bjoerns

I've built a mobile airpollution sensor together with a few other parents and their kids at my 8-year old son's primary school in London. Air pollution is a big topic as the school is right next to a busy road, a lot of children are suffering from asthma and Islington has been pretty useless in collecting/publishing data. So we've started taking things in our own hands and built a handful of raspberry pi based monitors in class rooms, the playground and children take it with them on their school run. geo-tagged data is automatically uploaded to little influxdb/grafana based web service wheneve the pi has a wifi connection. Makes it an interesting project for kids to look and interpret charts and stuff as well. Currently we're measuring PM2.5/PM10, temp and humidity only, haven't had the time yet to look into NOx sensors yet...

https://github.com/bstiel/airpollutionpi

I quite like daff (have used it for diffing Excel sheets):

https://github.com/paulfitz/daff

True. But isn't it a bit too easy to just tell people to stop using Excel (which they won't do anyway) instead of giving them something to help them solve their problems _within_ the Excel/spreadsheet world?

feel free to drop me an email: bjoern.stiel@zoomeranalytics.com

great story. I used to train on a C152 and knew a few guys at the local aerodrome who had done ferry flights from the US to the UK. Which is something I (as a shitty weekend VFR pilot) have the uttermost respect for. But that story about getting it from Oskosh to Gaborone, wow.

We have a tool that makes Git understand Excel workbook files: https://www.xltrail.com Is that something that might help?

Ok, here's he full story: We have a product (https://www.xltrail.com) that makes git diff and all that work with Excel workbooks. So you get a lot more out of it than dropbox/drive/sharepoint etc as you get to inspect content changes.

90% of our clients/prospects are in financial services (heavy workbooks with VBA, think applications, not spreadsheets) but today I spoke to a prospect from a different industry which made me think that there might be people outside banks, hedge funds going down the same path.

In general, yes. However, from my own experience, there is a (small and industry-specific) subset of "Excel developers", especially in investment banks, hedge funds etc that know both.

Commit your Excel workbook to a Git repository. I know it's a binary file so you can't diff or merge out of the box, but at least you get the commit history. Does that make more sense?

I would say that Joel Spolsky's video is one of the best ones out there (I used to be a die-hard Excel dev on a trading floor).

Other than that, we have a collection of open-source Git Excel workbook repositories that might be useful to go through: https://cloud.xltrail.com/

Some of them are pretty advanced, others less so.

It's an interesting idea (though I'm not sure how much appetite there is out there).

We've built a version-control app on top of Git for workbook files (www.xltrail.com) so I know a thing or two about parsing the cell content of sheets.

There is no Python library that supports reading cell formula content across all workbook files (for instance, xlrd does not support formulae at all).

If you can go beyond Python, have a look at sheetjs.com (open-source JavaScript based) or Aspose (commercial .NET solution).

That aside, "client" integration of such a tool could be a major blocker from our own experience. E.g. if you could build it in such a way that it's "fire and forget" (like uploading it to SharePoint and integrate it there) it is a lot more likely to being used. Or integrate it via Git (drop me a message if you want to chat). Or add it as a feature to something like rubberduckvba.

Revenue is vanity. Profit is sanity. Cash is reality.

Our server-side (commercial) product actually does handle spreadsheets as well as VBA code. If you want to see it in action, have a look at our demo repositories: https://demo.xltrail.com

op here: feel free to log an issue on GH, it's definitely on our agenda: https://github.com/ZoomerAnalytics/git-xltrail

It's the hosted version (free plan). But we also have customers who use the (free) self hosted version and we and them love it. Do you use GitLab yourself?

We (disclaimer: I'm the co-founder) have developed a Git-based add-on for Excel workbooks. Basically, it starts where Git left off in the Excel world. Here's a short blog post how it works out for one of our open source projects: https://www.xltrail.com/blog/xlwings-is-now-on-xltrail

The diff is actually not really straightforward - a sheet is two-dimensional and you have to work out column/row inserts/deletes (which themselves are intertwined) before you can even start looking at cell changes. So it's quite non-standard stuff.

We use S3 for our hosted version - but for our on-premise offering it simply has to be 100% on-premise, entirely on our client's infrastructure.

That's a very good point - but for what we do (on-premise version control for Excel where each workbook version represents one object) we won't be getting even close to that number. But yes, agreed, it entirely depends on your use case.

After evaluating a couple of options mentioned in the other comments here, we recently replaced our in-house built s3 clone with minio for our on-prem version of our app. Very robust and stable.

Fellow (ex-)banker turned founder here. I can relate to this... and it is very scary (the amount of spreadsheet craziness I've seen on trading floors is insane). But I still believe that Excel is a powerful and power-enabling tool. And I know from my own experience how fragmented the "Excel tools industry" is.

We've been working on a fully fledged version control system for spreadsheets for ~2 years now and finally getting somewhere with Pathio[0] (we used to be called SpreadGit but that wasn't ideal for many reasons).

It's taken us a while to get here as our aim was always to build a system that: - works with all Excel file formats (including the binary formats) - works with VBA - integrates seamlessly into (non-technical) people's workflow (which is different to how developers work) - is fast (I personally worked on 100mb spreadsheets when I was at UBS...) - gets the diffing right (which is different in two dimensional space where inserting/deleting rows/columns has a knock of effect on potentially a large number of formulas which are just... noise)

Feel free to drop me a note to see if we can be of any help to you - We also have an on-premise version (thanks to the guys at Replicated[1]).

[0]: http://www.pathio.com [1]: https://www.replicated.com

This is great. We went through a lot of pain keeping our cloud and on-premise offering [0] in check. Eventually took our infrastructure back to the drawing board and settled on a docker based solution with some custom orchestration on top. Cloud is docker swarm and on-prem docker on a VM. A service like Gravitational would have saved us a lot of time.

[0]: http://www.pathio.com

err, is it just me but shouldn't this be the other way round? as in use HTML for desktop apps?

Excel-DNA is a nice open source project that lets you write XLLs in .NET http://exceldna.codeplex.com

Nope, same here! I personally prefer the routes being near the views as they belong together and most times I can't be bothered looking the route up in a different file like in Django or Rails even. But I guess it all comes down to personal preferences.

That's a very good point indeed!

I guess it's in the eye of the beholder. Like so many things in life.

Being a practical guy, I wonder if banks' accounting systems can cope with negative interest rates.

"Insanity is doing the same thing over and over again, but expecting different results."

"The fact that they use spreadsheets suggests that they spend very little time worrying about accuracy." I don't get it. Why does the fact that a person uses a particular piece of software suggest that they don't care about accuracy? Does this mean that just because I use R or Python I worry more about accuracy? Really? I think it has more to do with loss of control - real developers hate Excel because it gives business people something they can work with straight away. Every time there's a bug which could remotely be related to spreadsheets, people start another wave of Excel-hate. As if this was entirely impossible in a real programming language. Don't get me wrong, spreadsheets are not perfect by any means. But neither is anything else in this world. It purely comes down to choosing the most suited tool for a given problem. And in Piketty's case there's nothing wrong with choosing Excel.