I was thinking about this earlier, and while I've been in the industry ~8 years, I've only worked at two companies, and they've both had some pretty fundamental dysfunctions from a strategy/leadership/management perspective (e.g. projects building software that no one likes/uses, major flaws in delivered products, huge delays/delivery slips), that I want to see what a "successful" company looks like.
But then I thought some more, and I don't know if any companies actually do this correctly.
Do companies exist that can deliver on projects at a fairly steady clip with teams that build solid products generally on time and within budget? Or are we all just floundering around, making buggy apps that take forever to release and don't solve any problems, except for the lucky few who can get everything to harmonize out of sheer statistical happenstance?
The few companies I worked in and with, and all the code I've seen, points to the latter case - most of us are delivering utter disasters, thanks to the mix of technical errors, management errors, shifting priorities, internal politics and time constraints.
Agreed and same. It's my opinoin that this issue is underdiscussed and that the mess that is often delivered is rationalized by "methodology" and "it's not our fault our requirements force us to deliver this".
But whatever the reason or rational, every time a story breaks about another security exploit or privacy exploit I read that as a condemnation on our profession.
In other professions, part of the certification process is gaining a basic level of understanding of the ethics one is judged against when associating one's work with that of the larger trades group, guild, or association. Often it is well understood that, to some extent, the topic is simply being paid lip service. But,it is also understood that those ethics draw a bright line which those the association serves will not tolerate when openly crossed. Take for instance investment professionals. Everyone knows insider trading happens and it's not uncommon to put profit before fiduciary duty, but when those lines are openly or egregiously crossed it is not tolerated under threat that the understanding between the client and advisor that minor infractions will be tolerated will no longer be honored.
The software engineering profession lacks this basic ethical covenant with its customer. Just look at the utter lack of product warranties. Sure there are SLAs, but there are virtually no warranties. And it shows.
As software begins to function more and more as the linchpin of our society, this issue will morph from technical debt to an Achilles heel. We complain about anachronistic laws. What about anachronistic code? We complain about absurd laws. What about absurd code? It's just as dangerous.
I'm not so sure about that. Almost every time I update IDEA I get an annoying bug. For example, on last major update something broke and dialogs/popups wouldn't display. I couldn't close the editor because of this. When I tried to close the windows a confirmation popup should show up, but due to the bug nothing was happening. To me, being able to close an application without having to kill the process seems like a critical feature, yet it somehow was not noticed during testing.
What I'm trying to say is that if you're serious about finding issues no project/product is safe from you :)
But then I thought some more, and I don't know if any companies actually do this correctly.
Do companies exist that can deliver on projects at a fairly steady clip with teams that build solid products generally on time and within budget? Or are we all just floundering around, making buggy apps that take forever to release and don't solve any problems, except for the lucky few who can get everything to harmonize out of sheer statistical happenstance?