Slashdot Mirror
On Getting Management Interested in Improving Quality?
npoole asks: "Like many of the Slashdot readers, I am a programmer and have been pushing out repetitive database content for about a year. The work simply doesn't stop and the more we get it seems the less we ensure quality work. I have been debating telling my boss that either we take less clients, less money, more quality work or I am leaving. Is this a smart thing to do? I'm making very good money doing quick hacks to push out websites, but it's not very project oriented as much as it's become 'throw in pre-written, pre-used functions'. Any advice on how to ensure quality in our work without telling my boss it's either my way or the highway?" Of course, improved quality in any product affects the bottom line, and it's the bottom line that managers are paid to keep up. How can a developer communicate to managers (both open and closed) the value of better quality in development, and how long should one try before giving up?
They decide how much money they put into developing quality. The customer decides whether to buy the product or to go with a better product. All you get to do is find someplace cool to work. If you have fun where you work, stay. If not, don't. Maybe there are moral issues about programming hack jobs, but that's up to your conscience.
The best thing you could do would be to start up your own company if you think you could make more money doing things your way.
Sounds like npoole wants to change the quality of his workday, not necessarily the quality of the software he produces. While I'm sure we programmers can sympathize, I think he'll have problems getting the point across to management.
-- If no truths are spoken then no lies can hide --
You might want to consider getting some QA in place (if you dont have one already). Also, there is a series of books published by Microsoft Press (Yes Mickey$oft!) called Software Development Classics that can probably help. The books are: 'Debugging the development process', 'Dynamics of software development' and 'Software Project survival guide'. The most useful being the second mentioned by Jim McCarthy who has plenty of sagelike advice, some of which will certainly be useful in your conversations with this project manager you mention.
? I'm making very good money doing quick hacks to push out websites, but it's not very project oriented as much as it's become 'throw in pre-written, pre-used functions'
Is this necessarily a hack? I could easily understand how it could be boring (as noted in a post above), but I was under the impression that being able to re-use your code across multiple projects was a Good Thing (tm) -- in order to get them out the door faster, among other benefits. Just because it's a new client shouldn't mean that you should have to re-invent the wheel.
AHHHHHHH! I'm burning with goodness again!
- Reakk, Sluggy Freelance
I see two things here: One, perhaps the boss is trying to get as much work as possible so that billing can be at a high level. Second, (s)he may have the same QOS concerns you do, but has reasons not to address them at this time.
What ever you choose to do, a calm, reasoned approch is always a better way than a hot-headed, "My way or the highway" attitude. You can leave if it bothers you that much, but don't leave in a huff. It won't do you any good and will cost you later.
Necessity is the plea for every infringement of human freedom. It is the argument of tyrants; it is the creed of slaves.
Not an impossible task, but you need to consider your approach. As developers, we like clean, pretty code. However, the people that keep us fed like profit. Saying, "I'm a geek, I like it like this" will get you nowhere. Instead, push for quality control - some sort of lifecycle methodology (in which writing code is a small part of the overall process). Point out that 80-90% of the life an application is maintenance, not original development. By pushing for a structured development process (requirements, design, development, QA, deployment) your projects will come out clean and well implemented. Of course, the bottom line is profit - if the "extra" hours to ensure quality can't be translated into billable hours, there's no hope. However, whatever you do, DON'T QUIT. The market is sh** right now. I repeat, the market is sh** right now.
The best thing about a boolean is even if you are wrong, you are only off by a bit.
First a word of advice: If you're getting paid, and the company you're working for isn't about to go out of business, then strongly consider staying where you are.
Perhaps you can approach the problem a different way. You could try talking to your boss about the issues you're dealing with; it seems the worst one is repetitious nature of the code you're working on.
Anytime you're doing something repetitive with a computer it's usually boring, and it's a sign that you're doing something wrong.
Perhaps you can work with your company to develop a more abstract toolkit for your application area. If your programming lanugage/library doesn't support abstraction very well, perhaps you can come up with a code generator.
It may be easy for you to push out quick hacks. But how easy is it to write a program that can do the same? That could be a good challenge, and it would benefit the company because they could complete projects quicker. You might also get to use some new techniques or tools.
An employee who's constantly eliminating his own job is highly valued by good managers. Not that really good managers are all that common either...
In my experience, the company's reputation does not suffer. I speak mainly of web development shops in this matter, but all to often there is oversell by the salesdroids, and there is no reasonable (sometimes even technilogically feasable) was to meet the targets.
But the company has this nice fineprint in the contract: "Quote is an estimate only; billing for actual hours worked will apply", or something to that effect. Basically everything is cool until the deadline, then the client starts jumping up and down demanding their software, which they get 100% late for twice the bill.
They bitch, moan and complain, but still come back to the company because of lock-in; not contractually, but because everyone else they talk to admits that to modify (update, maintain) the application will be difficult and costly for them without the original company's knowledge, design docs, etc (which, incidently, aren't part of the software and not purchased in the contract - bummer).
I know this goes for the UK and South Africa, and I'm sure there is at least some incidence of it in the US. "Churn it out fast 'n ugly, 'cause then they will pay more to maintain it" is the general rule. One day companies will catch on to this extortion, but they aren't at the point yet.
I know a number of damn good programmers who are simply not allowed to produce good code, because the company feels it is a waste of time. Sadly, the companys where they have produced good code have gone under in the dotcom slump ... why? Because they took longer to do the job, and (sometimes) quoted higher in the first place. The perception is that anyone who can quote can do the job properly, so you go with the lowest bidder or the most established.
/end_rant.
i-name =twylite [http://public.xdi.org/=twylite], see idcommons.net
...I would suggest that this economy is no longer the kind of economy that will support an employee dictating "my way or the highway". It is very likely your supervisor will pick "highway" even if you're very good, because there are many, many highly-qualified candidates now coming into job interviews. It wasn't like this a year ago. I realize other slashdotters may challenge me on that, because it's not a very nice thing to tell someone that they're possibly expendable. However, your boss may very well think that way, regardless. So be careful.
In addition, the "good salary" you claim to be getting may be due to the fact that you're churning out sites fast but charging the same rates you did back when you custom-built them. By asking to change the process, you may be getting a change in salary too.
Finally, don't forget that object-oriented, modular programming is supposed to make cookie-cutter work possible. If you're reusing your code over and over, sure, it could be sloppy, careless work, but it also may be that you've got a system working well and just object to the monotony more than the code. If that's the case, ask to be put on different projects, rather than taking a hard-core "reform-or-I-walk" stance.
My Greasemonkey scripts for Digg &
In just about any organization you have leaders and you have workers. It sounds like your boss is saddled with the responsibility of being a leader, and you have the role of a worker. For that reason *alone* it would be well worth your time, money not to stick your opinionated nose in where it doesn't belong. Don't get me wrong, I'm on the same side of the fence that you are, but rarely does an approach like you suggest end up making any change for the better for *anyone* involved. It may be that your company makes more money doing quick hacks; in the long run if they think they can make more money doing project based non-hacks, they will. If in the long run your clients finally realize that more careful planning up front is worth it's weight in gold down the line, they will go that route too. You can't control the average intelligence around you, you just sound like the fool on the hill... trust me on this one.
That said, I think a better way to look at this is
a) ignore the money aspect (both yours and your employers) Always trust that a business will do the thing that makes it the most money. You won't change this in the near term. If the money is that important to you, you should either stick it out or try to find a job that pays similarly doing something you *enjoy*
b) If you aren't happy doing what you are doing, look for guides on the web that give professional suggestions about how to bring it up, what to do and not do, etc. with your boss. do NOT just give an ultimatum, especially not in public company
c) If in the end, after rationally sitting down with your boss and explaining your position in a professional manner, you still aren't happy with the work, and your changes don't make business sense for them (even if you still know you are right -- you can lead a horse to water and all that) maybe you should consider leaving! It doesn't do much good to be in a job you don't enjoy.
A computer without Microsoft is like ice cream without ketchup.
Most of what I do doesn't involve databases but networks of terminals (serial, RF, and PC UI) are becoming more and more important to us. I have just defined a general-use flat data structure that allows virtually anything to be related to anything else within the sphere of what we do, without a DB engine and very fast, and with the ability to add virtually any kind of record to an existing set on the fly. It's harder to code this than a fixed-record field for a particular customer's app, but I only have to code it once, and then I can use it for everybody. In the long run, it will clean up a lot of old spaghetti while greasing the path for new jobs.
Brackets contain world's first nanosig, highly magnified:[.]
You're paid to program, you're not QA.
Quality is the responsibility of everyone. Do you think you can write whatever and however you want and expect the QA department to magically instill quality into the product?
QA assure quality, they don't create it.
Now, if this bothers your conscience or you just prefer to not be associated with "good enough" work, do seek employment elsewhere.
If the client hasn't budgeted for overages due to quality assurance, your boss will ignore your pleadings.
-- @rjamestaylor on Ello
Let software quality make your products stand out from others, not features. There are many "elitist" companies in every industry that use this tactic. Take the Leica camera company for example (I'm into photography). They make some of the simplest featureless cameras in the market, yet they are the most expensive cameras in the market. Why? Because people don't buy their cameras for features, they buy them because of quality.
You can do the same thing with software. Make it nice and simple, but make it stable and fast. Take the basic and most important features that people use all the time, and make them work the best they can. With a good solid base system, minor features can easily be compensated for or even forgotten.
Outdoor digital photography, mostly in New Engl
I have worked for a small software shop for 3 years now, and we recently started implementing Extreme Programming into all of the development.
Ths difference has been astronomical. Deadlines are more realistic, the our releases are far more stable, and basically, the whole "chaos" of development seems to have taken on an organized form that makes everyone happy, even our extremely hyper CEO.
It's customer-driven, it's organized, and it has simplified life at the company immensely. I'm not a shill for it, I'm just relating our experience.
I would highly recommend picking up a couple of the XP books. There are several chapters devoted to how to sell the idea to your bosses, as well as recommended means for showing them how well it's working, which is essential when dealing with people who are obsessed with ROI.
If you have to threaten to resign over quality, you're working for the wrong organization, period. Don't quit in a huff, just find another job. You're working for an organization where non-technical people are calling the shots on how software is developed, and they aren't accountable for when there's problems with the code. Since when is buggy code the manager's fault? If you push unreasonable deadlines and get buggy code as a result, you can always whip the programmers because it's their mistakes that caused this problem.
As a rule, I will *never* work for any organization where project management is in the hands of people who are not technically current. They are a non-entity at best, at worst, they can ruin the lives of their employees. I've interviewed enough of these "manager-managers" that I find their attitude towards managing the software development personally offensive. Mostly it's focused on perserving their authority and cracking the whip over their staff. "Beat the rock harder, beat it faster, more guys with sticks! Oops, wrong rock! Beat that rock over there!" No wonder so much software sucks out there. It's even worse that in most projects, the requirements are pulled out of management's hat, and have little to do with actual customers desires. How many of you have had to re-work a piece of software because of poor RA?
Now that I've ended the rant, enlightened management is always concerned about quality, and at the very least, is running the software on a regular basis to keep their thumb on the pulse of the software, and they have a good feel for what the customer experience will be. Customers don't reject products because they're short on features, they reject them because of crashes.
I strongly recommend Tom Gilb's "Principles of Software Management", which describes phased delivery of software, and rigorous requirements analysis. It's a forerunner of "XP" which is the latest trendy book, but it seems that some managers use "XP" as an excuse for "open-plan" offices, and more shouting, and kinda forget the automated testing and refactoring parts...
Since the mid-60s, my father has been a huge proponent of Quality Assurance and Total Quality Management, having followed the teachings of W. Edward Deming long before even General Motors had taken a liking to him (Deming, not my father). Since I was very young, I knew that my father's job was to make companies make better products. Sometimes he'd cost a company a few hundred thousand dollars in new quality programs that would, in several years, pay the company back millions of dollars in decreased support or re-work costs.
I also knew that when the United States fell on hard times (relatively so, like in the 70s, early 80s, mid 90s, and now), my father would inevitably spend several months looking for a new job because the companies he worked for could no longer afford the overhead that a Quality Assurance program introduces. There was never any question of a Return on Investment in quality, but there was always the question of how much cash the Quality Programs required. What's worse, Quality Assurance is a cost center: cash flows in but revenue never comes out. Most improved processes in all parts of the company can no be directly tied to an increase in revenue or a decrease in costs, so even though people understand that Quality Assurance is something beneficial, they don't know how to quantify how beneficial it is.
Because of this, when a company needs to tighten its belt, Quality Assurance staff are the first out the door.
It's a great thing to get management interested in improving quality. There are many people who truly believe the principles that were taught by W. Edward Deming, that are awarded by Malcolm Baldridge, and that are supported by the ISO 9000 certification process, but given today's economic situation, now is probably not the right time to be bringing this up with your management.
Oh, and if anyone knows of any upper-management positions for a long-time Quality Assurance guru with an impressive track record and who's been through the ISO-9000 process many times, send me e-mail. My father is, yet again, looking for a new job in the Los Angeles/Orange County/San Diego County area.
--
For example, Microsoft is the biggest software company in the world specifically because they realized early on that consumers are less interested in getting quality than they are in getting something that meets their perceived needs.
This is an important point, because unfortunately the most important aspects of software quality are usually hidden from the end user. Since most consumers actually do realize that they don't know jack about the inner workings of their software, they elect not to get inovlved in the esoterica of which kernel is more stable or which file system kicks ass.
Let's make the assumption that your company's clients aren't consumers. Let's say they're aerospace engineers. Smart peope. People concerned with quality. But the quality they are most concerned with is the quality of their own work. So their attention is primarily focused on how well the software you build for them will help them create quality aerospace products.
Also, keep in mind that the cash that runs your client's company and your company has to come from somewhere. Cashflow can often be a huge issue for clients. If a client knows that they can spend $50k for something they know they'll get in an imperfect form one month from now, versus $100k for something that they know is more elegant more cost-efficient over the long haul, they may elect to spend the $50k because their short-term financial concerns dictate that they deal with a lower quality solution now if it will give them just enough to do what they need done.
Even if your manager understands that the quality way is better, more likely than not she'd have a tough time convincing the client of that. In fact, in my experience, clients often don't really want to hear about deep quality issues. "Just tell me the tradeoffs, and I'll make the decision" they'll say. They simply don't want to really know the nitty-gritty details. "That's what I hired you for!"
The Bottom Line is money, and if your software is good enough that your clients can make money with it, most of them will vote with their pocketbooks. Most companies simply follow the buck, from quarter to quarter.
One thing you might try is to evaluate your next employer not just on the work conditions and pay, but also on who their clients are. There are companies out there that actually think long-term. An excellent book on the subject, "Built to Last" goes into detail about the characteristics of visionary companies. Yes, it sounds like cheesy business-speak crap, but these guys conducted extensive research, and they avoid easy answers.
Find a company that serves long-term oriented clients, and you may find yourself a lot happier.
Read the EFF's Fair Use FAQ
The problem could be that it takes time to build this kind of good, flexible, reusable, modular components. I've done the same thing myself. But rarely can this be done on employers time (who do you bill for the time ... when there's enough work to keep everyone busy). One might try to argue to management that if they spend a couple weeks putting together some slick modular tools, that over the course of the next few months it will pay back well with even faster deliverables. But when business is rolling in and customers are saying "the other company promised it in 3 days, but if you can deliver it in 2 days, you've got the deal" then management is loath to pay people for what to them seems risky. The answer may be to put together those tools on your own time, put them on some not-well-announced project on sourceforge using a "pen name" as the owner, then come to management one day and say "Hey look what I found, I think we can use this and speed up our work. We should try this out before anyone else discovers it". And the fact that it is already out there on sourceforge would prevent them from trying to take ownership of work done on your own time.
now we need to go OSS in diesel cars
managers (including the poster's) evaluate the costs and benefits of QA. the benefits of not pursuing quality include lower dev costs, and a shorter turnaround on investment. the costs of not pursuing quailty include customer churn, bad image, tech support costs. balancing these costs and benefits and their attendant corporate politics is probably quite tricky, so the manager therefore probably won't be that interested in being told how to do his job by someone who doesn't know how the company works (unless it's a blindingly obvious way to reduce costs - such as reusing old code ...).
As an example:
If your work is repetetive, this indicates to me that there is room to automate parts of it. You might talk to your boses about setting aside 10% of your time to improving the website creation software. This could make your work time more interesting and make the company more profitable.. a win all the way 'round.
What's possible is only limited by your imagination. Just remember that the easier you make the change for your management, the more likely that they'll agree to it.
Free Software: Like love, it grows best when given away.
That was a huge mistake. I was "talked to" by several people above me, and my superiors wondered if I was "on crack." When I tried to explain my standpoint, and how quality would improve and six-day workweeks would be unnecessary if we could produce more quality work on a consistant level across the development teams, here's what I was told:
So what it comes down to is that the profit margin is the bottom line, always, and the beauty of the insides of the machine you're building take a backseat to doing things the way they've always been done, as long as everything gets done on time.
Is it crappy? You bet. Am I comfortable working this way? Not at all. But like everyone else, I have bills to pay, and I'm looking forward to a future where I can start my own small company, and run things in a manner that I'm comfortable with. It's a sucky situation, but the more I learn about anything, the more I learn that the bottom line is always the trump card in every situation.
It's also true that the market does suck, but smart people will always be needed, and if you're smart enough, you can find a way out that both benefits your career and improves your workstyle. It may not be this week or the next, but it will happen.
This is one of the reasons I refuse to do work as a developer in most cases. My professional work is systems and network administration, although my 2 decades of programming does (except in the last few months) get me plenty of calls to come interview for some programming job. Right now I have enough cash and contract work to hang on for the next 2 to 3 years, so I'm working on developing something for my own business idea. Interestingly, while I thought I'd have plenty of time to do it right, I even feel my own pressure to get it done and get it to market sooner. That is a fact of life. The concept of getting things to market sooner really is valid from a business perspective, despite how much the geek in me says that's crap. At least during the economic downturn, it's less likely that competitors are doing as much that will compete against me, so I can take a little time to get this done right.
now we need to go OSS in diesel cars
I'm not sure what code reuse has to do with poor product quaility, unless the modules themselves are broken somehow. What you've described really sounds like a development manager's wet dream. Drop-in, pretested modules with a minimal amount of modification? That's the holy grail of the coders-as-cogs management mentality! If the customers themselves aren't complaining about quality, I doubt management is going to give a hoot what the rank and file thinks.
If you really insist on pushing this, I've got a few pointers for you:
Best case, management addresses your issues, and you look like a "team player". Worst case, they drop kick your arse out without even giving you a listen. The outcome depends on how you play your cards.
There is a trilogy of dimensions at the core of the issue: cost, time and quality. Every organisation needs to balance these. The management of your organisation, like most I have worked in, don't understand the quality issue in relation to software because it is more subjective than the other two dimensions and therefore it doesn't get the emphasis it deserves.
From the little I know about your situation, here is my take on what you need to do:
But the basic form of the business case is:
Another way of saying the same thing is:
If you have other people who have the same focus as you, pool your talents and resources together.
If you want this change and it's important to you (which it sounds like), then you need to put in some work to make the change. Don't make an ultimatum because it's an employer's market - they can just take you up on it and that won't help anyone, especially you.
Remember, anyone can be influenced if you can show them that what you want makes it better for them too.
I hope this is in any way helpful. I have had similar battles myself and still do, but life is always slowly improving!
Mark
Define "quality" before trying to make "quality" code. "Quality" doesn't mean one single thing. It can and does mean different things to different people. I've seen people use "conformance to spec", "fully documented", "feature rich", "crashproof", "fast", "easy to use", "surprising", "first to market", "bug-free" as all or part of what "quality" means.
Figure out what you mean by "quality", then find out what your boss means by "quality". You may be talking across each other. You might want to look at Gerald M Weinberg's Quality Software Management for a better discussion of the meaning of "quality". I'm not sure about the rest of the book, but the section on what "quality" means is relevant.
My other advice: ignore consultants and companies who peddle a Process (a process to reach SEI CMM level 5, or ISO 9000 status, for example) as a means to acheive "quality". They often leave "quality" undefined or vaguely defined because then they get to use opposing meanings as convenient. When convincing programmers to use The Process, quality consultants will use "bug free" or "speed to market" as the implied meaning of quality. When talking to managers, they use "feature rich", "on schedule" or "completely documented" as the implied meaning of quality. When talking to corporate leadership, the use "cheap", "speed to market" as meanings. Often, some tension exists between various definitions of "quality". "Cheap" often opposes "bug free" or "fully documented". "Feature rich" can oppose "high performance". "Speed to market" can oppose "fully documented". You get the picture.
The day I have to "be thankful to have a job at all" and not speak my very experienced and bright mind as I see fit is the day I stop programming for a living. I wouldn't be worth a damn without being able to speak up and actually make a difference.
Not taking a stand, if done by enough employees, guarantees that your job, or even your entire company will be the one with its head on the block next.
Not caring about quality because your company doesn't is a most excellent way to hate your work and lose your spark utterly. Don't do it. The paycheck isn't worth what that will do to you if you take such advise. I know what I am talking about.
'As a rule, I will *never* work for any organization where project management is in the hands of people who are not technically current.'
By far, the biggest problems in technically-oriented companies are the non-technically-oriented managers. They are generally making far more money than they would at a non-technical company. They are willing to do anything to keep their jobs, including making life miserable for everyone else.
The best acting I have ever seen was not in a Hollywood movie. The best acting I have ever seen was by a manager trying to make everyone believe that he could manage without thorough understanding.
When they sink their companies, they are generally able to get another job, because the people who hire them are faking it, too.
The dot-coms failed because they hired good actors and not knowledgeable people. The dot-coms did not fail because of highly complex situations that could not be understood in advance. They failed because they did extremely foolish things.
The use of non-technical managers will continue as long as there are investors who will put money into something they don't understand.
Bush's education improvements were
Dude, trowing out pre-written, pre-used functions is what good software engineering is about. That's the whole point of good design, preventing you from having to re-invent the wheel every time.
If you're getting bored, try to get higher up the design hierarchy instead of sticking in your lowly code production job.
It's a hack when you have to write from scratch essentially the same thing again and again. NOT when your are re-using stuff. Get with the program.
In 1986 David L. Parnas and PC Clements published a paper entitled, A Rational Design Process: How and Why To Fake It. Parnas and Clements present a strategy for imposing overlying order upon the often fractured development process; the goal of which is to produce better software. Doing snippets of work for managers/clients who don't care about quality as much as they care about costs is often a cause of this fracturing.
I couldn't find a copy of the paper online, but it has recently been re-published in Software Fundamentals: Collected Papers by David L. Parnas.
Money.
Let's face it: If you talk technical details most managers will get that blank stare after a while. Either they don't understand what you are talking about or they are not interested, probably both.
But if you can show them how to save money on the bottom line, they will listen to you. And yes, you can boil it down to money. Better code means less time spent on correcting errors, time which most likely is not billable. If that time (a.k.a. cost) can be removed or better yet converted to billable time, it will affect the bottom line positively.
So in order to be allowed to make better quality, you have to calculate how much it will save on the bottom line.
Also, be honest and don't oversell your stuff. You may think that what you propose will earn say a 20% saving. Tell your manager that the saving will be at most 10%. Why? Because you will most likely run into snags and teething troubles that will diminish the initial savings. And if you don't, well, performing better than promised is usually not a sin.
My opinion? See above.
The point was to emphasize that when a pizza is delivered to my house, I don't care if it was cooked in a Vulcan or Middlebe-Marby oven, if they used X or Y management style or if they were driven by this factor or not.
All I care about is that if the local manufacturer of circular pseudo-Italian cuisine:
delivered it fast enough;
delivered it as specified;
wasn't stuck to the top of the box;
didn't taste like the top of the box.
... yet amazingly enough, even in those cases, I didn't give a fig about how the restaurant was managed.
I mean if I wanted "quality" Italian, then I'd get dressed up, take the wife downtown, spend some time and some bucks
healyourchurchwebsite.com - WWJB?
Your boss wants
The trick is to find an approach which fulfils both these sets of goals. Several exist, but the most obvious one is to work over the course of several projects to turn what you have (which you say are all very similar) into an actual product. This means
Your boss gets
In addition you both get
To successfully sell this to your management you will need to be able to demonstrate that this can be done off the backof your regular project stream, and does not have to mean that some guy gets to sit in a corner contemplating his navel "writing the product" for a year. You will never sell that to your boss. Instead, devise a plan where you use the code for project A, and generalise it a bit and add customisations to support project B. By the time you have delivered C and D as well, what you have is a product.
To make this work, you will have to retain the IP on the software you write, which I guess you don't at the moment.
The best way round this is to tell the clients explicitly that they are getting a product (clients often like products because it means that the project delivery risk is reduced).
But the client will refise to allow themselves to be marooned without support. Hence you do a deal with them whereby they get a non-exclusive license to modify the code which is transferrable if the business is bought or sold. You can also tell them that this means that they are free to seek support for your code from elsewhere (but that they cannot sell the code on). They may well like this (it has several good features, e.g. insulating them from risk of your company folding). Your support agreement will need to be clear on the fact that you won't support the code if they have just hacked upon it madly.
In short: develop a strategy that benefity you, your boss and your clients, and think about hoe to sell it to all three.