One way to stop the Stewart's expansion and six ways to make it better

The new Stewart's store will be sited almost exactly in the footprint of the yellow house on the far right.

Stewart's plan to expand at Green Street and Fairview Avenue has been moving forward. We will see the latest version at the next meeting of the Hudson Planning Board tomorrow at 6:00PM. Stewart's ambitions were greased last year by the enactment of Local Law No. 5. Prior to it, Stewart's, as a nonconforming use in a residential neighborhood, was forbidden to expand. But with the law at its back and now two adjacent residential properties in hand, it is poised to build a vastly enlarged gas station and convenience store on the site.

I've been looking for a hole in LL#5, as I believe that good urban planning should facilitate the development of locally-owned businesses, not externally-owned businesses that remove local money. Not to mention that this will be an unattractive, suburban project inserted into a semi-urban site. Several times I thought I had found a loophole, only to be disabused on my next reading. The only obstacle to Stewart's ambitions I see is for someone to sue the city for "spot zoning," the largely illegal practice of preferentially zoning a specific property. This case seems to tiptoe very close to, if not over, that line.

Until that happens, I am heeding the frequent exhortations of Walter Chatham, the Planning Board's head, to "not let perfect be the enemy of the good." This exhortation has been embraced by Stewart's own representative, who echoed Chatham's axiom to him in initially engaging the board.

Stewart's proposed site plan

In this spirit, I took on a brief study of the site. First, let's look at Stewart's current proposal (right). It is a little hard to read, given that the proposed scheme is overlaid on the existing conditions. But if you know the site you can discern the major pieces. The existing gas pumps and canopy are at the lower left corner, at a 45 degree angle. The existing building, to be demolished, is tucked partly under the proposed canopy. Once you spot the two existing structures, you will realize how much larger the new site will be—double the size of the existing one.

My proposal looks superficially similar, as I could not move the basic pieces very much. But some key tweaks will go a long way toward creating a project that will look like it belongs where it was built. Here is my proposed site plan:

My proposed site plan









The following explanations are keyed to the plan.

1. New landscaped traffic island. The existing intersection is big, open, and ugly, with lots of asphalt, and the new Stewart's will bring even more paved openness. I propose a traffic island between the existing thru-travel and dedicated right-turn lanes on Fairview to provide softening and better spatial definition. Ideally, the island would be larger than what I drew, with the lower left corner (as viewed on the drawing) squared off to create a more triangular shape. However, trucks turning left from Green onto Fairview need the radius shown.

2.  Revised curb radius at transition of Fairview to Green. This adjustment will narrow this portion of Green Street NW-bound from two lanes to one. The space gained will become a planting buffer that reduces the scale of the intersection and screens views of the gas pumps. Incidentally, there is an existing bus stop on Green Street at the proposed curb cut location. If this is or will be an active stop, it will need to be moved a little farther up the street (for either Stewart's or my scheme).

3. New pedestrian crosswalks, signals, and signage. I've gone back and forth on the best crosswalk locations. Pedestrians usually should cross at corners, but this would force two of the crossings to traverse the traffic island and right-turn lane, which seems awkward and would reduce planting space on the island. So I have shifted them away from the corner to allow pedestrians to cross the streets perpendicularly.

4. Geometry tweaks. It's hard to tell on the Stewart's plan above, but the proposed gas pump canopy is skewed a few degrees off the geometry of the houses on Green Street. I suspect the proposed building is similarly skewed. That this is hard to discern might make it seem it can be ignored, but this is precisely the kind of thing that causes a building to subliminally convey, in real life, that it is alien to its surroundings. So let me put a finer point on things:

a. Set the gas pump canopy and the Green Street facade of the new Stewart's building exactly parallel to the facades of the existing houses on Green Street. Ditto for all parking spaces, interior curb lines, etc.

b. Set the Fairview Avenue facade of the new Stewart's building exactly parallel to the facades of the existing houses on Fairview Avenue. I suspect that Fairview and Green aren't exactly perpendicular, in which case the Stewart's building would not be precisely square/orthogonal. C'est la vie.  

Top: Stewart's suburban prototype. Above: an "urban" Stewart's in Troy, NY.

5. A two-story, flat-roofed Stewart's building. Stewart's initially proposed its ubiquitous, low-slung, suburban-style prototype (top right). Negative reactions at an earlier planning board meeting led Stewart's to suggest a version of its flat-roofed Troy store (right bottom). This is indeed more urbanistically appealing, although the second floor isn't really a second floor and the entry facing the sidewalk isn't really an entry. In defense of the latter, two entries are a security nightmare for a business that has only one or two workers on the premises. If you're Stewart's, you're going to have only one operating entry and it's going to face the parking lot. But if you're an urbanist, you want the entry on the sidewalk. With this in mind, I propose the following adaptation of the Troy prototype.

My proposal, view from Fairview Avenue



The major change is a single corner entry in lieu of two separate entries or one entry facing the gas pumps. This would give the store a reasonably urban sidewalk presence while allowing it to conveniently serve customers approaching from the parking lot side. Stewart's already has corner-entry stores, including one a couple miles south of Hudson at Routes 9G and 23, so this is not at all a stretch.

Schuylerville shop, 1946

A modest stretch is a walk-up window for ice cream service. But ice cream is in Stewart's roots, and so is engaging communities in a personal way. A gesture like this will go a long way toward making this store look like it belongs where it is built, and will help turn the sidewalk into a truly friendly space.

A greater stretch, which in an ideal world would not take this project to the breaking point: how about a real second floor with an active dwelling unit or commercial space?

If I had time, I would study the Fairview Avenue setbacks a little more. On the one hand, I am inclined to match the setback of the main building and entry "porch" to the houses and porches a couple doors away on the same side of the street. However, siting the Stewart's slightly closer to Fairview than the existing houses helps block their view of the ugly gas pump canopy, and helps give Stewart's an adequate footprint to do business.

6. New, mature trees along Fairview Avenue and Green Street. I believe I have heard that Stewart's will be making a contribution to the city to maintain the Green/Fairview intersection. I suspect that any such monies will be meager, and will go into the city's general fund—i.e., they will disappear. Instead, there needs to be a substantial, dedicated, up-front payment by Stewart's to cover the complete planting and landscaping of the site and intersection (trees, traffic island, etc.) as well permanent annual funding for replanting and maintenance. All the better if these monies and responsibilities go to a neighborhood group that takes personal responsibility for it.

And there you have it. Not fully urban, but more urban than Stewart's proposal. Not threatening to perfect, but definitely "gooder." Which should satisfy all involved.

A crown for Warren Street

Above: Route 9 (solid line) currently jughandles through Hudson because a narrow,
two-block section of roadway prevents a simpler, straighter route (dashed line).
Below, the existing intersection of Warren Street with Worth and Prospect.

In a recent post, I looked into rerouting U.S. Route 9 as a way of reducing truck traffic within the Hudson street grid. I raised the possibility of simplifying the route by widening a narrow, two-block stretch of Prospect and Fairview Avenues, and as an additional option, merging Prospect and Fairview into a continuous street.

In this post I'll look at the intersection where my rerouting began, the meeting of Warren Street with Prospect Avenue and Worth Avenue. Some Hudsonians complain about this intersection because of the three-way stop signs. Personally, I find communal value in the eye contact and informal coordination they call for, but at rush hour, when I almost never drive, such charms are perhaps lost.

Whether or not the intersection is a failure from a traffic standpoint, it is most definitely a failure from a civic standpoint. This is an important intersection in Hudson. It marks the head of Warren Street, the cultural and commercial spine of the city, a street celebrated for its architecture, restaurants, and galleries. But while the bottom end of Warren Street is anchored by a promontory overlooking the Hudson River, here there is no indication of the street's significance. There is no major building or space to complete the street axis and pull walkers to its top end; the strongest architectural gestures are an auto repair garage and a marginally maintained, 1870-ish flop house. Besides presenting a significant aesthetic problem, this represents an economic and cultural problem, as properties at this end of Warren Street have tended to languish.

Warren Street is the commercial and cultural spine of Hudson. It is is received by a park at the Hudson River, but it lacks a
significant marker at its upper end.

Below I offer a suggestion for anchoring the top of Warren Street. I've realigned Prospect and Worth Avenues into a continuous street, and extended Warren to meet them in a T. I also rerouted the alley behind Worth Avenue to connect with Prison Alley. These changes would result in the loss of two houses (one architecturally undistinguished) and the auto repair garage. The gain is a clear beginning/end to Hudson's main street and a large, developable parcel to the southeast.

A street realignment and new building and plaza offer a civic gesture at the head of Warren Street.

Portland, Maine

On the new site I've suggested a large building and plaza to preside over Warren Street. The location lends itself to a public building such as county offices, a police station, or a library. Unfortunately, these functions are all in the process of being relocated to other sites in Hudson. A church, grand hotel, or apartment building might work, and the proximity of the hospital suggests an administration building. However, some of these uses might not lend the right civic note. But it's fun to imagine something new and impactful at this location. A good street should have a good beginning. And a good end.

On gas stations, street walls, and beer gardens

The view from the park to the northeast/east is unwelcoming.

Urban spaces succeed through their relationship with their context. If a public space is of a type that is to be activated by workaday activities, it usually needs to be fronted by a well-defined street wall. Strong built edges generate life in a park, define it as a meaningful place, and provide a sense of "hereness."

Hudson's Seventh Street Park is defined reasonably well on most of its perimeter, where two and three story buildings directly front the sidewalk. But the northeastern/eastern side/corner is "spatially leaky." A homely, one-story Citgo station sits sixty or more feet from the public way. A train right-of-way immediately west of the Citgo station and some odd building and street geometries to the east contribute to additional spatial weakness. A park user walking toward this corner is presented with a destination that ranges from unclear to unpleasant.

Most of the park's perimeter has fairly well-defined street
walls, but on the northeast and east it is spatially leaky.

Of the several offenders, the Citgo station receives the most frequent criticism. I hear occasional, hopeful musings that economic pressures from increasing real estate values will cause it to eventually go away. I think this is unlikely. It is difficult to get rid of gas stations under almost any economic circumstance, because they earn more than almost any other use. If we are going to address the problems at this end of the park, I think it would be wiser to seek an urbanistically friendlier design for the gas station site that maintains Citgo as a tenant. If nothing else, the exercise is good practice, design-wise and emotion-wise, for learning to work with undesirable constraints.

At present, vehicles enter and exit the Citgo site on the same side, i.e., Columbia Street (below left). This design decision, made long ago, required that the building be sited on the back of the site. A more sympathetic building placement can be achieved by employing a drive-through circulation pattern, from Columbia Street to the alley (below right). 

Vehicles currently enter and exit the site via Columbia Street.

A through circulation pattern would permit a friendlier building placement.

This would allow the placement of a building on the Columbia Street edge, where a building stood years ago. Ideally, a new building here would be three stories high (the former building was two stories) and would house ground floor retail (including the gas station mini-mart) and other uses that would activate the sidewalk and park edge.

The Citgo site with a 3-story building fronting on Columbia Street

This would solve part of the problem, but the railroad right-of-way to the west (i.e., left) of the gas station would remain spatially leaky. At bottom, I've illustrated one way of strengthening the street wall here. I offer no warranties on the quality of any of the architecture represented, although I suspect some will find the beer garden compelling.

In a perfect world, we wouldn't need to concern ourselves with oddities such as a gas station and a railroad on our town square. If we were facing the decision today for the first time, would any of us, other than Citgo and ADM, choose to have them here? And yet many Hudsonians, even some staunch formalists, admit to liking the intrusion of the railroad. It is part of who we are, part of what the park is, part of what makes Hudson different from other places. Indeed, the things that by some measures don't belong in a city are the very things that make it authentic and interesting. Granted, it may be easier to romanticize a railroad line than a Citgo sign. But before declaring that a conclusion, we probably should check in on the question with the Red Sox fans among us...perhaps even some Yankees fans.

Some possibilities for strengthening the street wall at the northeastern side of Hudson's Seventh Street Park.

The style trap

Carole Osterink has a way of beating me to the punch. While I am still processing raw thoughts or gathering background information for a blog post, it seems she's already captured most of the issues on her blog, and more elegantly than I would have done.
     Her post today, Rebuilding the Second Ward, speaks to the housing to be built on Columbia Street, something I've been looking into for a while. The buildings will include three units by Habitat for Humanity as well as the eventual replacement for Bliss Towers by Omni Development, which will total about 132 units. What should these buildings look like? From where should their designers derive their design cues, given that the neighborhood is so broken apart? How can these buildings be contextually responsive when it's hard to say what the context is and whether some of it even should have been put there in the first place?

from 101 Things I Learned in Architecture School

     Carole covers these issues well, so I will only add one point: Style is the last thing we should be thinking about in addressing these concerns. When architects dwell on style, they invariably neglect more essential considerations such as proportion, massing, and texture. A building can be compatible with its surroundings if it hits all or most of these marks, even if it is in a different style. Likewise, it can be very incompatible even if its style is "correct." Certainly, we've all seen bad additions to old buildings that were built in the same superficial style but that simply look wrong.
     Important as such visual considerations are, however, they are merely that—visual considerations. The underlying reality of a building is far more important than its outward appearance, and this is where the design process is properly rooted.
     When I was a student of architecture a long time ago, this was the last thing I expected to hear. Architecture is a visual art, after all. As a confused student it only made my life worse to hear an instructor define architecture in a way that seemed to have little to do with bricks and mortar: Architecture is the physical manifestation of a social order.
     This didn't seem particularly helpful or relevant at the time, but I've come to realize that it's the most important value to bring to the design process. What is the social order the building needs to accommodate? How do the people who will dwell in it need to live their lives? How do they sleep and eat breakfast and raise their kids and relate to their neighbors? What about the people walking by the building and living near it? How might they get to know the occupants? Should they be able to look directly into the house and see them? Or can this social relationship be accommodated and enriched by creating gradations of private-public space?
     Such investigations of social order, when considered sensitively, knowledgeably, and patiently, ultimately will lead to the building that needs to be. Appearances will tend to take care of themselves if you do the rest right, but the doing-the-rest-right part is very difficult. And so far, I am concerned that the developers of the aforementioned properties don't seem to have the inclination to engage the design of their buildings in this way, or perhaps more dangerously, that they believe they are designing this way when they really aren't. I don't know how to convince them otherwise; maybe this is another thing Carole can do better than me.

If you build it...

When I'm not exploring the city on foot or having coffee with other Hudsonians who know the city much better than I do, I often find myself studying aerial views of Hudson on Google or Bing. I'm usually trying to make sense out of aspects of the city I don't yet understand, to identify hidden patterns, or to find ways to make Hudson a physically, experientially, and socially more coherent place. The mishmashed street geometries at the eastern corner of the street grid (are there sensible ways to realign them?), the truck route (any alternatives not yet identified?), the Promenade (can it be extended north-south?), and a million other things spill through my mind, only occasionally leaving me with greater clarity than when I started.

The other day I found myself studying the Second Ward, looking for physical factors that contribute to the social isolation of its residents. A few were easy to identify: It has more vacant lots and more random open space than are found in the rest of the Hudson grid, making for a less comfortable milieu for strolling, hanging out, and having casual social encounters. Too, the buildings sit farther back from the sidewalk and are much lower, taller, or longer than the Hudson norm. Such physical differences in low income neighborhoods tend to stigmatize their residents, increasing their sense of alienation from other city residents.

High-rise buildings are especially problematic, as it is harder for their residents to participate in the social life of the city than it is for residents of four-story-and-under buildings. Above the fourth floor or so, it is difficult to recognize a familiar face on the street or to call out a warning to a child endangered by a passing vehicle. One's visual field tends to be oriented away from the immediate street environment and more toward the distant cityscape and landscape.

Social interaction in high rise buildings is also limited by their internal corridors. No doubt there are valuable friendships and social interdependencies within Bliss Tower, but such relationships tend to arise despite the physical environment, not because of it. The biggest problem with corridor buildings is that their residents are either completely inside their apartments or completely outside them; there's no in-between. A solid fire door on a blank-walled corridor eliminates gradations between in and out, between the private world of the home and the public world of passersby. This reduces the opportunities people have to meet each other. The single woman in 7A hoping to "accidentally on purpose" bump into the handsome guy living in 7G, or the elderly woman looking for a friendly greeting from the maintenance man won't even know when they have come and gone. Such might not sound like a big deal, but it becomes a big deal for those living with it day after day. By comparison, apartments in buildings with porches and stoops and that open directly to the street offer many more opportunities for their residents to casually interact with neighbors and strangers.

As I further considered the Second Ward's isolation problem, I looked at the city's structured public spaces. Structured spaces have built edges on two or more sides, and are usually located on or next to a highly used pathway. Thurston Park on lower Warren Street is an example, and a similar one lies adjacent to Mexican Radio. The Seventh Street Park and the Courthouse Green are much larger examples, and the buildings that enclose them are across the street. Such spaces are valuable for fostering social interaction; they are the spaces people go to to take a break from work, eat an ice cream cone, feed the pigeons, or people watch. They are also the spaces people incidentally pass through, all of which fosters the casual encounters that build familiarity and knit a community together. Soft public spaces (such as Promenade Park, Waterfront Park, and the Cemetery) are also important to cities but they tend to be located more on the periphery of neighborhoods or districts and to support somewhat different social purposes.

A structured public space: Thurston Park on lower Warren Street

If my informal survey is accurate, Hudson's structured public spaces consist of the four mentioned above, plus the Parc Foundation park. As you can see below, all five spaces are south of Columbia Street, and three of them are south of Warren Street. There are no structured public spaces in the northernmost area of the city, where the Second Ward is located. (I couldn't find a ward map, otherwise I would have shown the ward boundaries.)

Structured public spaces (in red) are important for fostering casual social interaction.

This suggested to me a somewhat different approach to weaving the lives of Second Ward residents into the life of the city, particularly if Bliss Tower is eventually replaced by low-rise development as is often discussed: We shouldn't only be looking for ways to bring Second Ward residents toward Warren Street and the rest of Hudson, important though this is; we need to create reasons for the rest of Hudson to go to the Second Ward. Indeed, there are only dead-end streets beyond the Second Ward, which means many city residents almost never pass through it incidentally. But imagine a new structured public space in the Second Ward with a few mom and pop storefronts along the edges, some benches for the elderly to sit on, some shade trees, and all the neighborhood kids coming and going.

I would sure go there to people watch.

“There is a quality even meaner than outright ugliness or disorder, and this meaner quality is the dishonest mask of pretended order, achieved by ignoring or suppressing the real order that is struggling to exist and to be served.”

— Jane Jacobs, The Death and Life of Great American Cities

New City Jail, er, Senior Center on tap

Last week I attended a meeting on the proposed Hudson Senior Center. The meeting brought out over 30 seniors as well as a number of local officials, including Mayor Richard Scalera and Common Council President Don Moore.

The Senior Center is to be built as an addition to the existing Youth Center building at Third and Union Streets. If well conceived it promises to be a boon to seniors as well as others using the building and this part of Hudson. But from what can be discerned about the project at this point, there are many reasons for concern. Chief among them are a weak design concept and, I believe, an unrealistic schedule.  

The existing building (above) is not a classic beauty; it is rather brutish in scale and is not in the best of repair. It appears that some detail has been stripped over the years and replaced with mismatched brick. The upper windows have been replaced with bronze aluminum storefront while the front entry was mismatched with anodized aluminum. Nevertheless, it is a building of substantial presence in Hudson, and it lies at a significant entrance to our small city. As such, the addition needs to give far greater respect to its location than the proposal (below) allows. The addition put forth is badly proportioned (squat height, a meager 8’ ceiling on the first floor, windows more scaled to a suburban split-level than a public building, oversized rather than standard brick, and more) and arbitrarily detailed. Note, for example, the quoins pasted onto the corners of the addition and the dentils applied along the cornice line; no such features appear on its predecessor.

Existing Hudson Youth Center. Senior Center addition will be at left.



Proposed Senior Center: Cherry Alley elevation

I have to think that these aesthetic problems point to some more elemental problems in the design process. How carefully have the needs of the building’s users been considered when, for example, this proposal was put forth without benefit of a single meeting with the seniors’ group? How practical can the design be when it was created without knowledge of the specific programs that may/should/will go on inside? To what extent did the designers put themselves in the shoes of the elderly when they didn't think to put an overhang above the door? How much study was made of the relationship between the seniors and the youths who share the building? What opportunities and problems was this found to present? The apparently unaddressed design issues go on from there.

To be frank, this is what happens when a city hires an engineer instead of an architect to design a building. I do not mean this as a swipe at my colleagues in the engineering profession. Engineers can do things that architects cannot do; thank goodness for that. But the opposite is true as well; architects do things that engineers cannot. We are trained to work closely with a building’s users to uncover the core issues--practical, psychological, historical, material, all the rest--that need to inform the design process. Only after some real understandings are attained do we move the process forward. Crucial to this process is an understanding that architecture is not something that gets applied to a building after it is "engineered"; if anything, it is the other way around: The engineering of a building is but one of many concerns the architect integrates holistically into the design process.

As for the schedule, the mayor said the building will be ready next fall. Buildings can be designed and built in this time frame, but most often the process takes much longer. In fact, the more one rushes a design and construction schedule, the more things are likely to cost. The city will have to hire the first available contractor instead of taking time to find the right one. Cost increases during construction will be more likely, as oversights are discovered as a result of having rushing the design process. ("Oops, I didn't realize we'd have to relocate that utility pole...")

At the risk of parting with a cheap shot, I found myself wondering what this building will look like in three dimensions, so I Photoshopped the engineer's drawing onto a photograph. I then asked myself, if I were to drive by this building unawares, what would I guess it was? The only answer that came to mind was the city jail.

This isn't the right image for our seniors or our city. So before we walk farther down this path, can we all take a deep breath, step back, and engage the design process the right way?

New city jail? A three dimensional view generated from the engineer's drawing

Park Falafel and Pizza

I've enjoyed two trips to the relatively new Park Falafel and Pizza, on 7th Street facing the Park. The falafel was very good, the pizza a tick above average. I'll give the pizza a second try soon in case I happened to have it on an off night.

I remember seeing this building on the market earlier this year. As I recall, it was a foreclosure, asking price of 89K (I think). I am told that the building was pretty bad inside, but the location is excellent. So it was good to see it turned around so quickly, particularly given the demise of the pizza joint across the park sometime after my first visit to Hudson in 2009 and before I moved here in September 2010.

The storefront renovation is simply and effectively executed--a good lesson in hitting one's design marks without breaking the bank. Inside, large color blocks, muted somewhat from primary tones, strike an appropriately jaunty note for informal dining while helping create a brand identity. (Note to Kennedy Chicken: If you ever renovate, look here first.)

From what can be seen on the MLS, it appears that the apartments on floors two and three were renovated--good to know when there are so many substandard units in the city. But I am left wondering why most of the building exterior was ignored. While I don't expect the uneven settling of the brick facade to have been repaired (rarely a cost-effective undertaking), the rest of the building should not have been left in such a shabby state. 

Perhaps the building owner (I don't know if this is the same party as the restaurant owners) has plans to address this and hasn't gotten to it yet. With a bit of work, the building could be an amenity for the whole city, not just for those seeking a reasonably priced dinner, given its prominence at the busiest entrance to Hudson.

Personally, I'd like to see the signature color blocks used as a bold color scheme for the building exterior. The more visible the building, the more people will visit the restaurant and the more interesting and alive this little stretch of Hudson will become.

Imperfect, but OK

I came across this nifty little house on lower Union Street the other day. It is only about ten feet wide and the ground floor looks barely tall enough to stand up in, but I found it rather appealing. Note how carefully it is composed, with all elements in perfect symmetry: the center of the porch, the bay window, and the brackets above align precisely. Even the squatness of the ground floor contributes to the whole, as it makes the bay above it seem all the more ambitious.


Actually I'm fibbng; the image above was Photoshopped. The real house looks like this:

 

Notice that almost nothing lines up. While the bay window seems to be centered in the facade, the bracket above it, which one would expect would be centered, is several inches to the left (see below). Meanwhile, the middle porch column stands obtrusively in front of a window (I deleted it from the first image) and is even farther left of center than the roof bracket. And the left and right ends of the porch roof overhang differ from each other.



Why do such flaws occur? Why can't architects or builders get things lined up?

The reason is that when buildings abut, as they often do in urban environments, architectural features such as roof edges cannot turn the corner as they would on a freestanding house. The roof overhang on the front of this house couldn't be continued on the right side, because it would intrude on the neighboring house, which would create not only an aesthetic problem but a legal one. So the bracket on the right was pulled in a bit from the edge(below), even though the bracket at the left side is flush with the left edge of the house. To make the whole assemblage look more three dimensional, that is, to give a sense that at least part of the assembly turns the corner, the topmost molding was returned 90 degrees into the facade.

The abutment problem in urban places has a domino effect on the design of a facade: if you can't place a roof bracket quite where it wants to be, do you adjust all the brackets so they are evenly spaced, or do you let the spacing of the end one be mismatched? Do you place the windows so they align with the roof details or do you place them symmetrically in the facade without regard to the roof? Which "mistake" would be more or less noticeable?

No answers here, but on a completely different note, as I played with this facade, I found myself wanting to grow something out of the top of the bay. I had no idea why, until I realized that in removing a porch column I had turned the porch into a de facto table... a good place for a vase?

   
Mmm, maybe not.