Alex Hopmann’s Blog

More 802.11n draft technology is coming on the market and the router market continues to be super confusing. The big
announcement this week is that Dell will shortly have an internal adapter for their laptops that uses the Broadcom chipset. This
suggests to me that if I’m going to get a new router (and I’m getting really unhappy with my Belkin Pre-N) I probably want
one of the the Broadcom ones.

The other interesting new development is that Netgear finally introduced a router that has the high-end networking (N) and
gigabit Ethernet ports. Since I use gigabit Ethernet at home its somewhat painful for me to be hooking in a 100mbit router since I
hang a few devices off the router’s wired ports. The tricky part is that Netgear has two versions of their 802-11N routers- one
use the Broadcom chipset and one uses the Marvell chipset. The bad news is the gigabit version is only available with the Marvell
chipset and eWeek has reported that it has more compatibility issues talking to other devices than the other chipsets.

All this is made much worse by my experience with the quality of router firmware. It tends to just be terrible, and the bleeding
edge models are even worse- I’ve seen horror stories of routers that show up and have such buggy firmware they barely work at all. It
really feels like some event will have to happen to shake up the home router market since in its current form it really isn’t serving
customers very well. It sure is nice that the prices are so low, but I’d pay an extra $100 easily for a router that actually works 100%.

Tip- you can tell what chipset the NetGear routers have by the last letter of the model name. The NetGear WNR834B is a Broadcom router, the
WNR834M and WNR854T are Marvell.

]]>

I admit that I like to watch my server logs in real-time. I know its a sickness,
and Im not really trying to do anything to get over it. When I built a web-server back
in 1996 when there was still a market for web-servers, our killer feature in Boulevard
was that it had a bunch of cool graphs so you could sit at the console like a vulture
and watch the blips come in as users accessed your site.

Now that Im running several web-sites again I of course wanted to have something
similar again. Partly it is related to my philosophy of build, get in front of real users,
and iterate. Having a really tight connection to whats happening, what activity is popular,
what gets lift is important. But I think it also goes beyond the practical into the psychology
of building software live on the internet. I tweak the UI and watch for the difference my
changes may make. For example I made some changes last week and there was a pretty noticeable
difference in how many people found their way past the find feeds UI to an actual calendar

I’ve built a little WinForms app that downloads updates from the server logs every 5 minutes, scans through the
data and does various custom analysis including popular pages (broken down to specific parts of the site), bot analysis,
and whatever else strikes my fancy. Its really easy to update and add new stuff, especially since its pretty brute-force
on the actual data analysis side.

Calendar Feed for the day- Computer Industry Conferences including
the upcoming Apple WWDC in two weeks. I’m sure I’m missing many interesting conferences- again this calendar is in wiki mode
so any registered user can add new events to it.

]]>

Kat has been trying to get me to go tackle the Iron Horse trail for some time now.
Sunday we decided to go for it. First we went to the bike store to get some good
bike-lamps and I also got some gloves and some padded bike pants. We headed up
the pass and left my car at the end of the trail and took Kat’s up to the
beginning.

The Iron Horse trail used to be a railroad track through the pass and has
been converted into a hiking and biking trail. Because it used to be for trains
the grade is pretty even the whole way through- its the easiest way to go
through the mountains without lots of up and down. The direction we went is
overall slightly downhill- enough to keep a nice pace while still being a nice
workout. The trail starts off with an almost 3 mile tunnel. The weather in the
area has been close to 100F, but the tunnel was probably around 50 degrees,
cooled by the melting snow run-off. One thing that surprised me a bit about the
tunnel is that it was completely straight- you could see this little speck of
light at the other end over two miles away.

Mostly the path was very nice for biking- dirt and some gravel. There were a
couple of bridge overpasses that were much thicker loose gravel and those were
fairly treacherous. There are also several side-hikes available along the way
that are worth checking out sometime in the future.

]]>

Today Eric and I did the hike to Rachel Lake. The Rachel Lake trail is about 60 miles up I-90.
The hike isn’t that long but the last mile is pretty steep. It starts off with a
little up bit and then a nice long fairly flat part. The last part is a good
work-out and you are rewarded at the end with one of the most beautiful lakes in
the Cascade Mountains. There were a few people at the lake even on a weekday,
but next time I’m tempted to bring a float for the lake like some of the other
hikers did.

]]>

In building CalendarData.com the way I have, I’m doing a few wacky things.
In some ways its an experiment about how the progress of technology changes the
way that someone can develop and launch a business. First of all, I’m not
following the old model of “develop an app for a year (or 6 years) in secret
until its all perfect and then launch it to an amazed public.” With the new
site, everything is pretty much out in the open. I’ve been building it for a
couple months without announcing it, but the site has actually been up and in
public for a bit. I’ve also been rolling out changes to the site a few times a
week, sometimes more- so far today the site has already had two updates, first
to enable wiki-like functionality where you can mark a calendar as “open” so
anyone can add and edit events, and the second update to make some UI
improvements, get rid of unnecessary clutter, etc.

As I do this, to be honest the site so far looks really bad. Its still too
cluttered, the usability is poor, it needs lots of design help, and more.
However there were several factors that prodded me towards putting it up even in
such an unbaked form. First of all, lots of the functionality is already there,
and I find it useful and hope some others might too. Second, it can take a long
time for the search engines to find you and index you and its probably a good
idea to get that process going as quickly as possible. A spider isn’t going to
care about the UI, so there is no reason to hold up getting that going. Finally
I’m already learning a lot from the visitors that arrive at the site. I could
have spent a few more months designing the perfect UI, and would have been
totally wrong about how real users would be interested in the site. Now, I need
to be careful to not assume that the trickle of current visitors are
representative of the (I hope) future mass audience, but tuning a UI in the
presence of real data is a dramatically different thing from doing it in the
abstract.

So that’s how it goes for now- I’ve turned on the wiki features, so in
addition to importing a feed from some existing service you can create brand-new
calendars, and/or go add events to existing ones. In the next couple of days
I’ll add the ability to put descriptions and links into events, be notified when
changes happen to calendars you have edited, and display stats on how much you
have contributed to the community. The next goal is to get a set of people who
have actually come to the site and contributed in some way. Stay tuned as we all
see how that works out…

Calendar Feed for the day- The Seattle Art Events Calendar. Please feel free
to add your own Seattle-area arts related events to this calendar.

]]>

Today I put online another update of one of the sites I’ve been developing, CalendarData.com.
CalendarData is an experiment with several modern web-phenomena. It is an
aggregation engine for feeds of Calendar data. It is a wiki for creating
easy-to-find guides for events related to various special interests. It also
features a combination of client-side engines and web-based UI. The whole thing
is still very rough but now that I’ve added the ability to add new calendars and
events right in the site I thought I’d mention it here.

I’ve been thinking a bit also about my approach to building these new sites.
There are plenty of areas to pursue, but for now I’ve decided to focus on areas
that scratch an itch for myself or my friends. This doesn’t mean that I won’t
try to make them as general as possible for a broad audience, but I figure if
I’m building something that I would use myself, the chance is bigger that it
will be useful for at least someone.

There are several other things also in development but those will need to
wait until later…

]]>

Art Brut appears to be coming back to the US this October- I just got tickets for the Seattle show. Hopefully I’m going to see them
in Chicago in two weeks also. I’ve been inspired to update the tag-line of this blog to “Modern art makes me want to rock out” which is a
pretty good way to sum up both what is so great about the band as well as my more general attitude
towards art.

]]>

Mason Lake is a relatively easy to reach hike just off Interstate 90 45 miles east of Seattle. Despite being fairly
close you are pretty much in the middle of the cascades and enter the
Alpine Lakes Wilderness Area during the hike.

Eric and I drove up there- we had an older hiking guide that warned that the trail was in terrible condition but it seemed fun anyway.
It turns out the trail has been totally re-done since the guide was published and is in great shape. The hike was pretty steep but a good
workout and the only glitch was that my heels started getting blisters- I think the climbing shoes had weakened them a bit over the weekend.

At the end of the trail (for us- the trail actually continues on much further) was beautiful Lake Mason and we stopped for a quick lunch.
Round-trip the hike took us a bit more than three hours but with the elevation gain I was pretty tired & sore afterwards.

]]>

Yesterday Kat and I flew out to Glacier National Park with Michael and Anh in Michael’s Saratoga. It was pretty amazing how
much gear we packed in his airplane. The flight was beautiful; we landed and had a nice BBQ dinner.

Today we got up very early (at least by my standards) and headed up to Logan Pass which is on the continental divide. From there
we hiked about 3-4 miles on the Highline Trail to “haystack” little peak that Michael had scoped out for climbing. This was my
first time really climbing with ropes and all outside. I’ve done it many times in the gym but outside is a very different thing. It looked
like the climb was going to be pretty easy though which is just how I like it.

The first pitch was a breeze and was good confidence building. After that though the third pitch had a really tricky first bit. I was pretty
wiped out when I got to the top. After everyone got to the top of that we were trying to figure out how far it was to the top- we could see
what looked like an easy bit, but it wasn’t clear if that was the top or whether there were going to be some other tricky spots. Part of the
appeal of this climb was that the other side was a very gentle grassy slope. So we decided to see if we could traverse out to find an easier
way up to the top.

I suspect this wasn’t a great idea. We roped for the traverses but they were still pretty difficult and slow and things weren’t getting
any better. Eventually we could see what looked like the grassy slope close ahead, but there were also two ravines in the way
and those can be really difficult to cross.

So we decided to down-climb. I found the belay-down much scarier that in the gym at first. At the gym you just get to the top of the
wall and you already have your butt hanging over nothing. I finally managed to get going and we all got down but it took a while. We were
all pretty tired and eager to get back to the car and our camp-site.

For dinner we were really roughing it- since we were car-camping we had huge steaks, asparagus, baked potatoes, butter, a couple nice
bottles of wine and to top of off chocolate fondue. Its tough but someone has to do it.

]]>

Last week Google introduced their new Checkout service for online payments. I was
pretty excited to investigate since an easy way to accept payments without having to fork over 50% to someone else would be a real
bonus for those of us trying to build new web-services, etc. The site was easy to sign-up for and taking advantage of the $10-off
on any $20 purchase from various vendors was a pretty good deal to get me to try it out. The API stuff looks pretty good
too.

The one catch is that their terms of service currently only support transactions that involve “tangible” goods. So selling
a digital service is not allowed for now. They seem to imply they are looking at supporting these digital services eventually
and I find it ironic that they don’t support the very kind of transaction that they have built their business on so far. I don’t
for a second think that its because they don’t want people competing with them- Google has never seemed like the organization that
would take a short-signed position like that rather than eagerly get everyone selling web-services across the web to be tied into
the Google infrastructure. Stay tuned…

]]>