Matt Childs retweeted
Mon copain rebrode des vieux canevas pour y ajouter des éléments de jeux-vidéos et son rêve : faire une expo dans un event/conv JV ou autre. Si vous avez des pistes ou même juste un petit RT, lezgong !! 🔥
162
1,487
5,316
382,053
Matt Childs retweeted
The Toyota Production System (TPS, usually called "Lean" in the West) defines three sorts of difficulties, all of which apply to software development: Muri, Mura, and Muda. They're all worth thinking about.
Muri is long-term overburdening of the entire production system (e.g., too much work for the people at hand). That's not waste so much as creating a system incapapable of doing the work. To my mind, Muri also encompasses inappropriate organization architectures, where the management structure is fighting against working in effective—all too common in larger corporations, but you see it everywhere.
Mura is variation in the work to do. The biggest cause is usually random show-stopper bugs in the drop-everything-and-fix-this category. Variation can also happen at a macro level, for example, tax-related software that's used only once a year, so all the bugs surface over a couple of weeks rather than continuously throughout the year. In general, the more variation you have, the more buffering and slack time you need in your schedule to absorb it.
In the Muda category, TPS defines eight categories of waste (all of which apply to software). I use the acronym DOWNTIME to remember them:
* Defects (of all sorts, but bugs and technical debt).
* Overproduction (is related to Inventory. Producing software we can't sell—often because we're not validating products before we build them)
*Waiting (all sorts of dependencies cause waiting, as do processes that mandate lots of meetings and the like).
* unused taleNt (punnishing people as "unproductive" instead of training them is a big one here.)
* Transportation (when the work needs to move from one place to another—from product to dev or dev to test, for example—you have delays).
* Inventory (work done that has not been released to customers. Money spent without a balancing revenue.)
* Motion (too much process. too many meetings. Lots of motion, but no progress).
* Extra Processing (building more software than you need. futureproofing).
Of course, all this is just the tip of the waste iceberg, but it's a good start.
Muri is long-term overburdening of the entire production system (e.g., too much work for the people at hand). That's not waste so much as creating a system incapapable of doing the work. To my mind, Muri also encompasses inappropriate organization architectures, where the management structure is fighting against working in effective—all too common in larger corporations, but you see it everywhere.
Mura is variation in the work to do. The biggest cause is usually random show-stopper bugs in the drop-everything-and-fix-this category. Variation can also happen at a macro level, for example, tax-related software that's used only once a year, so all the bugs surface over a couple of weeks rather than continuously throughout the year. In general, the more variation you have, the more buffering and slack time you need in your schedule to absorb it.
In the Muda category, TPS defines eight categories of waste (all of which apply to software). I use the acronym DOWNTIME to remember them:
* Defects (of all sorts, but bugs and technical debt).
* Overproduction (is related to Inventory. Producing software we can't sell—often because we're not validating products before we build them)
*Waiting (all sorts of dependencies cause waiting, as do processes that mandate lots of meetings and the like).
* unused taleNt (punnishing people as "unproductive" instead of training them is a big one here.)
* Transportation (when the work needs to move from one place to another—from product to dev or dev to test, for example—you have delays).
* Inventory (work done that has not been released to customers. Money spent without a balancing revenue.)
* Motion (too much process. too many meetings. Lots of motion, but no progress).
* Extra Processing (building more software than you need. futureproofing).
Of course, all this is just the tip of the waste iceberg, but it's a good start.
7
45
284
24,263
Matt Childs retweeted
2 minute rant about how senior engineers should care more about the construction of the business than the implementation du jour
35
57
477
50,851
1/6 @British_Airways, you left us stranded! Our flight was canceled due to "weather in Heathrow area" (rain), & the next available flight was 3 days later. Meanwhile, my other family members flew home without issues. @UK_CAA @SimonCalder @ConsumerReports
3
0
3
352
6/6 The staff deserve better employers, & customers deserve better experiences. @WhichUK @MartinSLewis @consumerchamp @MoneySavingExp
1
0
0
77
@WhichUK @MartinSLewis @consumerchamp @MoneySavingExp Bonus tweet, so many people complain to @British_Airways that there’s a whole industry scamming unhappy customers.
0
0
1
93
