Protein interface – how to change aproach to building software?

Re:Protein interface – how to change aproach to building sof... by catinblack on Sunday October 16, 2022 @12:10PM (#33224700)

I read this article recently. It fell into my hands by accident and I found this quote there:

-- Assuming that we should launch many pilot projects, let’s consider for which of them it is possible to build minimally viable products – and then try (where possible) to use low-code and no-code tools. With these, we can get less technical people prototyping quickly (of course, more complex integration pieces may still require the help of an engineer). Thanks to the fast implementation, managers or people responsible for the functioning of the process in the company are able (within a few weeks) to check and analyze the sense of an idea. They can test it in a natural environment, and if everything works, they can create a dedicated solution developed by an experienced team. --

One of my close friends, whom I advise in the field of digitization, owns a factory. I can't convince him for anything to try this approach, which I think is absolutely right. The client believes that only their IT department and CTO can change anything in automation. Or only they can be responsible for data analysis. How to change this, how to convince him?

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by csours on Sunday October 16, 2022 @12:50PM (#33225075)

IT and the CTO have tried very hard to convince them of this because IT and the CTO have seen the hell-sewer that shadow IT creates.

The hell-sewer is real, but banning shadow IT is not the right answer.

---

Above is the answer to your question, but it raises the question - what is the right answer to shadow IT?

The problem with this question is that it sounds like ONE question, when in reality it is a many segmented question. You need a way to characterize the shadow IT in question so that you know what approach you should take.

If shadow IT overlaps with governance, then it should not be shadow IT.

If shadow IT does not, then it's probably fine to stay that way or be replaced at some point by a MVP.

---

That answers the question I raised, but now I've raised another question: WTF does "overlaps with governance" mean???

And this is where the nightmares begin!

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by catinblack on Sunday October 16, 2022 @01:05PM (#33225230)

It's not even that they are supposed to use some solutions without the approval of the IT department. The IT department could, for example, prepare some sandbox for people to play with. They can create separate systems and with their consent and approval people could "play" with ideas. I feel under the skin that there is huge room for innovation.

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by csours on Sunday October 16, 2022 @01:20PM (#33225346)

I don't dispute that at all. This could happen. It should happen.

IT has an allergic reaction to Shadow IT because they know about the times they had to deal with it. Understanding what IT has had to do in the past to fix things helps you understand why they don't think in these terms.

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by ok_dad on Sunday October 16, 2022 @01:27PM (#33225404)

Shadow IT happens when the user isn’t getting what they need. I wrote software in excel at one job that did the priority planning for incoming inspection, because we didn’t have anything we could use to do that other than SAP and I wasn’t going to engage with that system due to effort required.

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by clcaev on Sunday October 16, 2022 @04:32PM (#33226952)

Shadow IT could be seen as ignoring externalities such as security, operational resilience, and process coherence in order to "get things done". Eventually the bill comes due (security incident, key employee exit, incorrect results). Questions are: can you recognize failure? who pays to fix it? what happens while it's being fixed?

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by openfuture on Sunday October 16, 2022 @12:50PM (#33225080)

I think what you are describing is basically a good development environment.. where you can contribute your insight at depth of skill level appropriate for what you have to offer. From my point of view the path to the eventually persuasive version of this enviornment (i.e. where your friend and the CTO will accept to use it at their factory) starts by setting a common standard for data interchange between the tools. So partial work can be commented on and different options for completing it can be assessed relative to one another. I'm just one person but I am on my way and datalisp.is my starting point ... I am certain that the road goes by your place on the way to prosperity.

[ Parent | on HN ]

Re:Protein interface – how to change aproach to building sof... by catinblack on Sunday October 16, 2022 @12:59PM (#33225169)

Thanks for this vote. Good point with the standardization. I take it.

[ Parent | on HN ]