Hi I’m a Python Developer, tired of doing internal applications using Excel as a UI. I’m thinking of proposing to turn most of our projects into internal web apps instead. Has anyone gone through this sort of problem?

My team is quite pro at using Excel, so naturally they prefer to use the tools I build from Excel. Some of those tools are also used by external teams, but they are not as capable with Excel, so they need supervision and guidance.

There are multiple concerns that arise:
- I code on Mac, but they need to run it on Windows, so compatibility issues
- Some of their laptops might not have enough resources to run the tasks
- Errors are harder to trace and could be very user-specific.
- New developers might not be familiar with Excel and the way to integrate with Python

I would like to know your opinion or criticism

  • 1
    the big problem with many different webapps in a company is that there is no directory of what apps exist.
    They usually just exist but nobody got an overview.

    So make an website with all served webapps in your corporate network with description what each of one does. And make ot mandatory to update this app directory. Talk to your higher ups and cyber security.
  • 3
    Speaking from experience, do not develope anything dependent on Excel. Sure it is a great app, but when it comes to automation dont rely on Office applications.
  • 3
    Won't work

    They will end up by asking to recode Excel in web.

    Usual :

    "No I want you to present me ALL DATA on the same webpage and a filter, just like Excel, for grou[ping and filtering"

    "Oh I can't do cross table here, can you export this view into Excel?"
  • 2
    @NoToJavaScript If i run across a scenario where they want an export of the data I will ask them why. Find out the real reason they need it. Are they looking at something specific? Do they have other data they merge with it? Maybe the just like the columns in a different order.

    I've found many times an excel dump is not what users need (but it is what they think they want).
  • 1
    Yep, exactly what I told my self when ( worked exactly on similar project.
    Response was : “Yeah here we can filter, then sort”
    Me : I specifically added filtering and sorting. What’s wrong with it ?
    Them : It’s not like Excel we don’t want to retrain the staf.

    Trust me, I feel you, I’m even rooting for your project, because they actually need iot and they don’t realise how much.

    But roadblocks will be hard !
  • 0
    Nw.Js for a 'front end', and use semantic UI if you want a semi-clean neutral design and dont give a fuck about css.
  • 1
    @NoToJavaScript It'd probably be cheaper to license a professional web component package that includes a datagrid. Especially in a professional setting. Gonna save you time and thus money in the long run. Couple grand tops.
  • 1
    @Wisecrack Yes, kendoUI or similar.
    But they all have the same limitation (If you can find me one without this limitation, it will replace like 3000 code lines of custom) :
    They cannot have Virtual Scroll AND grouping at the same time. It is one OR other.

    I need to fucking display 1000 groups, each having 100+ sub-items. (Yes we do it on group level at this time but not even virtual scroll, just pagination)
  • 1
    @NoToJavaScript cooking steak atm. once dinbers done I'll go take a gander at the options. I seem to recall reading this exact same question before just cant remember what the answer is atm.
  • 0
    Workflows exist for a reason.

    People know what they do and when they do it long enough they feel confident even though they maybe don't have a fucking clue what they're doing.

    I think it is not wise to break existing workflows for the reasons you mentioned.

    It becomes a whole other thing when you _integrate_ the workflow in the (web-) application you build.

    Excel export and import isn't that hard.

    Hard is to achieve a state where one source of data is dominant - eg. everyone is aware that the database of the applications is the source of truth, not their locally modified copy.

    So the goal with the applications is __not__ to take away Excel and the workflows at the first stage, but to make sure that there is a single source of information that should be the database of the web application.

    As soon as you've reached that milestone, you can start working on improving and migrating the workflows.
Add Comment