Hacker News Reader

Comment by collyw | Hacker News Reader

Stop encouraging people to use Excel for tasks where it is a poor fit. I have wasted so many hours of my life because of Excel and people using it for non-spreadsheet purposes.

View on HN

You may be a dying breed.

As much as everyone wants to build full stack for every application, Excel is fantastic for complex math out of the box.

Excel is the average persons database.

I have found being able to expand on this has been incredibly useful and modular.

And as a note, VBA is definitely a real programming language, it up to the developer on how to treat Excel.

A surprising number of people have difficulty understanding Java or JS or Haskell or any "traditional" programming language but immediately grok complex Excel workbooks with VBA macros

Excel has some real limitations with large datasets and multiple users accessing it. It's far better to use something like Access or SQL for these purposes. It's also nicer in access when creating a form for people to use that pulls data from multiple input tables. For smaller data sets or sharing access quickly Excel is a champ and even when it fails it tends to try to do so gracefully. VBA does add a great amount of usefulness and is wonderful as an entry scripting language.

I must be part of the dying breed too. Having been involved in migrating very broken Excel/Access/VBA into properly formed web applications for companies I've worked at, the problem is people using the wrong tool in the first place. It may meet their immediate need but with no viewpoint on it's future usage, then fragility sets in and it costs the company more in upkeep and resource. I can see how that happens, but perpetuating it seems a bad idea.

Worked on a project last winter to add a fully featured compliance management system to a state regulatory agency in Access. They have moved to a sql server backend but the support for database operations is incredibly limited by the Access JET DB engine. Ended up just using ADO in my VBA scripts for everything instead of the Access database tools because it was not capable enough for the complexity. This client doesn't have the budget to migrate from Access, but has the budget to continue slapping garbage to it every year. At least as the primary NGO involved works with other states to create a web-based solution, my client might be able to move away from Access by 2025? I would say the majority of orgs that encounter this really have no choice in perpetuating the use of bad systems.

Looking at it from the perspective of a developer tasked with making a proper app from a spreadsheet, you're bound to come to that conclusion. But that's with the benefit of hindsight and with a massive dose of survivorship bias.

Of all the spreadsheets that start out exactly like the ones that need to be converted to an app, how many were thrown away having served their temporary purpose just fine? How many of them never outgrow the immediate needs and skills of its creator?

Perhaps spreadsheets should be seen as tools for business process prototyping. Starting a proper software engineering project only after a task has outgrown its spreadsheet based implementation may be a pretty efficient approach.

Maslow's Hammer in action.

I share your opinion. As a developer, I do not like Excel and the scripting environment around it, but I realize that I am not the target audience: it’s an incredible tool for analysts, managers and whatnot to explore and analyze data.

I work with a lot of data in my daily life, and see most BI tools struggling to recreate the same functionality that Excel offers, often failing in the process.

People try to recreate the Excel experience online too much, forgetting that Excel itself is a really solid solution and it’s worth taking serious.

VBA might be a real programming language, but imho, its syntax and tooling are awful.

Why would VBA be better than Python, Julia or R for handling complex math? Seems like the language would be really cumbersome and lacking in libraries relative to those languages.

I once worked for a company where the owner had written a Point of Sales system in Excel. It handled mixed orders where some product was given to the customer immediately, some they could come back later and it would be packed for them to grab, and some would be shipped after the event. There were hundreds of SKUs, up to 8 computers taking orders, and most scary of all, they were networked together so that another computer running excel could collect sales data and help the production team figure out how many of each product they should make next. Also the order fulfillment team used the same collected data.

It was brilliant. Excel was also the wrong tool. We constantly had problems with file locks, network shares, etc. After one year I worked with another guy to rewrite it as a web app.

My coworkers have wasted years of their life because people decided to write software that could feasibly be a medium-complexity spreadsheet.

I've wasted years of my life because of the web and people using it for non-scientific-paper-publishing purposes.

your point being?

Reply on news.ycombinator.com