How to Keep Crappy Programmers

http://www.codeanthem.com/blog/2010/04/how-to-keep-crappy-programmers/

by AmberShah • 16 years ago

105 48 16 years ago

etherael• 16 years ago

The most depressing thing about this series is that they seem to be reflections of standard operating procedure at the median corporation.

digitallogic• 16 years ago

I've only worked for small companies/teams, and I've experienced everyone of those traits from various managers/team leads. A coralary to "You can write Fortran in any language" could be "You can manage poorly in any environment."

euroclydon• 16 years ago

I guess I'm a crappy programmer since I'm not always yammering on about wanting to test something

Glide• 16 years ago

I think it's mostly the yammer about testing to actual yammering ratio.

You also have to look at how many conversations you can bring up testing in. If you can do it for at least 90% of your conversations you're doing pretty good. It's probably not worth it to cover 100% of conversations though.

j_baker• 16 years ago

TDY (test driven yammering) would suggest that you always shoot for 100% coverage of your yammering.

mjw2• 16 years ago

Of course, if you yammer about the tests after writing them, that's just not acceptable. Yammering must drive your practise.

mattmiller• 16 years ago

I disagree with part of #3. I really like being included in some of the business decisions, or at least being aware of why they were made. I don't want to spend 8 hours a day in customer meetings, but I do like having quick meetings with BD people on why I am doing what I am doing. A lot of times it gives reasons to tasks that seem stupid. (ie we are implementing this bad feature b/c it will give us some sort of negotiating advantage)

LiveTheDream• 16 years ago

Being aware of business reasons behind decisions is very important. I can think of two main reasons. 1. Poor translation of customer intent/business goals can lead to some terrible software because the real purpose of it is obfuscated. 2. Ability to comprehend which urgent deadlines are actually important and which are exaggerated.

GFischer• 16 years ago

The thing is, a truly great technical manager will sit in most of those meetings for you, and give you a digest.

You still want to interact with the customer whenever possible, but not so much so that you can't get anything done!

roundsquare2• 16 years ago

3. The ability to come up with an even better solution for a business problem.

In my experience, a lot of business people don't really know what is possible so their solutions aren't always that good.

Example: "I need to know if I own more than 3% of a company. Give me a report that tells me what percent I own of each company. " "How about a big flashing light or warning if you own more than 3% (or set it at 2% so you know when you are getting close) so that you don't need to spend time each day looking at the report."

mjw2• 16 years ago

(Posted on the article too, but might be more use here) Nice set of articles, but a teensy bit one-sided in places.

What about the programmer who comes across as being a prima-donna about introducing some flavour-of-the-month practise, but who isn’t able to justify it in terms of measurable impact on the bottom line to (say) a small business owner. Often they’ll be right, but frequently they’ll be wrong too. I’ve often been guilty of this as a programmer–I want to change some code to make the design more elegant, add more tests etc, but if pressed it’s hard for me to say whether the improvement adds any value for the business. Sometimes it will, sometimes it won’t. Sometimes despite its value the opportunity cost may be too high. Either way as a ‘good programmer’ I will feel strongly about the matter, and it’s my boss’s job to try and temper that programmer “I only want to work with good code!” gut feel with a bit of business-oriented practicality.

So yes, the extent to which you indulge the 'great programmer' is always going to be a tricky call, especially for the non-technical, or even the technical-but-non-specialist. On the one hand you don’t want to piss off the talent–and let’s face it, some (not all!) of this stuff really is just that, in the same way a hollywood producer lays on whatever perks the best actors demand. On the other hand you need to have confidence that there’s real business value in activities that are being proposed. Or if not immediately demonstrable value, a way to measure reasonably objectively over time what impact they’re having. For example, Google (from what I hear) have an amazing system to measure (in some reasonably useful sense) the value of every test that gets written.

On top of that, sad truth is that a lot of companies don’t actually need great programmers. Sure, all else held equal they’d like them, but they’re expensive, they often want to spend your money bringing you into line with the latest and greatest–and sometimes it really is sufficient just to keep some old ball of code ticking along. I wouldn’t really want to work for one of these places, and doubtless neither would you, but it’s an acceptable deal for some programmers and some companies, and there’s a more nuanced perspective available than just to patronise them.

The other thing with great programmers is that if you don’t give us a sufficiently interesting problem to work on, we’ll often go and invent our own (whether consciously or not).

That might be a plus for someone like Google, or sometimes for some kinds of start-ups, but not so much for some more pedestrian smaller businesses.

Or perhaps behaving like this makes one not a great programmer, depending how you define it. It's not a one-dimensional quality I suppose.

strlen2• 16 years ago

> The other thing with great programmers is that if you don’t give us a sufficiently interesting problem to work on, we’ll often go and invent our own (whether consciously or not). > That might be a plus for someone like Google, or sometimes for some kinds of start-ups, but not so much for some more pedestrian smaller businesses.

If you actually like programming (as oppose to pursuing it for the money), you should only seek to work at places where you, as a technologist within your particular field will make a difference.

Problem is that it's tricky to tell such places apart from the rest. Everyone calls themselves a "technology company". Mission of the company isn't a good heuristic: there are e-commerce retailers that are run like "IT shops" and there is Amazon (a true technology company). Nonetheless, from this message I can tell that something is different here and better here:

http://groups.google.com/group/mi.jobs/msg/d81b6c1fa8f361fc?...

In most other cases it isn't so clear; you have to read between the lines. Job descriptions only asking for years of experience with specific technologies, non-technical interview questions or interview questions that only involve language syntax and APIs but don't touch on algorithms and data-structures, inflexible working hours, mandated technology stacks (with no exceptions for new projects or experimental code) are a way for an employer to signal "hackers not wanted". Fortunately these are very easy to spot and avoid, but I've been surprised when people (who have other options) choose to work at companies that require them in their seat at 8:30 and clearly under-use their talent.

On the other hand, I am afraid excessive commitment to methodology (e.g., daily stand-ups, inability to write any code without meetings + permission of the "scrum master" + index cards, pair programming, mandated test coverage percentages and "always write tests first") is another such sign, but unfortunately I am in minority.

yummyfajitas• 16 years ago

Some good signs in a job description:

    Send us your resume in raw text format. Word resumes will be discarded.
    Experience with an OO language (Java, C++, etc) is required. 
    In your cover letter, tell us about a non-mainstream technology you love.

hassenben• 16 years ago

You have to go through at least one company like that.

It is often the only way to find out what type of "employee" you want to be. (environments, coworkers, projects, schedule etc.)

Or, it will motivate you to become self-employed or launch your startup and do things differently :-)

jrockway• 16 years ago

How to write bad blog posts: Make your readers understand 4 levels of double negatives.

SandB0x2• 16 years ago

Did you mean

How not to write a good blog post: Don't permit your readers to decipher the article without unwinding 4 levels of double negatives.

WilliamLP• 16 years ago

The sarcastic tone of articles like this seems entirely unnecessary and detracts from their effectiveness to me. It should have been simply something like "Management mistakes that drive away good programmers".

erlanger• 16 years ago

Some weekend comedians think less is at stake if they make the joke easy. Going for the joke and accepting that many people will not get it takes some guts.

steveklabnik3• 16 years ago

Ugh, I just watched this happen to one of my friends. Actual quote: "I know we discussed that you like to come in at 11 and stay late at the interview, but the adjustment period is over now. I don't care how good your work is, I care that you're at the office by 8:30."

Luckily he got the hell out of there. I'd fear for his sanity.

ido2• 16 years ago

I had a boss (years ago) who would scream (!) at me if I came even 10 minutes after 9am.

I would also occasionally notice him standing behind me looking at my monitor over my shoulder (he was a 'hands on' kind of manager).

He was really surprised when i quit after a few months.

dantheman• 16 years ago

I'm surprised you lasted that long -- there is no reason for anyone to scream at anyone unless they do something incredibly dangerous or careless; you only do this so they understand the severity of their action. For example smoking next to flammable materials, improperly handling chemicals, etc...

arethuza2• 16 years ago

A few months into my first job (20+ years ago) had me going out to a customer site to fix a problem at about 11pm, getting home again at about 3am. Next morning I got a bollocking from the same guy who had asked me to go onsite about coming in 5 minutes late!

I managed about 7 months in that job before getting a research job. However, I did learn a lot about the politics of workplaces in that time.

strlen2• 16 years ago

> "I know we discussed that you like to come in at 11 and stay late at the interview, but the adjustment period is over now."

Wow, outright lying?

steveklabnik3• 16 years ago

Yep. Lot of shitty employers out there.

chc• 16 years ago

This applies to so much more than just programming. Almost any skilled production job has the same properties. Quality of output should be any manager's highest priority, and when it isn't, you've created a lousy work environment where the only people who will be happy are people who don't care.

(Just to clarify, I'm not saying punctuality is unimportant. Something that isn't in on time to be useful should not be considered high-quality work. That that's different from requiring a specific set of ass-in-chair time.)

alttab• 16 years ago

This sounds like everyone that was still at IBM when I left.

SandB0x2• 16 years ago

"Use Lotus Notes"

synnik• 16 years ago

Hey, I've built an entire career on decommissioning Lotus Notes platforms. It is a lovely niche...

trevelyan4• 16 years ago

This was a sticking point for me in my last job as a programmer, where the owners had actually instituted a punchcard system I routinely ignored. I'm more sensitive to it now that I'm running my own business and hiring people though. The problem isn't generally having a single person come in late. It's that this can affect social perceptions of what general work effort should be. Especially in a small office it is easy for working hours to creep down and soon everyone is starting work an hour late and leaving half an hour early.

It is easy to say, "I'm only going to hire self-motivated people and if they can't get the work done they're fired." But the reality is most people aren't like that. And it's a hassle to have to fire people if they can be productive with more management.

So I'm not sure what the best way to handle this is, but I think HN would be better served by an actual discussion of practical solutions rather than rants by disaffected programmers howling with rage at being asked to keep regular hours. On the employee side, I still believe doing really good work solves pretty much all of your problems if you can effectively communicate what you're doing to the manager (and put in the required time). On the employer side, giving people a bit of slack (30 minutes), a fair degree of autonomy and coming down on them hard if/when they miss their deadlines is the best I've been able to come up with. I'd welcome a better discussion of this since as a manager I don't want to have to waste my time micromanaging this stuff either.

mmt3• 16 years ago

Especially in a small office it is easy for working hours to creep down and soon everyone is starting work an hour late and leaving half an hour early.

This illustrates one of the attitudes I believe the OP is satirizing: that butt-in-seat time is equal to work time.

Programming[1] isn't a profession where one stops being productive just because the keyboard isn't under the fingers. Even if it were, the office is far from being the only place a keyboard is located.

[1] and perhaps all creative endeavors

trevelyan4• 16 years ago

Programmers aren't the only people in an office and there are social norms which affect general productivity. I don't give people deathmarches and don't ask people to work overtime (a social norm!). But I will get rid of someone if their behavior leads the rest of the team to slack off or miss targets.

Personally, I think the poster is wrong to complain about not being allowed to work 30 hours a week when he has a fulltime job, but that's his issue with his employer. In my experience, if you want to rock the boat, you've got to be able to manage the communication and signaling problems that creates in the workplace. This guy is offloading that problem to his manager and doesn't seem to understand the issues he might be creating. So it isn't that I don't understand his point. It's that I don't think his sarcasm offers any practical advice for the people he's complaining about other than "let me do it my way".

eoin_murphy• 16 years ago

Programming is a very wide field. In some cases creativity is a priority; the stereotype of the genius rock star developer hacking a business defining solution. For the majority of cases however, programming is all about communication with your team and with your stakeholders which is why core business hours are useful.

That's the problem with articles like these that they treat all programming equally. What's appropriate in one setting or for a particular developer is not appropriate for everyone.

That being said, if you can't allow your developers some flexibility and rely on them to behave like responsible individuals (re: internet access or time keeping) then the problem is with your staff and no amount of strict policies will fix that.

I still laughed at the article though.

dschobel2• 16 years ago

One thing I've seen work is to make it clear that, while you have the liberty of working odd hours if you so please, your work product will be under greater scrutiny to make sure you can "handle it".

Make it clear that it is a privilege which can be revoked. This lets your alpha programmers who like to work odd hours excel and keeps your mediocre programmers who might be tempted to mail it in from even trying it (because there's nothing mediocre devs loath more than heightened scrutiny).

trevelyan4• 16 years ago

Thanks for the thoughtful reply.

arethuza2• 16 years ago

When I managed a team of people who had "unusual" working hours I made a point of trying to subtly communicate to people who might get upset what was going on in the team:

"Thanks to Bob working to 3am this morning we are all set to give the demo to XYZ"

Do that before Bob comes in at lunchtime an he get treated like a hero rather than being resented. If Bob is always working funny hours then after a while people accept it and assume that he has always been doing something heroic when he comes in at lunchtime.

Of course, being the manager of such folks you actually do have to be able to know that all this is actually worth it from a business perspective.

johnl• 16 years ago

The article might add that having lots of department meetings is also a excellent way to keep crappy programmers.

dschobel2• 16 years ago

I'm surprised he didn't touch on "give your devs decent hardware & tools".

This should be such a no brainer but it's far and away the most common complaint I hear.

$900 on a workstation every 2 years -> happier, more productive devs.

rythie• 16 years ago

What if they want Macs though?

xiaoma• 16 years ago

6) Pay well above the market norm

Crappy programmers are well-known for choosing money over career development, intellectual challenge and cultural fit.

How to Keep Crappy Programmers

Related Stories