion $3.73 billion -13.3% 15 18 Micron $3.22 billion $2.45 billion 31.2% 16 12 AMD $2.61 billion $3.79 billion -31.2% 17 19 Hynix $2.57 billion $2.34 billion 8.3% 18 17 Sony $2.50 billion $2.47 billion 1.1% 19 21 Rohm $2.39 billion $2.21 billion 8.3% 20 20 Sharp $2.36 billion $3.36 billion 0% 21 22 Sanyo $2.10 billion $2.03 billion 3.6% 22 15 Agere $2.03 billion $3.14 billion -35.5% 23 23 Analog Devices $1.94 billion $1.93 billion 0.6% 24 28 Qualcomm $1.85 billion $1.39 billion 32.8% 25 30 nVidia $1.80 billion $1.29 billion 39.6% 26 24 Agilent $1.60 billion $1.53 billion -3.2% 27 26 National $1.57 billion $1.51 billion 4.1% 28 25 LSI Logic $1.51 billion $1.56 billion -3.5% 29 29 Fairchild $1.35 billion $1.34 billion 1.1% 30 27 Atmel $1.21 billion $1.48 billion -18.5% -- -- Other suppliers $34.14 billion $33.95 billion -0.6% -- -- TOTAL $155.35 billion $153.06 billion 1.5% Source: iSuppli Corp. [RatcliffeBlog: Business, Technology & Investing] 6:04:24 PM    comment []

"P2P" streaming The End System Multicast (ESM) system developed at Carnegie Mellon University sounds like a promising approach to delivering streaming video from the network's edge. Each client takes on multicast processing rather than forcing the process back into the network, specifically routers. In principle, this should allow packets to flow between clients -- peer-to-peer -- rather than through any bottleneck on the Net (like a router). The problem will be in network performance overall, because traffic can build up. The ESM team writes: "End System Multicast introduces duplicate packets on physical links, and incurs larger end-to-end delay than IP Multicast." This is definitely something to watch, because a pure edge-multicasting system can enable self-organizing networks of "broadcasters." Will major content providers adopt it? Not likely, to put it mildly. However, for the rest of us, it may open new channels for access to all sorts of information in audio and video form -- speeches, demonstrations, porn.... [RatcliffeBlog: Business, Technology & Investing] 5:56:44 PM    comment []

Financial firms align on messaging and applied presence Credit Suisse First Boston, Goldman Sachs, JPMorgan Chase, Lehman Brothers, Merrill Lynch, Morgan Stanley, Salomon Smith Barney, UBS Warburg and Communicator Inc. are building what they hope will become an IETF standard for initiating secure messaging sessions, messaging and applied presence (MAP). Using a company the firms own together, SecuritiesHub, they've launched a MAP working group with these goals: Foremost among MAP's goals will be efforts to foster the development of a shared vision for the application of presence within messaging and other applications. MAP will do this by collaborating with multiple industry consortia and actively supporting the work of the Internet Engineering Task Force (IETF) and other formal standards bodies to bring best practices together from across industries. It's not clear how this proposed working group will interact with the IETF's Instant Messaging and Presence Protocol, Presence and Instant Messaging Protocol, Session Initiation Protocol for Instant Messaging and Presence Leveraging Extensions, and Extensible Messaging and Presence Protocol working groups. It's not even clear why the IETF needs another working group in this general area, except for the political leverage it gives the banking industry in the consummation of a final standard. Who says engineering is apolitical? [RatcliffeBlog: Business, Technology & Investing] 5:55:28 PM    comment []

Cable: No more strong subscriber growth In a retrospective article about the pains of 2002, Cableworld reports: The days of strong basic subscriber expansion--the fuel that drove the cable engine for 50 years--can no longer be counted on. Kagan World Media, which is owned by Cable World parent Primedia Inc., figures that cable penetration will slip from 65% in 2002 to 58% by 2012. Few MSOs can claim they haven't lost customers. AT&T Broadband lost 525,000 of its 13 million subscribers through the third quarter. Charter lost 4% of its customers in the 12 months between Sept. '01 and Sept. '02, and Adelphia's sub numbers slipped 1.4% in the first nine months of this year. The few operators that didn't lose subs--Comcast Corp., Cox Communications and Time Warner Cable--saw only modest gains or flat growth. While cable was losing a half million customers in the second and third quarters, DirecTV and EchoStar Communications (they of the year's biggest failed merger) added more than 1 million new subscribers, according to Leichtman Research Group. In a survey of 1,250 homes capable of receiving cable service, Leichtman found that 69% of DBS customers who had subscribed to satellite service for less than a year had switched from cable. Leichtman also found that 32% of current cable customers have looked at getting a dish, and 10% said they'll likely cancel their cable service in the next six months. According to the article, 10 percent of cable customers are ready to cancel their subscriptions in the next six months while direct-broadcast satellite experiences 20 percent churn every year. The clear message here is that modes of delivery of information and entertainment are in flux. There are many, many opportunities for upstarts and for changes in the balance of power between cable MSO, telcos and new players. [RatcliffeBlog: Business, Technology & Investing] 5:52:27 PM    comment []

Half of top Web sites are from local newspapers Neilsen/NetRatings found that half of the top 20 Web sites in November are run by newspapers. Local news, whether it geographically local or topically local (such as about finance or technology), is still the most engaging thing people go looking for each day. While blogging covers many of these topics, the core research and reporting process paid for by newspapers will not be replaced by blogging (or any "decentralized" approach to news) until there is a solid economic foundation for paying for the creation of original content by individual contributors (that is, not employees and not freelancers who need to have assignments made and facts checked). [RatcliffeBlog: Business, Technology & Investing] 5:49:44 PM    comment []