The Next America

Ethernet over Powerline

If you're considering it, don't. Or at least know there's a 50/50 chance of returning it. Pretty compact house -- dismal signal one room away.  Ended up hardwiring CAT5 -- which is what I should have done all along. Second house in a row that I broke down in the end (though actually G with a newer laptop with integrated antenna would have solved the problem in house #1). The open-source Linksys replacement firmwares really help with signal with the ability to max out power.

Its somewhat amazing that I have three things under my TV with ethernet (PS2, Tivo, Media PC).

Job Tips

This is a list of tips for job searchers in high-tech fields I compiled based on my experiences working for a major employer in New England. I've seen a lot of people come in through the door in the past couple of years.

1) Get on Monster.com. Brutally reject companies/situations you're not interested in. If you like the company, but are dubious about the job check it out. Many respectable large companies use Monster far more than is appropriate. Make sure your buzzwords are appropriately called out in the resume -- don't get too technical or it may not pass the initial HR screen. Don't go overboard or you'll be tortured on the interview.

2) Craigslist. For my fellow Harvard alums, harvard-jobs and harvard-startups mailing lists. Spotty but sometimes gold.

3) Informal groups like Bar Camp, Pito Salas' occasional geek meetups, Providence Geeks meetings at AS220, Brown Forum for Entrepreneurs.

4) Temp-to-perm rarely works for big companies (but is great for others). Many big companies specifically prohibit permanent contractors or temp conversions for legal reasons. Temp benefits will always suck and rarely get better. Understand this up front.

5) Apply directly to jobs in the right field at large companies. They are probably using the recruiter who will fill the job you want. They may drop hints about other jobs to apply for. If someone you've talked to previously tells you about an opening, they may be indirectly trying to tell you to apply (see #6).

6) However annoying it is, apply to each position separately at large companies. They are moving to highly legalistic processes to prove sufficient applicants were considered and that applicants weren't informally screened out using discriminatory practices. We can't even accept resumes at job fairs anymore -- all we can do is hand out business cards and ask people to go online. This is having a devastating affect on recruit quality.

7) Stay away from major companies with weak financials. Even if they are hiring, the process will be slow and painful. And then you might get laid off.  On second thought, apply but don't hold your breath.

8) You can't fight the tide. The market has gone from gangbusters to tight to approaching gangbusters at the beginning of this year (those who didn't prepare for their 2007 hires got burnt). When times are tight, be happy at the job you have -- the entire process turns slow, crowded, and painful.

9) Look at major venture capital websites in the area. Find who they've funded. Haunt the funded company websites and try to network through #3. You'll meet prefunding high-risk people in the bunches through #3. Might find the perfect opportunity, but hard to be patient if you're unemployed. Better to stick with funded if you have no job.

10) Check specialty job boards. Like joelonsoftware.com or a specific new field you're good at. Once everyone finds them, they're useless. But in the beginning, there can be some spectacular finds. They're looking for a mindset and a community outside the norm, not the riffraff from craiglist or monster. Play the card if you fit.

11) Either hop immediately if its a bad job (and neglect on resume or straight up explain as bad fit) or wait it out 18-24 months.

12) Figure out a way for people to contact you. I recommend a prepaid cellphone. I've got a Virgin Mobile one in a desk drawer that I can activate for $10 in 10 minutes. I have seen some great potentials get dropped cold after too many conversations with a stay-at-home spouse while trying to connect to Ms. Right. Doing screening interviews in the parking lot isn't pretty, but it works. Though check reception first.

13) If you are off the market, pull your resume from Monster, etc. We are unlikely to try again when you're back on. Off the market means not returning emails/phone calls quickly and in a professional manner. A quick professional no is always understood.

14) If you are in the slightest doubt, do not quit your current job until after you have passed the drug test. If you are going for a job requiring a security clearance, talk to a friend about the process to see if you have any red flags that will hold things up (foreign spouse, extensive foreign travel, credit problems are a start). 75% of people can get an interim clearance in two weeks after computer background and credit checks are completed which is more than enough to get lots of real work done. Some people will take years for no discernable difference. It may be 2+ years before you're in the job you were hired for. Your new career will be much more likely to take off if you're part of the 75%. This warning does not apply to entry level (some delays are assumed in hiring).

15) Figure out if you're a big company or a small company person. Neither wants to hire the other type if possible. Do not mix the two resumes and get advice from someone knowledgable to rewrite your resume and retailor your interview if you're looking to switch. If you're a big company person, look up the top 5 or 10 employers in the area. Figure out which ones suck (work conditions, products, management, etc.). Watch the survivors' websites like a hawk. There are only a few actual choices for most big company people.

16) Make sure you get hired into the right group in large companies if career prospects are a major concern. There are some very tempting job writeups that are absolute dead ends. You want to be in big groups where you have a mainstream skill set. Army of one is great if you're happy to stay where you land.

Blackberry 7250

If you have a Verizon Wireless Blackberry 7250, you need to run...not walk...to the Verizon Wireless website to download the new 4.1 firmware. It completely resolves the difficulties with bluetooth headsets and these Blackberries. My first headset (an Allways) worked great, until it broke. The second (Scala 500) and third (Plantronics 640) had been disasters. After a week of messing with each, I had given up and gone back to a wired headset.

Then on Monday I was looking at some 7250 forums and discovered complaints about almost every single headset imaginable with the 7250. Meaning the problem was the 7250. Whose new firmware had come out last Friday. It actually took 3+ hours for the upgrade (1:45 for the upgrade, another 1:30+ for enterprise activation). But it was well worth it and can be mostly ignored during that timeframe.

I am now back in love with my 7250.

My DVR Broke

My Motorola 6412 DVR (Pioneer software) broke on Friday. It actually started dying on Wednesday, but much plugging/unplugging and rebooting got it to work once more. Friday morning it was dead (permanently downloading new firmware and occasionally rebooting itself). A frustrating call to Cox on Saturday. The rep swore that if I just waited it would fix itself. I got five days to wait for a technician to bring out a box and swap it out (5 minute job). A regular cable box can be swapped at a retail counter -- a DVR cannot.

1) $10/month sure beats Tivo/satellite when things like this happen

2) I'm glad I had my backup Tivo (normally use to tape SD programs)

3) What I missed most were the programs lost. Many could be recovered from Bittorrent. A few (including some top 20 shows) cannot.

4) The technician who came out was the same person who did my original install 3 years ago. An amazingly knowledgable guy who went beyond the book to do a very professional install. He prides himself on never installing a purely HD box. 90%+ he upsells to the extra HD DVR box in his truck. For $5/month an incredibly easy sell. I forgot to ask if he ever installs CableCards.

5) Modern entertainment systems are becoming increasingly hard for big companies to handle. "Can I plug it directly into the TV?" "No" "Why not?" "You won't get any sound -- my TV has no speakers".

Quicken Medical Expense Manager

Announced at Web 2.0. Genius. My father had two severe accidents and would sit down with his "folder" one or two days each week. Fiendishly complicated rules for certain categories of non-HMO people (in his case, there is a federal law prohibiting charging more than 115% of the medicare rate). Perfect example of a greenfield consumer-focused vertical.

First step: solve consumer problem to build relationships. Second step: establish automatic data transfer and sell it as a benefit to health plans or Fortune 500. Third step: use data gathered to compete health care providers.

In 2015, click a button, and local health care providers (backed by insurers) compete for your business like they compete for a mortgage. Your medical history is irrelevant -- its the claim history that counts.

My Telecoms Manifesto

I wrote this for work to address how our antediluvian telecoms group can make itself relevant to the business. There are a lot of opportunities once you look beyond telecoms to collaboration in general. While this touches on issues and capabilities specific to my company (Secure and Easy Collaboration is the name of an ongoing internal project in this space), it should be of general interest. Credit to Kevin Werbach, whose article "Using VOIP to Compete" in HBR triggered the question.

The Telecoms Manifesto

The first problem is the title. However, if I hadn't it called it this, no telecommunications person would have ever opened the document. And they are the people this document is directed to. Because unknowingly, unwittingly, they hold in their hands the key to a very valuable capability:

High-Availability Real-Time Collaboration

To be distinguished from Secure and Easy Collaboration. Which is pretty much accomplished. Which was focused on technology, on virtual spaces shaped and guided by technology, on better file cabinets, bulletin boards, and wall calendars, to shape our virtual lifestyle. It was designed to bring us along the path to how we should collaborate. Or at least how professionals who identify themselves as "IT" professionals think we should collaborate.

But we don't collaborate that way.

In the 1990s we were fascinated by email. We had mailing lists. People emailed jokes, or chain letters, or pleas for help. They started flamewars on Usenet. We thought email would change the world, that personal and professional relationships would flourish. And for a time they did. But now for the most part we don't. And it's not that our email boxes are so filled with spam that we cannot distinguish the missives from our loved ones, or those who might make us rich. It's because email is a very weak form of collaboration. Those missives rarely come.

It turned out the early adopters corresponded to the "cheap adopters". They used email not because it was particularly effective, but because it was easy/cheap/free/simple.

When I grew up in the 1970s, little kids didn't talk on the phone. You might say hello to Grandma for 60 seconds. Long distance was expensive ($2+ minute in 2005 money). Local calling was expensive (30c/minute for metered plans). Now they make specialized phones for little kids with preprogrammed buttons so they can call their relatives and friends (and hopefully no one else).

What changed? Price. Metered local gave way to unlimited local calling (which is largely responsible for why dialup internet exploded in the United States unlike most of the rest of the world). Deregulation gave way to long distance price wars. To 25c/minute with 10c nights (1996). To 15c minute (1998). To 12c minute (1999). To 5c minute (2001). To 2.6c minute VOIP with no local line charge (2005) or $200/year for unlimited nationwide calling (2005). I knew people in 1999 who routinely paid $200/month for their phone bill.

So to summarize...my home phone bill went from $75-90/month for unlimited local and 250 minutes of long distances to $17 (metered Vonage). With the difference I bought a cellphone (where long distance is meaningless), which I use to the extent that my wife and I never exceed the 500 minute/month cap (extra minutes = 3c). I picked up another line last week just to receive faxes, which are automatically sent to my email account. Price=free -- its a come on to get me to use the 2.6c/minute outgoing VOIP. My new business cards will cost more than my fax line.

Voice is a more powerful form of collaboration than email. Our fascination with email wasn't because it was better -- it was just cheaper. And given the choice, people prefer voice. Sure, email is great across timezones, or to maintain a paper trail, or to a large group. But who has a staff meeting over email? What sane manager would fire or promote or mentor or shape an employee over email?

In person collaboration is even more powerful than voice. But the expense is great. A team must accept productivity draining travel or loss of diversity and talent through colocation. We still fly home to see our relatives at Christmas.

Video seems a good compromise. It has never caught on at Raytheon, largely due to the $15,000 units and dedicated T1 lines we declare as the minimum acceptable. A suitable minimum for talking to the CEO, overkill for engineers and middle managers.

But IT narrowly focuses at electronic methods of collaboration when we seek to improve it. That if our eRoom was a little more powerful, somehow we would be collaborating more. No doubt we could manage our data better. Better workflow is desperately needed. But these are just efficiency tools to make our actual collaboration more efficient, to minimize the phone calls and shorten the meetings.

Collaboration is more than the core methods. There are the enablers: especially the phone and VTC bridges, which allow group collaboration. There are the electronic ones: the eRooms, the docushares, the shared drives where most of our data lives. What we call collaboration, but are really our electronic storage lockers, some nicely painted with a calendar on the door to track our comings and goings. There are the bolt-ons: the Netmeetings, Sametimes, and Webexes which are useless in themselves but allow powerpoints to be finished at the last minute and be kept away from the prying (and remembering) hands of the audience. And there is even lowly email.

So when an executive surveys his $4B chunk of a Fortune 125 company, where should he turn to ensure his collaboration landscape remains neatly groomed. To analyze how his employees are working. To see whether the 10 minutes wasted at the beginning of many meetings as Sametime burps can be recovered. To figure out whether any people who don't belong are on his conference call.

They shouldn't turn to network engineers. While collaboration requires networks and servers, an enterprise collaboration infrastructure could just as easily run on a wireless system provided by a third party. A good collaboration infrastructure is networking agnostic.

They shouldn't turn to web programmers or mainframe programmers. Day to day management of the collaboration needs of 10,000 or more individuals and maintenance of 99.999% reliability is not linked to designing business applications. Many perfectly good collaboration solutions have been employed by these groups, but never do they provide dial-tone reliability.

They shouldn't turn to desktop technicians, whose wizardry keeps working the equipment on 10,000 desks, because their focus is on the thousands rather than server farms, load balancing, and continuous monitoring.

Rather they should realize that collaboration is about conversations, not workflow. Who has managed our conversations for the last 75 years? Telecommunications.

The first step is changing their name. Because telecommunications is only a small (and clearly understood) segment in achieving High-Availability Real-Time Collaboration.

Renamed "Collaboration Services", the true scope of their new task begins to come into focus. They are responsible for the corporate directory, which is how we find each other to collaborate. They are responsible for collaboration spaces (formerly conference rooms), to ensure adequate spaces are available in each physical location. They are responsible for electronic, voice, and video bridges and spaces. They are responsible for webcasts. They are responsible for surveys. And they remain responsible for our telephones and supporting infrastructure.

And their legacy skillsets begin to show their relevance to modern problems.

Telephone Operators become Collaboration Enablers. Able to look up information for far-flung employees on mobile phones. Able to intervene into Sametimes when a meeting password is forgotten. Continually monitoring the phone bridge for conferences nearing capacity. Able to solve permissions problems in eRoom. Able to track down a conference room when a last-minute conflict arises.

Telecoms Engineers become Collaboration Engineers. Their focus spreads beyond voice to the entire range of collaboration. How can we ensure every employee can get a webcast? Do we need collaboration phonebooths to allow workers without enclosed offices to participate in low-level video collaboration?

Telecoms Managers become Collaboration Managers. Is the directory accurate? Why doesn't Sametime work better? Why don't we route the voice for our Sametimes over the data network?

Long-haul telecommunications, the T1s and T3s of the world, are given to networks where they belong. Most of the hundreds of thousands we pay for long-haul voice could be gone within a year. Networkings role is to provide bandwidth of appropriate quality for text, voice, video, and beyond. The demanding performance requirements (and easy visibility of failure) force them to properly tune and engineer the network, smoothing the pathway for both collaborative and informational.

So what would I ask Collaboration Services to do in 2006? Drive collaboration. Here's how:

1) Provide VOIP capability to all workers with laptops. Abandon USB phones in favor of the bluetooth headsets increasingly becoming common. Due to unreliability of the Avaya VOIP stack and service, go to a 3rd party for a commercially proven and frequently updated VOIP stack (SJPhone works great). Consider using them for SIP gateway services as well (having a "corporate" phone # is not nearly as important as telecoms thinks). Set up direct VOIP connectivity at the audio bridge for higher performance, no phone charges (VOIP to VOIP is free) and less trunking used.

2) Provide unified ring to people who travel extensively. This feature taps into the PBX to ring a person's desk phone, cellphone, and VOIP phone simultaneously.

3) Upgrade the audio bridge to world class. For historical and cultural reasons, this is our collaboration sweet spot. They're contaminated by "be-boops" and arbitrary feature selections made years ago, while hosts have no idea how precisely the behavior can be tailored. Our most senior executives don't use ours in favor of AT&T. Why? Because they've been burned in the past by capacity restrictions and service failures. Engineer 99.99% uptime. Dedicate personnel to real-time monitoring and uncap port restrictions on directors and above. Give them the killer feature that AT&T can't deliver -- real-time viewing of conference participants. Give all users a fully automated solution to set the parameters on their bridge accounts. Setup voice recording capability so that people unable to attend important meetings can replay them online.

4) Buy our own video bridge that supports ad-hoc conferencing of low-end VTC equipment. Install $2500 VTC suites for the office of every director. Enable web-based scheduling of our video bridge.

5) Link email lists and audio. Managers should be able to send a voicemail to every one of their employees just as easily as an email. With no additional list maintenance required.

6) Fix Sametime until it provides the 99.999% reliability they demand of Voice gear.

7) Provide access to "Collaboration Enablers" at every center, not just Massachusetts ones. Freeze hiring until value can be proven in this new paradigm.

8) Install a fax to email gateway so every group of employees can have their own fax #.

And here's a few glimpses of the future:
1) Automatically install the closest printers to a user when they connect either by wire or wirelessly. Show them a map with emergency egress instructions, facilities, coffee, etc.

2) Automatically degrade conferences from video to voice to one-way voice (with IM backchannel) if there is a sudden loss of bandwidth on the network.

3) Sit down at any phone in the company, type in your directory services id and password, and receive all phone calls at that location. Enter a conference room and the screensaver on the plasma screen shows you the number of voicemail messages in your box by reading your badge with a proximity sensor.

In the 1970s, telecommunications was the experts in High-Availability Real-Time Collaboration. Then they gradually lost the bubble due to the flashy technologies of the Internet. Instead of declaring victory for the network and abandoning telecoms, we should be taking advantage of telecoms' historic skill sets to drive us forward.

The Famous Ipod Interface

So my wife used her Ipod mini yesterday for the first time.

I expected resounding praise. Or complaints about lack of selection (I haven't moved over most of her music since its in FLAC or OGG). Instead there was a resounding flurry of UI complaints:

"How do I turn it off?"

"The volume kept drifting up and down as I run."

"Can't create my own playlist" (she has running-specific music)

"It's impossible to navigate through the menus"

She's coming from the Rio Karma which has a "more button" approach. But I found it startling that one of the alleged great user interfaces of all times was so strongly panned on first use. It is beautiful, though.

Bizarre Ipod Setup Problem

So I get my wife's new Ipod Mini (green). And I go to set it up, expecting the legendary Apple experience.

It turned out to be excruciating. Install the software. Plug it in when requested. USB picks it up. Ipod begins formatting and never ends. It starts flashing between remove and do not remove every second or so. Resetting the Ipod doesn't help. Reflashing the software fails. Manually formatting fails (even though it is seen as a drive).

An hour or two of troubleshooting later (interspersed with several hours of doing other things while occasionally checking on it, I discover the problem. As part of the initial beta testing of Windows XP, I decided to do something different. My hard drive is on F:, not C:. So the first removeable media device installed is mounted as c:. This causes some problems with Windows 95 era software that expects a fixed path (which is why it was good for beta testing). Once I switched the Ipod to a normal drive letter, it worked like a champ. But there is no mention of this problem on any Apple website or third-party forum that I could find. It appears that there is internal logic in the software that will not allow formatting of c: but does not provide an error message. In the "old days" (and still for most factory installations) refusing to format c: made perfect sense. But it caused a heck of a problem in my case.

Now I get to transcode a ton of OGGs, FLACs, etc. to AAC. Delightful. I have yet to find software that will perform a very simple task -- automatically transcode and maintain synchronization during file movements, retags, etc. from FLAC to OGG (RIO Karma) and MP3.

VoIP Wireless Phones

The new Vonage/UTStarcom phone is interesting. It actually looks like something we could consider giving executives (actually this could be very big for telecommuters -- current solutions like software phones are awkward at best).

My big worry is how it works with "non-open" wireless points. Probably supports WEP and encryption. But how does it work for the Hampton Inn style one-time password situation (password is on the keycard jacket when you check in)?

Also, there is a clear market for a more advanced home router that would allow you to segment traffic. It would offer unencrypted public access (with logging) but not to the "private"=wired part of the network. If you need the private part, use VPN. I would be more than happy to share my logged access point with the world if I didn't have to totally revamp my open shares.

HP DesignJet 500ps Problems

I spent several hours today trying to troubleshoot the plotter for our facility. It's a nice 42" plotter -- great for the posters we need for a corporate event this weekend. Unfortunately, our media person got called away suddenly due to a personal crisis. So it ended up in my lap -- and I'd never printed to it. Worked great; but the pictures were cut off at 14" or so horizontally. Nice sharp demarcation between 1/3 of a great poster and 2/3 of a blank page. Every paper size in the book produced the same results. Reinstalled drivers, bypassed print servers, etc.

Lesson Learned: The basics of troubleshooting apply even for a $4000 printer with a 17 day uptime. Cycle power. Worked like a charm afterwards. And HP/google was absolutely no help.