Week three had some good workouts and social running.

• 12 hours, 9 minutes all training

• 23.9 miles running

• 3,268 ft D+ running

Yoga on Monday, as usual. Hill sprints Tuesday on a moderately steep dirt ramp at Pineridge Open Space. Elliptical spin and sets of back squats on Thursday. A long easy bike ride Friday.

Saturday, yesterday, I went for a long run in the snow at Horsetooth Mountain with friends. I kept it super easy until the last mile and a half, when I jumped onto a train of younger, faster runners who were coming down from a different route.

Today my legs felt great, and I took a new pair of NNormal Kjerags to the dry trails at Maxwell Open Space for a test drive. I think these could be my Quad Rock shoes. They're very light (my lightest trail shoes by 50 grams per shoe) and fit me well. I felt secure going quickly on rocky technical sections of the trails, where I did not in the Tecton X 3 that I wore last week.

My left hamstring feels 100 percent today, so I'm looking forward to two full sessions of hill sprints next week. I had some knee swelling early in the week, but it subsided before the week. My only pain point at the moment is an arthritis flare-up in my left shoulder. I'm taking ibuprofen at bedtime and hoping that it doesn't spread to my knees.

The first two weeks of my Quad Rock training program went by quickly. My running mileage is still low, as I'm emphasizing power and speed, with generous recovery, and biking, treadmill, or elliptical chuggingfor easy aerobic base building. The quality of my running workouts has been good, in a way that doesn't show up in these numbers.

In week one, I did hill sprint sessions on Tuesday and Thursday on a block of Wallenberg Drive with a 3-4% incline. Running up at nearly maximum effort took about 25 seconds. This was not an aerobic workout in any way. I was focused solely on power and turnover.

• 11 hours, 7 minutes all training

• 26.4 miles running

• 1,545 ft D+ running

Week two was complicated by cold and snowy weather. I curtailed my hill sprints and did less dynamic workouts. Wednesday I did a two hour run at Lory State Park in a new pair of HOKA Tecton X 3 shoes. These were fun, and as propulsive as hyped on non-technical stretches of trail. I found them to be just barely stable on technical descents, however. The heel had a tendency to tip over. I'll give them another chance, for sure, but they are not an immediate choice for Quad Rock in May.

• 12 hours, 41 minutes all training

• 21.7 miles running

• 2.464 ft D+ running

In October 2025, I started daily mobility and core strength sessions based on a post by Joe Uhan. I do a little Vinyasa flow, spinal twisting, some diagonal chops, and then push-ups, crunches or dead bugs, and eccentric heel drops and calf raises. I'm still doing this morning routine about six times a week, which accounts for almost three hours of my weekly training time. I've come from being able to do zero respectable pushups to being able to do sets of ten impeccable pushups. I'm a big person, 90 kilograms, so body weight pushups aren't super easy. It's been fun to make progress on basic strength and fitness.

I've made my long range plan for 2026. It's different from my plans from 2023-2025, which were about recovering from injuries and trying to finish a 100 mile race. I succeeded in the former, but not in the latter.

In 2026 I am trying to be more intentional about training for speed. I'm going to do some workouts that develop power and neuromuscular adaptations, without any concern for building aerobic capacity during those workouts. I'm going to do more strides during longer runs and sign up for some shorter 5-10K races. I'm going to remind myself that running briefly, but regularly, at 100 percent can translate to running faster at 80-85 percent during a longer run.

That's my process goal. If I'm fortunate, I will reverse my recent slowing trend, and develop some relaxed, comfortable speed. Maybe I'll get a remarkable time in a race this year. Honestly, after three DNFs in a row (Bear 100 in 2023, Never Summer 100K and Bear 100 in 2025), just finishing would be nice. I haven't crossed a finish line since July 2023. It would be lovely to break this streak.

Like last year, I am mapping out my long range plan using a spreadsheet. I have signed up for two "A" races: Quad Rock 25-mile in May, and Run Rabbit Run 50-mile in September. Quad Rock I know very well. Run Rabbit Run, which starts at the base of Steamboat Mountain and goes up into the Park Range and back. It will be a new one for me.

Each race gets its own 16 weeks of training, organized into 4 blocks of 4 weeks. In the first block I focus on power and speed. The second is about longer hard intervals. The third is dedicated to longer zone 2 efforts. The fourth block is race-specific training: lots of time on hilly trails developing muscular endurance.

As part of my build up to RRR, I'd also like to do a classic Colorado alpine loop in August, like 4 Passes in the Maroon Bells Wilderness, or Pawnee-Buchananan in the Indian Peaks Range.

I've also marked down time for volunteering this year. I expect to help out at Never Summer and have signed up to be a volunteer at Hardrock. I'm also interested in crewing friends at races this year. Hit me up if you need an experienced hand, whether as a pacer or aid station crew.

Version 1.5.0 of your favorite Python library for reading and writing classic GIS raster data is on PyPI now. Since Jan 5, in fact.

Among other new features, this version adds support for 16-bit floating point raster data, and HTTP cache control. Please See the release notes for a full list of bug fixes, new features, and other changes.

Once again, major credit goes to Alan Snow for managing this release. Thanks, Alan!

Hello, my name is Sean Gillies, and this is my blog. I write about running, cooking and eating, gardening, travel, family, programming, Python, API design, geography, geographic data formats and protocols, open source, and internet standards. Mostly running and local geography. Fort Collins, Colorado, is my home. I work at TileDB, which sells a multimodal data platform for genomics and precision medicine. I appreciate emailed comments on my posts. You can find my address in the "about" page linked at the top of this page. Happy New Year!

Running in 2025 started out badly, but I hung in there, rode my bike and chugged on the elliptical machine when I couldn't run, did a lot of physical therapy, and finally got into good enough shape that I could plausibly try the Bear 100 mile race again.

I had little margin for error at the Bear, and misplayed my hand. I went out too fast and suffered for it after mile 35. I did manage to battle on for another 18 hours and 40 miles, and reached a new personal distance best. The best part of the whole event was the road trip with Ruthie, my crew chief, and staying with my aunt in Cache Valley before and after the run. And the fall colors. There were a lot of good parts. Crossing the finish line, sadly, was not one of them.

I did finally finish an ultra-distance run in November, a 32 mile trip from my home to the summit of Horsetooth Mountain, around the reservoir, and back. My house is behind my crazy hair in the photo below.

In July I got to spend a weekend with my friends David and Marin at Never Summer, a great time. After the Bear I ran regularly with my local friend Dana as he was ramping up his running, and paced him during his DIY marathon. I did a day of trail work with local runners in May, and volunteering at Black Squirrel and Blue Sky Trail Marathon in September and October. I train alone, mostly, but I really do like to spend time outside with other runners.

My running numbers for the year:

• 965 miles

• 144,864 ft D+

• 231 hours

That's less than two-thirds of my running volume for a good year. Things are looking up, however. I've been running and biking consistently after the Bear and am in good shape. I'm eager to get the 2026 season started. More about that soon.

Rasterio 1.4.4 is on PyPI now. The first release in a year and ten days. If you're using Rasterio in 2025, shoot some thanks to Alan Snow, aka @snowman2 on Github. He's leading the effort to get Rasterio caught up to recent GDAL and Python changes, and it's not a cakewalk!

Release notes are here: https://github.com/rasterio/rasterio/releases/tag/1.4.4. That's a lot of bug fixes.

2025 was another good year for gardening. We had only a tiny bit of hail damage in May and June, and a lengthy period of warm fall temperatures. I had the usual great green bean harvest, plentiful cucumbers and melons, and exceptional crops of peppers and tomatoes.

The difference-maker for me, this year, was commercial fertilizer. I'm a zealous composter, and we funnel our kitchen scraps, coffee grounds, garden debris, fall leaves, and such into our vegetable beds. And yet the productivity of some of our crops has declined. Tomatoes, in particular. I'd almost given up on tomatoes. This summer I used a modest amount of Dr. Earth fertilizer in my tomato and pepper beds and was rewarded with a bumper crop from just six San Marzano plants that my neighbor gave me. In September I put 18 pint-size freezer bags of 6-7 peeled plum tomatoes in my chest freezer along with eight pint containers of tomato sauce. Paul Bertolli's recipe, with some tomato leaves and stems for an extra summery kick.

Massive quantities of excess tomatoes is my big takeaway from this gardening season. So many curries and stews are going to be based on this surplus over the next six months.

I've got good news for people who love news about Rasterio, the Python package for reading and writing classic GIS raster data. Alan Snow is the release manager for 1.4.4 and 1.5.0 and has shepherded a release candidate with 40 wheels and one source distribution onto the Python Package Index: https://pypi.org/project/rasterio/1.4.4rc0/. The release notes are here.

Please try these out and let us know if they work as expected.

Some things went very well at the Bear 100, or in my case, the Bear 75.