Slashdot Mirror
Slack
Summary: A highly entertaining, and informative survey of the state of the high-tech and software industries today, which suggests that companies have been taking exactly the wrong actions under pressure and further decreasing their ability to handle rapid change. The book is peppered with interesting asides and examples, but is always informed by the central thesis that companies need more Slack built back into their structures.
Check your sources. Tom DeMarco is an established industry management guru who has the respect of many of the technical community. He's written several previous titles, including the notable Peopleware and the collection, Why does software cost so much?. I'm not normally keen on any books in this genre, but have always found DeMarco's writing very readable and though-provoking -- most importantly for me, he has a habit of trying to find NUMBERS to back up any claims.
What's this book about? This is a 2001 title, and I find it slightly shocking that, in a maturing industry, we still need a book on this topic (from the blurb):
"To most companies, efficiency means profits and growth. But what if your 'efficient' company - the one with the reduced headcount and the 'stretch' goals -- is actually slowing down and losing money? What if your employees are burning out doing the work of two or more people, leaving them no time for planning, prioritizing, or even lunch? What if your super-efficient company is suddenly falling behind?"
So far we're just talking about the state of the modern software industry right? What's he proposing we do about it?
"[...] what you need is not more efficiency, but more slack. What is 'slack'? Slack is the degree of freedom in a company that allows it to change."
It seems a very simple concept to me, but then I'm an engineer, his writing is persuasive, and I have the benefit of 20-20 hindsight when reading. How can he get a 220 page book out of such a simple concept? After all, all we programmers know that your general purpose solutions always sacrifice speed for flexibility right?
What he discusses is a business model where you keep people, say, 70% busy. This leaves time for unexpected business, for reflection on why X takes so long and how to fix it, for self-training, for discussion about how things are done. These are all good things -- but the winner is that when people are stressed by sudden change or a deluge of new work, they have some slack to take in. Things change, you suffer a reduction in productivity, but hey, you had some slack to take in so the week's work is still getting done, you've just dropped that Ruby book for a week or two. You're swamped by a rush on finishing Product X before a competitors Product hits the market first -- just drop that tinkering with a novel memory pooling thingy you were considering slotting in to replace the adequate-but-inelegant solution in your product. I'm simplifying and reducing his argument here, but that's the idea. The other corollary to the 70% busyness level is that the system is responsive -- some nodes are 100%, some are 20%, but overall things are flowing. A system where most nodes are at 100% means some nodes are hanging waiting for other nodes to catch up -- total throughput drops. This'll make more sense reading his version ('underworked but responsive secretary' vs '100% busy, cannot help until Friday secretary'), but it's a good central topic -- simple, but not trivial.
220 pages isn't much -- he states that the book should be comfortable reading for a business trip -- and the bulk of the space is taken up by rationale for his suggestion, and discussion of the consequences. What I found valuable about the book was the description and subsequent debunking of several management techniques -- for example, he has a severe go at management-by-objective. I recognise it. I suspect you too will recognise it, and several other common variations.
Let's have a quick skim of the contents -- this isn't a technical book, more one massive opinion column, so the section titles aren't that useful, but I feel like I'm cheating if I don't do this in a review ...
- Slack
Madmen in the halls, busyness vs business, the myth of fungible resources. This section sets up the case by setting out the assumptions, and describing what actually happened to most businesses when put under pressure in the last 10 years. I loved the word "fungible" too -- describes a resource that can be freely interchanged -- like paperclips are and software designers aren't. - Lost, but making good time
The cost of pressure, aggressive schedules, overtime, culture of fear, quality, management by objective. This is a meaty section and basically describes how the heck things got to be this way, what practices were adopted, and how they made things worse... - Change and growth
Vision, leadership, fear and safety, trust, what middle management is there for, change management. This section talks about change, specially why a lot of the measures adopted to prepare for it help make things worse, and how we should instead consider other approaches. - Risk and risk management
Working at breakneck speed, learning to live with risk. This seems like a short section from the contents, but it's reasonably long. There's less to discuss here for what we have is a 2-by-4 to head of businesses who refuse to plan for failure. A discussion then follows of the classic problem -- scheduling -- and why you'll never do a decent job of that without risk management. This is the only section where the tone is hectoring rather than persuasive -- or else that was my own frustration at the experiences I've had coming into play!
Target audience It's aimed at a particular segment according to the cover: "A handbook for managers, entrepreneurs, and CEOs." Well, I'm none of those, but I enjoyed it and found it useful. I'd prefer that my bosses were reading this than most of the other pap from the same shelf, but let's face it, change comes from all levels in the organisation, and if you can't spot mistakes being made within your team then you can't plan for your own career either. Read this book, it'll come in useful either when your managers start going awry and making you suffer, or it'll come in useful when you float up the org chart and have to start dealing with a team of your own.
What's good? Most of it. This is a highly entertaining read, and does present some genuinely useful ideas. It's also great as a collection of management anti-patterns. I think any career programmer in a medium-sized or above business would find this book interesting. Actually, come to that, anyone who enjoys Dilbert will enjoy this book.
What's bad? Not much. There were a couple of areas where I would have liked more case studies or evidence. As I said above, the recourse to surveys for the truth is something of a trademark of DeMarco -- he certainly references quite a lot of material in this book, but doesn't produce any solid evidence to back his ideas. Granted, probably hard to experiment on this scale!
You can purchase Slack from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
And a history of corporate suits ignoring what he says. If you do read and implement what he says it will save you time and make you money.
These are the great books of Software Engineering written by people who know, and can prove it. headed by The Mythical Man Month and Peopleware everytime I re-read them it depresses me. Another year on, and still the same mistakes as 30 years ago.
An Eye for an Eye will make the whole world blind - Gandhi
I have first-hand experience with this kind of work environment at my University job. It has really helped me produce much higher quality work (I'm a Web Designer and Developer using Perl, PHP, and MySQL). Of course, in my case it's somewhat accidental, but nevertheless I see a lot of this at the University.
The culture here is such that people are hired to handle a set of responsibilities rather than to produce 40 hours of solid work every week. Because there is no one clear goal in most University departments, you find a wide disparity of workloads.
I think there is one crucial distinction between people that needs to be judged before such a management is widely deployed, however. There are some people, when given spare time, will increase the quality of their work. Others however, will simply waste their extra time. I'm inclined to say that techies, being generally more interested in their work than the average full-time employee would fit into the first category. Upon reflection, however, I do not believe this is true. I think it just boils down to personal work ethic. I've seen people in what I consider to be dreadfully dull positions (retail management, facilities) coming up with all kinds of great ideas to further the goals of the organization. As with many things in business, hiring seems to make all the difference.
Yeah, hiring 3 people to cover a department that should have 10 people is monetarily cheaper since you only have to cover salary and benefits for 3, the overtime and burnout will bite you in the butt in the longrun.
Where I now work we have exactly THREE people to cover a backlog of tickets (some going back almost SIX MONTHS) along with the current issues of a 10 building, 250+ computer WAN. They wonder why we get stuck working a bit of OT (average of 1hr/week - and that's usually divided between the 3 of us), but they also expect us to get the department totally caught up (hey - there's 3 of you now instead of the just 2 of last year).
The world really needs to kick a few of those highly paid corporate officers out of their palaces and make them work a week or 2. I bet that would let us start seeing a change in working conditions (or at least pay).
I don't mind WORKING but this whole "we're going to cut your department, and your pay but you sill have to get everything done on time without overtime" idea is nuts.
We didn't get it from any book. My boss is just a laid-back guy. Hell, until business was picking up to the point that it is, people were drinking every day... he offered me a beer at 3:00 on my second day!
Now, business has picked up, and before our last two large projects, he's hired somebody to help me with them. Now, I've got a close-knit team of 3, and I'm still doing the same amount of work as always. I get a little stresed about busy weeks, but a "busy" week usually means cutting the hour of Unreal Tournament, coming early and leaving a little late -- not working 80 hours a week. As a result, I'm always "on". I don't feel burnt out. I even enjoy my work most of the time, though it can be monotonous. (web scripts are all the same after you've written too many)
I think it just depends on the temperment of your particular boss. I work in a marketing department with 4 non-techies (well, one is a graphic designer), and I LOVE the degree of control they give me. Of course, it only works because they trust me, but I would not trade it for a job working with techies anytime soon.
(web scripts are all the same after you've written too many)
;-)
This is a big hint from the universe that you need to abstract further. Spend some time factoring out the similarities, and you can make those drudgery scripts more quickly, with fewer bugs, and move on to more interesting problems. Plus, the challenge of factoring the functionality is itself an interesting problem.
Just trying to be helpful; I have no vested interest in you listening or otherwise
How the heck is that unethical?
/. poster anything. It's exactly how affiliate programs should work--he referrs them to Amazon, and if they buy from Amazon he gets money for it.
/. did the same thing...
RedWolves2 pointed out to a populace reading a book where they could save some money on it. Pointing them to his affiliate page doesn't cost the hypothetical
Besides which, it looks like
What is the big deal? I mean its not like it makes the book more expensive for you. He is doing exactly what the affiliate program is supposed to do... bring business to amazon for a small cut.
Spencer Ogden
Slashdot should have a policy allowing moderators to discard posts from people who pull this kind of crap.
/.
This is great. We've got a microcosm of the whole Libertarian-accepts-sketchy vs. Ethical-demands-authority debate that always crops up on
RedWolves2 is a whore. I hate namecalling, but that's the fact. As soon as I saw him shill for Amazon, I checked to make sure that he had included his affiliate code in the link, and sure enough. He will make some money on being the first 3+ post on the article, with a link (unless he subsequently gets modded down). I have a friend that, despite my chiding him, maintains an Amazon affiliate link for a book that he recommends as his sig, and pulls in a modest ~$15 a month. That's not even half as shady as RedWolves2 though.
What upsets us about RedWolves2's action? The fact that he didn't disclose that he had included his referrer tag? I'm not bothered that he referred Amazon for a discount, and it would only make sense that he'd include *someone*'s affiliate code. If he was a true humanitarian, he would have picked a worthwhile organization that had an affiliate code and used theirs, and disclosed that fact. No, I'd say it's the fact that he used his own, and didn't disclose it, that has got our hackles up.
That being said, that it's fairly unethical behavior that many of us would find distasteful, do we need a new method of dealing with it? He was modded up by folks who apparently wanted to reward his post, dubiously in the public interest. He could be modded down by people who want to punish it. Ultimately, you'd believe that most people clicking through his link would understand the Amazon referral program, and would realize who they were rewarding. Maybe people just keep a list of affiliate codes for worthy charities by their monitors for just such purchases.
Some AC mentioned that the Tattered Cover is a more worthwhile book store to support, and that they have it for $14. They might have an affiliate program, perhaps someone should find a decent organization and buy it from there.
I just double-checked, and sure enough, RedWolves2 has already lost 2 mod points. Looks like the people are speaking...
"Premature optimization is the root of all evil." -Knuth
Companies, especially in the cut-throat US market, consistently choose immediate gains over long term gains. This is why we can have billion-dollar corporations just crumbling within days. At some point you can no longer borrow from Peter to pay Paul, and it all falls apart. Companies should be looking not only 1 or 5 years (or god forbid, just months!) into the future, but 10 or 20...not only with respect to human resources, but all the other resources and strategies available. Unfortunately, when you are surrounded with competition which will gladly eat your lunch if you attempt to forego immediate efficiencies for long term efficiencies, this can be very hard. Somehow this premature optimization needs to be disincentivized, but I'm not sure how that can be done. Also, with such "premature optimizations" the damage is long done before the long court process can resolve any wrong doing (HOW many years has the MS trial been going on without any ramifications or reparations so far?) Perhaps corporations should be forced to submit long term business strategy documents or have their charter revoked (maybe make this public record, so that companies cannot eat each other's lunch?) Who knows. But it a larger issue than just human resources. The free market optimizes very locally (and while some may argue the failure of those that optimize too locally, and the subsequent emergence of other companies support, not detract, from the free market - remember, big giants make BIG fucking holes when they fall...maybe we should be wary of letting the giants get that big without looking where they are going)
It's 10 PM. Do you know if you're un-American?
Another important aspect of slack pops up in Repenning, Understanding Fire Fighting In New Product Development; without it, the system is unstable against transients that tend to draw resources from upstream development (where they are more efficient) into downstream firefighting (where their benefit is more timely). A death spiral results.
Decent paper. Yes, he tends to belabor points the reader should see coming, and the model is clearly simplistic - but these very points might make it accessible for managers, particularly those still enamored of their MBA degrees.
I decided that behaving ethically was the most nihilistic thing I could do. - Paul Pavel
It'd only be unethical if he had defrauded people about the book's value or done so in a story submission.
;)"
Hmm. Why do you draw a distinction between shilling for amazon in a story submission and doing so in a post? I think it's the same behavior, either way. The only thing that would make it worse in my book about a submission would be that the behavior had the tacit approval of the story's editors.
I think what rubbed some of us wrong, including myself, was that he sounded like he was motivated to help slashdot readers to save a couple bucks, but upon realizing he's trying to reap 15% of a large number of sales, you realize that he has much shadier motives. You wonder why he didn't disclose it to begin with. It suggests that he didn't want people to know he was shilling, that he had a financial interest in the behavior. In the US, we usually require our politicians to disclose conflicts of interest in the stocks they hold when recommending certain companies for public works contracts, or journalists to indicate whether they have any interest in a story, such as when they're reporting on a company that owns or is related to the company they work for. It's just integrity.
Do you have anything against people making a buck?
I think he could have had it both ways, by saying something like
"Buy from Amazon, it's only $16! Click my link (don't forget I referred you
Works for me. I might have bought it from his link if he'd done it that way... that would be a great way to quickly make a lot of bucks, without compromising your integrity.
I don't usually post on stuff like this. Redwolves2's posting history seemingly speaks to the "pimping Amazon to Slashdot users" angle.
i d=4447040 i d=4445470 i d=4426834 i d=4426775 i d=4424239 i d=4420982 i d=4416765 i d=4411581 i d=4409864 i d=4404197 i d=4375345
My opinion is that continuously trying to hijack referral benefits away from Slashdot is not ethical, and damaging to the community as a whole.
Here in all it's naked glory is evidence that half of the most recent 24 of Redwolves2's posts are attempts to drive Amazon affiliate benefits to Redwolves2:
http://books.slashdot.org/comments.pl?sid=42004&c
http://books.slashdot.org/comments.pl?sid=42004&c
http://books.slashdot.org/comments.pl?sid=42014&c
http://books.slashdot.org/comments.pl?sid=42014&c
http://books.slashdot.org/comments.pl?sid=41395&c
http://books.slashdot.org/comments.pl?sid=41811&c
http://books.slashdot.org/comments.pl?sid=41392&c
http://books.slashdot.org/comments.pl?sid=41789&c
http://books.slashdot.org/comments.pl?sid=41393&c
http://books.slashdot.org/comments.pl?sid=41660&c
http://books.slashdot.org/comments.pl?sid=40460&c