Wednesday, November 30, 2011

A maze of twisting passages, Part Three

Gregory (Scotland Yard detective): Is there any other point to which you would wish to draw my attention?
Holmes: To the curious incident of the dog in the night-time.
Gregory: The dog did nothing in the night-time.
Holmes: That was the curious incident.
-Silver Blaze, by Arthur Conan Doyle


One thing about the building that seems kind of weird when I think about it is problems we don't have.

For starters, this place is not dingy overall. Computers are newish (even if software isn't). Office supplies are plentiful. My desk chair here is nicer than the chair at my computer at home. Most of the furniture here would not look out of place in a nice law office. (Admittedly, I assume that the desks, shelves and office furniture of my peers would be for interns or behind-the-scenes paralegals or whatever, and only a dozen people in this building have furniture fit for a lawyer in private practice, but still.) My cubicle doesn't give much privacy, but some peoples' cubicles do, and some people have actual offices.

This compares favorably to similar workplaces in private industry, at least in my experience. The one time I went to the home office of my contracting company, it was in worse shape than this. My previous job was in publishing, and the desks there were described as "Soviet-era" for their sturdy construction, dented metal exteriors, and general ugliness. Only the owner of the business had his own office, and most people didn't even have cubicle walls to make their own, just a desk. Computers were more out-of-date there than they are here. People at that previous job would be jealous of my workspace even as they looked down on the larger building.

There are a couple problems here that I didn't mention in previous posts, though. First, the walls, especially in public spaces, take forever to get fixed. Just standing up I can see a bunch of scuffs on one piece of wall where a chair has been repeatedly bumped up against it for years, and a segment of wall with a couple feet of in the plaster or whatever where wiring was removed. Both have been like that as long as I can remember. On a door around the corner there's a gap in the paint where a sign used to be. (Not where the paint was faded; the sign was presumably attached to the door before the paint, they painted around it, the sign was removed.) The paint jobs are even worse on the stairs. There's a huge range in quality of conference rooms: from worse than my own office to better than my boss's boss's office.

All in all, there is widely variation of quality of different things here, and on boring afternoons it's kind of interesting to think about why.

Tuesday, November 29, 2011

Why I would give Congress a low approval rating

One annoying thing about the doomed project is that it relies so much on the exact text of a Congressional mandate, which is annoying because Congress apparently doesn't have a style guide, or if they do it doesn't match ours.

There are two specific examples. First of all, the word "shall". My office does not use it any more. I don't know exactly when we stopped, but it's bad. It's archaic; you'd never ever hear it in conversation and it has long since fallen out of daily use. It's vague; "shall" probably usually means "must" but sometimes it could mean "can" or "should", and how can the reader tell which? Regulations have to be clear. And this is exactly what they have to be clear on - whether something is or is not required. Occasionally we get subject matter experts using "shall" because they're in the habit of reading or even copy-pasting from 15-year-old material, but we tech writers just correct it and move on. Writing regulations in plain English is often a challenge and is sometimes not worth the trouble, but it's always something to aim for, and avoiding archaic, vague language on important points is a great place to start.

Congress, however, freely uses "shall". At least they did in the mandate on the doomed project. It's hard for me to lay down a definitive, blanket rule in the project about it when Congress is doing the reverse.

Also, there's the phrase "interim final rule". To us, it means nothing. As I've said, the normal process for a project has two publications: a notice of proposed rulemaking followed by a final rule*. Sometimes, for something where we're allowed to or really need to skip the notice phase, the notice is replaced by an interim rule, which is effective immediately but has a period for public comment and will be followed by a final rule for any necessary revisions. "Interim rule" is a defined term in our internal guidance, and is meaningfully distinct from everything else, and just makes sense in layman's terms. However, "interim final rule" is none of the above. It doesn't appear in our handbook or on the fancy laminated poster-sized flowchart of the rulemaking process and it seems intuitively contradictory, at least in my opinion.

But, again, Congress uses that term. I'm not sure what they think it means. It appears in the mandate for the doomed project more than once.

So the doomed project has "IFR" all over the place, and "shall" keeps popping up. Obviously, these are not major problems, but they're my problems, so they're annoying.

* Sometimes there's also a supplemental notice between those two for when things change signicantly before the final rule, and sometimes there's an advance notice before the notice for when our ideas about what we want to do aren't firm enough for a notice. But the notice (or IR) and the final rule are two basic steps that almost every project has.

Monday, November 21, 2011

A maze of twisting passages, Part Two

It's worth noting that my comments about the building are not just my opinion. Like I said, there's one floor which was recently remodeled. The windows there are waist-height to ceiling, wall to wall. Clean, well-maintained exposed piping on the ceiling allows most of the floor to be lit more brightly and breaks up the monotony. As for colors, there are more browns, yellows, wood patterns, and genuine wood. Some conference rooms even have wallpaper that looks kind of cool.

People who work on this floor are lucky compared to people on other floors and they know it. The window nearest the set of elevators on my side of the building has a styrofoam cup, and over it there's a notecard taped to the window right where people would come to bask in the sun on nice days. The notecard riffs on the old Mastercard ads, with prices for things like "quick peek" and "longing look", and the final line is "escapist fantasy... priceless!"

But here's the funny thing: there is money in the styrofoam cup, and I'm pretty sure it's not just a joke or aspirational seed money. So clearly, a lot of people share my opinion of the state of most of the building.

However, there are still at least two problems with the remodeled floor, both of which happen to be problems with the layout, that make it clear that it's still in the same building as the maze of twisting passages and designed under the same pressure, even if they don't actually have twisting passages up there.

The first problem is that the numbering scheme of offices, cubicles and conference rooms on the remodeled floor is... insane? Incomprehensible? Nonexistent? Well, I'm sure there's some logic to it, and I totally realize that the floors I'm familiar with could be confusing to some*, but the remodeled floor is a huge leap in confusion. I think even-numbered rooms might be on the opposite side of the building from the room with a number one higher or lower than them. When I want to go somewhere there I normally just pick a direction at random and walk until I find the room number I'm looking for.

And second, the middle section of it is blocked off. You need some kind of security clearance to get there. This is a problem because it's the middle of the floor. If you take the elevator there and happen to be on the wrong side of the building (and I never know about that until I get there because, like I said, incomprehensible numbering), you can't take the most direct route to where you need to be, you have to walk all the way around. They couldn't have blocked off one quarter of it to the side and left unimpeded traffic through the center of the building? No, apparently not.

So all in all, the remodeled floor would be rated much higher than the rest of the building by anyone who cares about interior design, but it still is never going to win any awards and still reveals the pressures that shape government buildings.

* I see people who probably don't work here much getting off the elevator near my office confused, probably because the only direction is badly designed. A floor plan is posted near the elevators closest to me, but it was designed from the point of view of the other end of the building, so there's a "you are here" sticker and there's an intuitive place to assume you are when you look at it and they are on opposite sides of the map. They should have either used a much bigger, more obvious sticker, or designed two different maps, one for each side of the building. It's symmetrical, so it's not like two maps would have been too hard.

Thursday, November 17, 2011

Confrontation

We had a meeting about part 12(c) of the minor project near the end of October. Verdict: we got to do the table our way. Victory. The meeting was more than a bit aggravating, though. The senior tech writer (sort of my manager but sort of not, the most experienced writer who is kept in the loop on all this stuff but he doesn't actually manage anyone) came along but left me to do most of the talking. Optimistically, maybe he would have jumped in if he thought I was screwing it up, but he doesn't seem much like the type. He's a very bad speaker. Bad in a public speaking sense, although he manages to ramble and trail off in one-on-one conversations too. It's just as well I kept things under control there.

In the end, I think I did OK. My arguments may have shifted a bit over the course of the meeting, going from "You can't use these previous rulemakings as a guideline because this one is different" to "Well, you can use these previous rulemakings as a guideline, because they actually mean something completely different than you thought." Underneath it all, of course, my real reason for it was "Do it our way because this is our choice, and our way is this way because my boss says so". My boss's way was probably the best way, but I could have made a good argument for my way if we weren't already behind schedule, and it wouldn't be too hard to come up with an endless variety of tweaks to the formatting of a table. The differences between them would have been almost trivial.

It was downright weird, maybe the most decisive I've had to be so far in this job. It's a pity it was over this. There are lots of problems that actually matter in this job, but the only significance to this problem was the internal turf thing.

Wednesday, November 16, 2011

Literally a maze of twisting passages, all alike

One insidious mind-numbing part of this job is the building. It's almost a perfectly typical base of operations for a faceless bureaucracy. There is one floor (recently remodeled, more open floor plan, better lit) out of seven that differs from the rest, but even it is mind-numbing in some ways.

Probably the worst thing about it, especially as the winter is starting, is the lack of windows. There are none in my office, nor any offices I go to regularly. The cafeteria has windows and I get lunch there most days and bring it back to my desk. (Maybe my mood would be greatly helped simply by eating down there. But then when could I blog?) Even in the cafeteria or offices that do have windows, high cubicle walls and generally bad layouts make it hard to see the sky in anything more than just narrow glipses. The only time I see the sun during work weeks in the winter is during the five-minute shuttle ride on my morning commute, and when I get lunch or deliberately go outside for some reason.

In the halls of most floors, the walls are off-white or eggshell and the trim, floors and ceilings are various shades of gray. In offices on most floors, the grays continue with eggshell cabinets and are supplemented by conventional shades of peach, tan and light brown on cubicles and metal filing cabinets. I think the desks are wood-patterned plastic.

Lightlessness and monotony aren't the only problems with the layout of the place. Halls on either side of the building twist around the elevators and stairs and bathrooms. In a building with a sensible layout, going from one hall to a parallel one requires making two left turns (or two right, depending on which way you're going), but here it requires turning left, then right, then left, then left (or the reverse, depending). The layout is simple enough on a map, but when you're walking through the halls they almost feel labyrinthine.

Various offices have seals or flags outside the door as appropriate and there are decorations, but there aren't too many and they are mostly staid, conservative and cheap. Paintings depicting the history of this department's field or people in this department doing heroic deeds, but mostly in a boring, representational style, at least a step below the quality of art in a hotel lobby. In addition to the art in the halls, the main hall on the first floor of the building is taken up by a permanent exhibition about the history of the department, and there are also several smaller historical or educational exhibits on other floors. However, these confuse me a bit. I don't understand who they are for. This place is hardly a tourist attraction. In fact, you need employee ID or an escort to even get into the building. The only people I can think of appreciating the exhibits are visiting dignitaries with unaccountable free time, and new people in the first month or so of their time here.

So my office building is mazelike, gray, drab and windowless. There are few decorations in public areas, and those we do have are either cheap and crappy or pointless, if not both. Admittedly, us writers in cubicles get the worst of it, and some people with offices and windows have downright nice workspaces, but still, from where I sit, this place is soulless.

Tuesday, November 15, 2011

Correction

Oops. I was way off in my "Alternate History" post, and probably in other posts about the doomed project as well.

Today in a meeting people referred to another requirement mentioned in the Congressional mandate that I hadn't thought of when I wrote Thursday's post. After hearing that, I dug out the actual mandate and read it. I had thought it only mentioned the size limit, but no, it clearly says a lot more than that. It says that we have to ensure the safety of individuals and the environment and specifically mentions certain staffing requirements. So I have to reevaluate that gray area. It's not as big as I thought, and it starts within the mandate. I thought a lot of the stuff in the rule was scope creep, but we really are required to do a lot of it. Which is kind of funny, because the meeting today began with H. complaining about scope creep and at the time I was thinking to myself "Of course it's scope creep. Limit this project to the mandatory size limit change and closing the standards loophole and it would be 30 pages, published six months ago. Oh, the mandate also mentioned staffing? Huh, I wonder what else it mentioned? OK, with staffing maybe it would be 50 or, hell, even 100 pages, published three months ago. The remaining 200-plus pages of the document and ongoing work is all scope creep."

That's obviously not the kind of thing anyone could say out loud, let alone me, but all these meetings about putting in just one more part or changing just one more thing get aggravating. But the staffing was brought up in the meeting, and that made me wonder just what else exactly fell under the mandate that I don't normally think of, and so I actually looked it up, and, wow, the mandate is actually fairly broad. And to be clear, there really is quite a bit of scope creep. Anything that's not truly essential for safety, rather than just being kind of nice, is scope creep. Anything that applies to things below the former size limit is scope creep, especially anything that only applies to things below the former size limit. Anything related to cross-platform compatibility is scope creep (I think). It's just that lots of stuff that I thought was scope creep probably isn't, and it's objectively harder to tell what is.

So anyways, what does all this mean? That the idea of meeting the deadline was always crazy to begin with.

Thursday, November 10, 2011

Alternate history

Obviously, the doomed project is a mess. But how should things have gone?

To be clear, the original timeline was impossible. The team was formed in December 2010 or so, it was a normal project in all respects except for the deadline pressure and one piece of red tape Congress cut for us, and the timeline called for publishing an interim rule in the summer and a final rule by the end of 2011. That can't be done. Not by this agency, not with this kind of project, not without more time and/or more red tape being cut for us. Two projects I'm on have both been under review by other agencies for more than a year, and that's just for one stage of them. Rulemaking projects can not be completed in a year. The second version of the schedule, though, with the interim rule published by the end of the year, actually might not have been crazy.

When Congress gave us the mandate, the Congressmen behind that amendment (or rather, the industry lobbyists, but that's a topic for a whole other post) probably pictured something less than 20 pages long, almost all of which would have been organizational stuff or legally mandated boilerplate restating our justification for doing it. They probably thought the amendatory instructions could be less than a page long.

That's unrealistic in this case. The amendatory instructions couldn't have been shorter than three pages. Blame the current regulatory status quo for that. I don't know whether it was created by one particular regulatory project or by accretion of regulations over the years or what, but either way it wasn't logically organized. Add in another page or two to the preamble to explain that we had to fix a previously existing mess, and that's at least a 25-page document.

And you know, if we had done it like that, the deadline doesn't look crazy. Still difficult, considering how many people need to approve of it, and unfair, considering how much of it is out of our control, but not crazy.

But then there's at least one more element, which Congress didn't foresee but should have. Basically, it's a conflict between two standards. The mandate was phrased in terms of the current, most modern standard. But another standard exists that is also used in some parts of the industry. If we just did what Congress said and ignored the other standard, it would carve out a huge, bizarre exception. It would give people an incentive to use the older, deprecated standard. It would probably be objectively bad for the country, and would certainly be bad for this agency. But since Congress didn't actually tell us to do this, we would have to spend time to close that loophole and justify our actions. So that would add about, say, five pages to the document and one week to the timeline. Now we're right on the edge of the "not crazy" line.

And then we get to the stuff that does not fall under the mandate at all. (Which, again, is a gray area.) Some of it probably should be included no matter what because it really is necessary for basic safety, environmental protection, or to enforce other requirements. Some of it would be a good idea, but if we had to wait a few more years to do it, the delay probably wouldn't cause a disaster. Some of it probably isn't a good idea at all. There is a lot of stuff in this gray area; like I've said, the mandate applies to a category that this agency was already working on broad revision of, so there's a long list of things they have been waiting to do. I can make educated guesses about what falls where in the gray area, but I'm not an expert and it's not my decision.

But the thing is, if you put anything from the gray area into the rulemaking, the idea of meeting the deadline is crazy. If we thought that the deadline was seriously important, we should have identified the really most essential bits in the gray area and planned on going over the deadline by just a month or two. On the other hand, if we want to say that the deadline doesn't matter and we're doing this rulemaking "right" regardless of how long it takes, we should have had an aggressive but otherwise normal timeline. Due in part to competing priorities in different offices, our current approach to the deadline seems to be the worst of both worlds.

Monday, November 7, 2011

Objectivability

I'm ambivalent about the unhelpful guy's fight with his critic. Obviously, picking on someone because English isn't his first language is unfair. And for the record, the unhelpful guy speaks with an accent but his command of the language seems fine. I can always understand what he says and vice versa as far as I can tell. His typos are within the range of normal English speakers. His formatting is a mess, but that's not a native-speaker-versus-non-native-speaker thing. There's basically no good reason to complain about his language. On the other hand, if the criticism of his language was meant as just good-natured ribbing, then him blowing up wasn't really fair either. Understandable, maybe, at the end of a tiring meeting especially if it's something he's sensitive about, but still, not exactly fair.

However, I have plenty of other problems with him. In this morning's post, and the previous posts I link to, I point some of them out.

It's hard to get a straight answer from him about anything. Someone asks him a yes or no question, choose right or left, one option or the other, but instead of saying yes or no he gives an example or explanation that seems to imply "sometimes", and it's only apparent that he really meant "yes" after five minutes of talking about it.

Version control problems seem to mysteriously follow him around. In addition to the usual basic unnecessary work, I also don't like it because it's confusing when he asks me to fix for something that has already been fixed and he just didn't realize it because he doesn't work in the same document as everyone else. Worse yet, version A was a month ago, the team has since decided on version B, and then the unhelpful guy asks me to do C. It's up to me to figure out whether he disagrees with the team about B or just thinks we're still on A, and it's usually the latter.

More generally, he seems pretty technically inept. Admittedly, sometimes in an absent-mindedly innocent "what's the shortcut to copy and paste again?" way, but also sometimes in a "can't be bothered" way. When a spreadsheet has a column for peoples' names and another column for inputs and he fills in his input in the name column, for example, that's just plain carelessness. Trivial, sure, but annoying.

He's often out of the office, gives us little notice of it, and is apparently the only expert we have on certain things so we just have to wait for him to get back.

Every two weeks or so I get an e-mail from him asking for apparently random changes in the document. This is a mix of him catching unambiguous errors by me, him catching errors by me that I'd say were caused by miscommunication (see above) so it's sometimes technically my mistake but I can't make myself feel guilty about it, and him asking for new substantive changes. As for the first type, fair enough, but it's still hardly fun. As for the second, obviously no one would appreciate being in my position. And as for the third, it's his right and his job, but why is he doing it now, and in a way that makes more work for me and other people? Everyone else manages to do their job without adding that extra step of more work for me, but he can't, apparently.

And this part isn't my business, but I have the sense that a lot of what he's doing in the doomed project is outside its mandate. As I've said, one problem with the project is people putting in a lot of things that aren't strictly required by the mandate. It's a gray area, a continuum - some stuff even a layman can tell is a good idea, some stuff the experts can make good arguments about why it's necessary, and some stuff still doesn't seem like that good an idea even after I've heard the arguments. And it seems like a lot of the unhelpful guy's work falls on the far side of that gray area. Like I said, this isn't my business, but whenever someone brings up deadlines I wonder how much better we would be doing without the parts that really just don't look justified... many of which belong to the unhelpful guy.

So I feel completely incapable of being objective about whether the unhelpful guy is being mistreated due to his language, when I have a whole bunch of other reasons to dislike him.

Friday, November 4, 2011

Reinventing the wheel, axle, camshaft, and stereo system

Every project has a certain amount of flexibility in it. Different people have their own personal preferences, different projects have different needs, etc. So certain parts of the process get reinvented by every single team, but there's nothing wrong with that.

But the doomed project is such a mess that we've had to reinvent the process itself from scratch, and it has been, obviously, a mess. In particular, there are two different Microsoft Excel spreadsheets we've used. They're keeping track of the same kind of data - input on the document from at least some of the same people - for use by the same team members, in the same ways. And yet the two spreadsheets have different colors in them, for a while they had different numbers of colors in them, the colors mean different things, they have different numbers of columns.

That's partly my mistake: I should have been better prepared, because it's not like being asked to make them was a surprise. I should have sat down and thought, maybe asked for help or looked for other examples, and come up with one layout that would work fine for both of them. And it's partly me just being adaptable: the first one worked less than perfectly, so I made changes in hopes of improving it. Good for me.

Partly, though, it's because the spreadsheets need to be used by different people in different ways. Personally, I'm not an Excel guru. When I saw the economist use the formula tool in it in a certain way a few weeks ago, I marveled at the elegance of what he had done, and I'm still picking up minor tricks with it that I wouldn't have known a year ago. But still, I'm capable of copying and pasting things, and if coloring a cell or row is a useful way to store information then I can handle that and keep it straight just fine. So I do.

Some people, though, clearly either aren't capable of that much attention to detail and command of Excel, or just don't care. One of the changes I made to the spreadsheet in the second version was dropping a color. In the first version, one problem was that people would make cells the wrong color. Maybe they misunderstood what the colors meant, maybe they just added a new row and Excel automatically gave it the color of the neighboring row and they didn't think about that. Either way, too many colors was apparently a confusing variable that I thought we could do without on the second spreadsheet, so I tried to keep it to just two. I feel pretty justified about that because even with two colors some people still made cells the wrong color, but I don't think I can simplify it much more than this.

But in the first meeting using it, we got stuck on how to show a certain change in status. Someone suggested turning it yellow. I didn't have a better alternative in mind right away (wish I'd thought of something better on the spot, but I didn't), so that's what we went with. Unfortunately though, it doesn't quite match the meaning that yellow had in the previous spreadsheet. So now we have three colors again.

This is a specific instance of a common general problem: in any given instance, better planning could have made things better. But considering how much we're making things up as we go, anticipating everything would be impossible.

Murphy's meeting

We had a meeting in the doomed project on Wednesday in which very nearly everything that could go wrong, did.

First of all, I wasn't prepared for it. That's only partially my fault - the meeting couldn't have been put off, and it's not like I forgot to do my stuff, it's just that I had started it but didn't have time to finish - but I should have pushed harder to get through it in time, and whether or not I could have done anything about it, it made the meeting go a little worse.

Also, it was in an unfamiliar building to me. I had only been in it once before so I don't know the layout, and big chunks of it are off-limits to people in my division so we have to sheepishly follow authorized people through the automatic security doors. It was held over there because it was scheduled on short notice and that was the only meeting room we could find. In fairness, I didn't have any problems with security this time, and some people on the team work over there so it's better for them, but still, I think my building would have been better overall.

It was also almost four hours long. That is never fun.

And despite all that time, we didn't get much done. We had 60 or so comments to discuss and resolve. We got through about 20, maybe 30. (And note that me preparing for it involves preparing more comments. So even 30 would be less than half of the real total.) People spend 10 minutes talking about things that I think could be yes or no questions, I wasn't the only person who came unprepared, our bosses were still discussing things that could lead to substantive changes, etc.

But all that is just usual unpleasantness of bad meetings. 15 minutes from the end, though, it grew into unusual unpleasantness. The unhelpful guy blew up at another team member. Now, the unhelpful guy's first language is not English, and he speaks with an accent and seems a bit sensitive about his command of the language in general. At the moment, we had just gone over some text of his and corrected errors in the process. This is my job, I do it without mentioning it except for really egregious examples, I had been doing the same thing a couple hours earlier with another team member's product while the team was talking about other things. But apparently someone else in the room was a bit too punchy about pointing out the typos in the unhelpful guy's product. He probably meant it as good-natured, old-boys-club teasing, but it had the opposite effect. The unhelpful guy just blew his stack. Got out of his seat, began pointing fingers, complaining that he shouldn't have to put up with this because he's so close to retirement, the whole deal. Wow.