441–450 of 455 posts
Earlier quoted context omitted.
Don’t feel too bad because if you would have created a product from it, today this product would be deprecated..
I don't think that's necessarily true. They built VSTO for C#, but everybody was using ExcelDNA instead, an Open Source solution that was much better!
Earlier quoted context omitted.
How large is pretty big? And who said it must be installed by default? Hyper-V is enabled by default on supported hardware.
> How large is pretty big? Python data science containers are on the order of GBs. > And who said it must be installed by default? I did end my comment with thoughts that I think it would be nice to have as an optional install. Mentioning why it isn't likely a default install is relevant because they want this Python feature to be accessible to everyone and they want people to collaborate and the UX for an optional i…
Yes. And SSDs are on the order of 100s and 1000s of GBs.
> Mentioning why it isn't likely a default install is relevant because they want this Python feature to be accessible to everyone and they want people to collaborate and the UX for an optional install isn't what they were looking for and would complicate things.
How do you know these things?
It won't be accessible to everyone until it's installable. Collaboration is not cloud exclusive. The UX isn't what they were looking for is circular.
Earlier quoted context omitted.
Sandboxing python enough that millions of users could safely pass around excel workbooks is a super hard problem, especially when you allow third party libraries. Forcing it to run in an isolated environment they control simplifies the problem greatly.
Windows has it's own implementation of containers, and with a docker-like structure, and xlsx being a zip file, excel could simply have a dockerfile like structure "FROM excel-python:2023" "COPY xlsx:scripts/" etc. new versions of excel could keep the last few 'base' containers, or if they are in love with the cloud, the signed, verified base images could be pulled on request.
Earlier quoted context omitted.
> It seemed like Microsoft's product complexity and dedication to backwards compatibility of software would make them the ones to move software management forward. That's precisely why they _didn't_ go with the package manager model of software distribution. Package managers in most distros assume that everything you download via apt/dnf is a part of the operating system and integrates with whatever dependencies the…
The state space doesn't have the conveniences that linux and language tools leverage, that's why I wanted to see them try to advance the field beyond what we have now.
Earlier quoted context omitted.
I cannot remember the last time I had some long number and thought to myself, "You know what will be helpful, convert this number into scientific notation!"... It's asinine Excel has this behavior with any large number. Try to work with a list of EAN/UPC codes... it'll wreck every single one every single time, unless you take significant care to guard against it via formatting, special characters that trick it into u…
> I cannot remember the last time I had some long number and thought to myself, "You know what will be helpful, convert this number into scientific notation!"... Okay, so set formats appropriate to what is useful for your data. Defaults don't cover all cases, otherwise there would be no non-default options. > It's asinine Excel has this behavior with any large number. Try to work with a list of EAN/UPC codes... Codes…
Your offered work-arounds only apply in a very narrow workflow. I work with Excel nearly every day, all kinds of files (CSV, TXT, XLS/X, etc). There's so many ways this issue can bite you unexpectedly - causing you to save and permanently destroy data.
If I type/paste/import/download/whatever `01234567890987654321` into a cell, 100% of the time I do not want `1.23+E18`. I don't want to even see that, I want to see the original value I entered.
It's asinine.
Excel may be the most used Office Suite app, but it's by far the most painful to use. The Excel Team has a long history of justifying downright perplexing behavior. Remember when Excel was the only Office app that didn't support Snap-To back in the Win Vista/7 days? Joel had a long-winded explanation that boiled down to "just cause"... it was perplexing even at the time. It really makes one wonder how bad the rat nest of a codebase Excel is.
"There's ways around our astonishing default behavior so just deal with it". That's not a great answer.
Half joking: Now we know why Microsoft hired Guido.
Earlier quoted context omitted.
It includes things that aren't selected, which is other random data. If I select, copy, and paste, the only desirable outcome is that I get a copy of my selection.
I don't understand what you mean. In my example I select the cells that contain 1, 2, 3, 7, 8, 9, a, b, c, then copy, then go somewhere else and paste, and Excel pastes 1, 2, 3, 7, 8, 9, a, b, c. I don't see an other data.
Earlier quoted context omitted.
Are you saying that excel does preserve digits? I must be doing something very wrong (entirely possible), cause once the document is saved, the digits are gone.
That reads as a normative "should" to me. It doesn't preserve digits, but it ought to.
Earlier quoted context omitted.
You can just ship a copy of Python with Excel, no need to ask anybody to maintain their own installation.
Anaconda distributions aren't exactly small. I'd also assume there's some sort of environment isolation, so you're getting multiple copies. Maybe one per workbook?
Earlier quoted context omitted.
It's the dependency hell problem as packages are added to the sandbox environment. If they were just shipping python (standard library), it would be fine, but all of the useful libraries have large numbers of dependencies.
doesn't this dependency hell still exist, but just in the cloud?