Agile scrum, etc.

[Pyrrho]

My bosses are considering implementing this. I distrust anything that relies on cult language. Anyone else been through the "agile" methodology? I'm really not keen on having to learn weird terminology that reads like North Korean press releases just to do my job.

"Shakin' the bushes here, ScrumMaster..."

[specious_reasons]

At a crisis point, our project tried some of the Agile/Scrum techniques. Sometimes it worked as a management tool to identify the critical problems and identify the tasks needed to keep the project going.

In at least one group, the Scrum meetings quickly dissolved into horrific, long status meetings which became a circle of blame and wasted everyone's time.

To be fair, that group was being run by a new manager who didn't actually understand the process. The "Scrum" meetings were turned over to a senior engineer - who cancelled them after calling a vote to see who wanted to continue them.

I did my part to make these Scrum meetings work by actively avoiding them unless I was specifically invited or had a specific problem that could be resolved by this group. Since these things were supposed to be fast, I didn't say a damned thing unless I was specifically addressed. These meetings don't work very well if the person who's leading the meetings is long-winded, unfocused, or rambling.

If your management has discipline, these things aren't too bad. I'm sorry, Pyrrho.

[Waluigi]

One of my co-workers has to go for his daily scrumming (or to get scrummed, or some other use of the word that makes it sound unpleasant).

Like any methodology, I get the impression that if the team buys in to the concepts and they're not too religious about all the minutia, it can be a useful way of developing software.

[Clutch Munny]

Originally Posted by specious_reasons

If your management has

...enough of a clue to implement them in the first place, you probably don't need Lean/Agile/Scrum/BlueSky/Transformational/smurf.

[Doctor X]

Not . . . that there is anything wrong with it.

--J.D.

[seebs]

I have used some of this methodology, and frankly, for a lot of stuff it's really excellent.

There are areas where it makes more sense to have a full plan in advance -- areas where there is nothing new. If there's anything new, even halfway-competent agile will beat waterfall, generally. The fact is, if you're doing something new, you don't know how long it will take until it's done, and estimates made in advance are bad estimates.

So it comes down to whether management is serious. If they are, it's probably a big win for developers.

[Pyrrho]

In my organization, they will use timelines as clubs to beat us and not as timelines that actually make sense.

I have low hopes for the future.

[seebs]

Well, if they do, that's not agile or scrum. Scrum means that you get to say "sorry, can't do this on this timeline" and drop features, moving them into the next cycle. If you can't do that, they're not actually doing scrum.

... Which is pretty common. Because managers often do not understand that the barrier to work getting done is not insufficient motivation, but managers not getting out of the way. You should see what happens to productivity when the manager stops trying to chivvy me into working harder, and instead focuses on finding obstacles in my path and clearing them.

Hint: I get more work done. A lot more work done. (Like, factor of five or more.)

[ceptimus]

Why do we have to suffer these bullshit made-up management style terms? Why can't these fuckers use ordinary English?

If we must have them they could at least be more entertaining.

[Doctor X]

Originally Posted by ceptimus

Why do we have to suffer these bullshit made-up management style terms?

One hides bullshit with terminology.

--J.D.

[seebs]

Originally Posted by ceptimus

Why do we have to suffer these bullshit made-up management style terms? Why can't these fuckers use ordinary English?

If we must have them they could at least be more entertaining.

Read the original scrum book, and it's pretty clear that this came out of engineers being given a chance to design a process that they thought would work.

Terminology helps shift mindsets. But only in people who are basically willing to risk thinking...

[ceptimus]

People have been working at doing things ever since there were people. Sometimes they work alone; sometimes they work in teams; teams can be organised, hierarchical or flat. We don't need new bullshit terms that some people misunderstand, and others have to read a book before they understand, to describe what ordinary people already know.

I'm fucking grumpy today.

[Vivisectus]

when the manager stops trying to chivvy me into working harder, and instead focuses on finding obstacles in my path and clearing them.

Funny thing: I have just listened to a VP of sales for an entire theatre give a sort of anti-peptalk to his entire theatre sales team, basically telling them that he just did not understand how they could only achieve 10% growth in stead of 20%. Maybe it is a culture thing, he said.

What on earth could he have been hoping to achieve by that? The good sales will simply fire up their resumes, while the bad ones just start to cover their asses harder.

He did not identify any core problems that he intended to have fixed, bar a few marketing quibbles. It was just a case of "you guys need to run faster" and that was it.

I assume this American became VP of emea sales because he is a fairly competent and intelligent man. Why then do something so utterly stupid and self-indulgent? What on earth would this achieve, except make him look like a pillock?

[seebs]

That sounds like he's in trouble for it not going well, and has no clue how to actually improve this. And the thing is... A lot of managers suck. If they're conveniently to hand when something goes well, people assume it's because they're good. If this happens a few times, you get viewed as really good, whether or not you had any influence...

Methodologies are not a silver bullet, but a good one can help avoid some of the usual failure patterns.

[Pyrrho]

Many managers fail at management because they mistakenly believe things get done because of them. There is a difference between a manager and a boss. Anybody can be a boss. Not everybody can be a manager. The VP described above is playing the classic boss role. If he was a good manager, he'd explain the need to increase sales and ask his team to develop solutions and he'd provide his team what they need.

I deal with a lot of bosses who place importance on getting their way instead of on producing a quality product on time and under budget. One aspect of this is developing realistic timelines and realistic expectations. Unfortunately, too many people want the moon tomorrow instead of an excellent Camembert next week, and too many people think that barking orders and acting all frustrated and impatient and freaked out is the way to get it.

[Goliath]

[Waluigi]

Code Monkey have boring meeting
With boring manager Rob
Rob say Code Monkey very diligent
But his output stink
His code not functional or elegant
What do Code Monkey think?
Code Monkey think maybe manager wanna write god damn login page himself
Code Monkey not say it out loud
Code Monkey not crazy, just proud

[seebs]

For a look at something a lot closer to "real" Agile development, go look at the history of patch notes for Rift. The devs are very prone to trying experiments to See What Happens, with the understanding that if they're unhappy they can change it. And this leads to much faster development, as well as fast fixes for problems. ... and arguably more problems. But overall I'm really happy with it.

[naturalist.atheist]

I've seen so many "methodologies" come and go I have a pretty low tolerance for them. They all start out with the germ of a good idea, and then they try to learn from past mistakes, (which is always a good idea) and then take it completely overboard. It doesn't look like scrum will be any different.

Projects that need project management to succeed are like little wars. And like any good battle plan, you do your best to plan for the first engagement with the enemy and when it begins you toss the plan. You only win if you constantly adapt to the ever changing situation. It's not about the plan, its about the planning.

I will say that things are many times better than they were thirty years ago. Back in the day there were many project failures because the technology just wasn't there. It was common for projects to go on for scores of man-years and generate a lot of code and be canceled before going operational. Nowadays the software gets finished and goes into production and either gets shelved because everybody hates it or it sinks the enterprise.

Today it's much easier to build software right. However building the right software is still tricky. There are still plenty of ways to fail and no sure fire ways to succeed. The other thing that compounds the difficulty is that there are still a large number of influential users that came of age prior to the fourth age of sand. They just don't understand the technology that well and are hoping to hold back the tide until they retire.

Anyway, for those of you just getting a taste of scrum, FYI, it's old hat. The next fad will be kanban software development. Juran is rolling in his grave.

[BrotherMan]

Kanban is old technology. They had some implementation of it at my last place. Though they had a major leaning to 5S that sort of overwrote it.

[naturalist.atheist]

5S is a requirement of kanban. It is impossible to maintain predictable cycle times without it. And yes it is old hat. I implemented my first kanban in 1984 on my electronic manufacturing floor, and Toyota had been doing it for at least 30 years prior to that. It never occurred to me that anyone would try to apply it to software development. I've always felt the there is a difference between building the same thing many times and building many different things once.

But it may have its uses in building CRUD systems. Yes CRUD systems are all different and yet they are all the same. So maybe it could work well there.

[BrotherMan]

Oh man, I totally forgot how long 5S had been around. Damn.