Wondering whether “best practices” has wrecked SW engineering
2 comments
You are definitely not alone, this is very applicaple to many fields as it comes down to lack of critical thinking by majority of people, that is why I actually think it is normal, not that I approve but it is just the natural way of how things go.
I see that the norm in IT is too much cargo culting but it is at different levels in different shops (different people, company culture etc.). It is especially funny when there is so much unclear and contradictory opinions/implementations formed around supposedly the 'best practices', but of course it is never funny :) to those people who know it all. So many people lack the notion of opinions and tradeoffs, feasibility of applying some 'best practice' depends on context but no, all it takes is someone to read a copy-pasted article and zealously defend the bulletpoints without even understanding the essence. People probably feel a lot of proud when they conquer (or at least when they think they got it) some complex unnecessary pattern that is supposedly the bestest practice that is practiced by only the select bestest developers, this probably adds a lot to cargo culting.
Anyway though, I found that being easy going sometimes (read: always) is the least painful way of dealing with such people. Of course it might be hard for some people, but at least try to play the game of being politically correct, for the sake of your own sanity. For example you can just reply that it is not necessary to apply complicated patterns on simple showcase project, this would be a pretty political answer I believe.
I see that the norm in IT is too much cargo culting but it is at different levels in different shops (different people, company culture etc.). It is especially funny when there is so much unclear and contradictory opinions/implementations formed around supposedly the 'best practices', but of course it is never funny :) to those people who know it all. So many people lack the notion of opinions and tradeoffs, feasibility of applying some 'best practice' depends on context but no, all it takes is someone to read a copy-pasted article and zealously defend the bulletpoints without even understanding the essence. People probably feel a lot of proud when they conquer (or at least when they think they got it) some complex unnecessary pattern that is supposedly the bestest practice that is practiced by only the select bestest developers, this probably adds a lot to cargo culting.
Anyway though, I found that being easy going sometimes (read: always) is the least painful way of dealing with such people. Of course it might be hard for some people, but at least try to play the game of being politically correct, for the sake of your own sanity. For example you can just reply that it is not necessary to apply complicated patterns on simple showcase project, this would be a pretty political answer I believe.
You are correct that many of the Software best practices have no scientific evidence. Also if you step outside of the software world into best practices of Business management, sales and accounting. They also not backed up by any scientific evidence.
Most of the best practices get developed by a smaller group of companies normally larger ones and everyone copies them since it worked for google it will work for us.
Having worked for different size companies and in different industries. I can tell you that most of the time they don't work well.
Some huge successful companies got big and do well not doing any of the normal standards. It was mostly luck and creativity.
Most of the best practices get developed by a smaller group of companies normally larger ones and everyone copies them since it worked for google it will work for us.
Having worked for different size companies and in different industries. I can tell you that most of the time they don't work well.
Some huge successful companies got big and do well not doing any of the normal standards. It was mostly luck and creativity.
At this point I've seen this kind of feedback one too many so, given that I'm just a click away from giving the whole current SW dev party the middle finger and do something more productive with my life, I'd just want to see if I'm alone in feeling that:
1. cargo-culting (uncle Bob et all) has poisoned the field to the point of destroying even unrelated paradigms (at this point I won't be impressed if someone tries to apply SOLID to Elixir)
2. SW eng roles are gatekeeped (this was a take home test) by evangelists of a certain type that make sure that things will keep being done in that way.
3. That way being the way Java ecosystem made sure that nobody with a creative vein and a slight aversion to authorities and bureaucracy would ever set foot in it again. Consistently delivering overengineered BS that build nice CVs and careers and leave companies with unmaintainable overengineered messes.
4. There is yet a shred of actual damned scientific evidence that all this 90s and 00s "best practices" actually lead to better SW. There has been no double blind experiment to measure whether the SOLID principles actually work or their best use is to insure a fat paycheck for evangelists and build CVs. For the record I conside most of SOLID and a lot of DPs just self-indulging rules conjured out of thin air. A case of an academic trying (and succeeding) to impose his hugely impractical views on the real world. Shit that looks good on paper but is actually offering little value and a lot of times is akin to splitting hair.
Anyway. Bye and thanks for the fish.
I think I deserve much more than munching on canned BS by this world's uncles and their minions.