Not to mention that sheets is far from a replacement for excel.

May I ask what you do with Excel that you cannot do with sheets?

Libreoffice and Sheets are both dead in the water non starters for all of the finance guys that I know.

The ability to open a million line csv file with 75 columns and execute a search and replace without leaking battery acid out of every orifice in your machine.

The hundreds of long complex well vetted plug ins that are drop and play with Excel which serve niche hard to reason about corners of regulation legalese. There has been so much cross pollination between the major firms that everybody has these tools and implements their own little tweaks.

The ability to quickly spin up a VBA macro to do some heavy lifting in a client provided 250 worksheet workbook each with 50000 line items (none escape quoted so every address with a comma in it malformed the data on client export).

The CSV support in Excel is a nightmare. And it still does moronic destructive things if it mistakes anything you're doing for a date.

That is my experience too, and I also find LibreOffice to be much faster than Excel for huge datasets. The biggest difference for me is clearly file compatibility and that is why I still use mostly Excel.

I'm glad that you have found success, however I would like it if you could explain further because this is exactly why LibreOffice is considered unusable (large data).

I have 32g of memory and a SSD on this machine. If I attempt to open a 1,000,000 .csv file, add a filter at the top, and attempt to filter by a column, my computer is done. LibreOffice will go gray and never recover no matter how long it is left. Eventually it will crash. Do the same in Excel and it will successfully search the column and return just the line items with said column.

That example is egregious as well. LibreOffice can not handle 500,000 line items files with 30 text heavy columns. It crashes on numerous basic spreadsheet operations, however Excel handles them (and quickly).

I am not a Microsoft shill, I'm running open source from Linux to Ruby/Rust and all the way down to Postgres. I would love nothing more than to stop having to keep a second windows machine setup next to me just to handle large csv files.

WHAT? CSV support in Excel is just completely broken. Million line CSVs are the only reason I have LO on my machine.

If LO would just make .xlsx a native format, they could probably take some market share from Excel. But, as long as there are little issues with interoperability, 100% of the world will be on Excel. Alas.

For me personally, sheets is fine although slow. For our finance people, vba and dynamics GP integration is a requirement.

We switched from o365 to gsuite in the last year, and have run into some things that make it feel like gsuite isn’t quite polished as a business product. Examples off the top of my head:

- Can’t share individual folders from a team drive. Not a dealbreaker, but hobbles g drive as a Dropbox replacement.

- Email to a suspended user bounces. In exchange, the mail goes through but access is cut off.

- inviting a google group to a meeting doesn’t work as smoothly as inviting an exchange group

- can’t delegate account access administratively, need to do it from the user account

There are big pluses too, and overall I’m not sad we switched. But it feels like MS has a much better handle on what corporate IT needs.

I'm sure the actual list is long, but I'd be happy if I could tweak graphs appearance, or do polynomial regression.

Plotting an X/Y scatter chart with lines linking the points.

Fit a polynomial trendline to a chart.

Choose the number of decimal places when showing a trendline's equation. Copy that equation, without re-typing it yourself.

Apply the solver to a nonlinear-but-continuously-differentiable system of equations.

You know, the stuff I could do in Office '95 on a 120MHz Pentium.

Absolutely. Sometimes you just want to output a quick CSV, open it locally in familiar spreadsheet environment, and explore.

For email? Yet to find a situation that requires it.

Desktop email clients like Outlook let you do things such as: select all email messages -> save as .txt -> feed the .txt to a python script that extracts useful data.

If we were using GSuite I guess we could use the Gmail API to get the message body data, but at the end of the day having local access is just more convenient (edit: faster).

Edit: for everything else that is not email (spreadsheet, word processor, presentations) there are also similar advantages related to having full fledged desktop applications.

Interesting use case. What exactly are you extracting from email that you’d need a python script for computing?

We do clustering on unstructured text and we output data in a Jupyter notebook. It's a proof of concept for now and the production pipeline will be a bit different. The whole Python ecosystem is awesome for this type of prototyping.

And that you couldn't easily do with Microsoft flow and office 365?

See my response to parent

The answer to this is fairly simple: Offline access to email. You need a client for this, and an arbitrary browser won't do.

Solved problem on iOS and Android with whatever mail app you want to use, but those are of course apps.

Solved problem on the PC with Outlook or "Gmail Offline", but "Gmail Offline" is a chrome-only feature so is, as far as I'm concerned, an offline email client built into Chrome, not a feature of Gmail per-se. See also: Any IMAP mail client.

An arbitrary client won’t do either. It would have to be a specific one.

So now that we’ve removed the need for an arbitrary client in both cases why can’t localStorage/etc solve this to most users satisfaction?