LibreOffice 6.3 Released With Better Performance, UI Enhancements

[uid313]

I like the fact it's not MS Orafice.

Seems pretty weird reason to like LibreOffice. Sounds like you don't even like LibreOffice and only like it because of Microsoft Office.

Personally, I think Microsoft Office is great. It is very easy to use and you can easily do pretty things.
The thing I like about LibreOffice is that it is free, open source, and cross-platform.

[uid313]

I like how it's possible to connect to it with Python, and automate lots of stuff.

What I miss — and a lot of peoples in scientific community do — is support for LaTeX imports. Right now if you want to convert a LaTeX document to an office format, you have to use pandoc. But pandoc is far from being feature-complete, also I doubt its support for various office formats is anywhere near the LO's one, simply because of man-hours people put into these two apps.

Oh, I didn't know it was possible to script with Python, that is really cool.

Personally I find LaTeX pretty difficult, but it is a cool piece of technology. Example, you have to declare encoding type, it is not UTF-8 by default, to use simple common things like hyperlinks you have to import packages, for somethings you have to compile the document twice for it to work, I don't think you can directly compile from .tex to .pdf without first compiling to PostScript, etc. Could have a good WYSIWYG too.

[Hi-Angel]

Oh, I didn't know it was possible to script with Python, that is really cool.

Yeah… UNO API (the API to write macros and scripts in LO) is not very good though, but maybe in part that's a guilt of python even allowing such a code. For example, lots of types are hidden, so you can't access them, yet you do work with them all the time. It leads to variouis problems, such as not being able to catch an exception by name, or not being able to test whether the object in hand has particular type (because you can't refer to the type to begin with). I also suspect that not being able to check whether an object is an instance of Iterable class has the same root reason.

That said, I still think python is the most pleasant way to use LibreOffice UNO among all supported at the moment langs, because you can fire up an interactive shell, and find out various properties of an object at hand interactively.

The biggest problem with using UNO is not being able to find a way to access a given object from the top of hierarchy. This takes 90% of your coding time, this is very-very bad.

Personally I find LaTeX pretty difficult, but it is a cool piece of technology. Example, you have to declare encoding type, it is not UTF-8 by default, to use simple common things like hyperlinks you have to import packages, for somethings you have to compile the document twice for it to work, I don't think you can directly compile from .tex to .pdf without first compiling to PostScript, etc.

Yeah, I agree. Note that you can use e.g. XeTeX to simplify things; and also, no, you don't necessarily need to convert to PS. You can use pandoc to convert directly to pdf (it's possible that it does use internally PS, but it's hidden from you anyway).

Could have a good WYSIWYG too.

Yeah, well, you can just you office if you want WYSIWYG and save it as an odt. The whole idea of LaTeX is that it's plain-text format to describe same things as an odt.

FTR: it could be useful to have a split-screen, where you could see your text rendered as you type (and I'm sure there're online editors that allow you to do it. For one, the math.stackexchange uses subset of tex, and renders it as you type), but it's not a WYSIWYG editor.

Honestly, I don't really like typing in LaTeX. I'd prefer to use Markdown, and then use a script to embed this into LaTeX (probably something with pandoc conversions). I like LaTeX because I can see exactly what formatting parameters I have set. While I was a student, our teachers were overly strict about paper formatting (ironically so, much more than about their actual content), and exact parameters for every piece of text were documented. And it takes a lot of effort to make sure you haven't mistaken anywhere while using WYSIWYG, but almost no effort if you use LaTeX or CSS. And I was very sad that they didn't accept PDF but only office formats (and for some reason not even an odt), so I had to go into great lengths trying to format the text with LibreOffice API.

[Mateus Felipe]

I would like to see more love given to Base. It's a nice software, but it's almost the same as when it was Open Office (I event saw some days ago a label that was saying "OpenOffice" instead of "LibreOffice".

[kneekoo]

Ok, what version is the standard version? The version numbers are only given for the basic version.
http://libreoffice.soluzioniopen.com...86_64.AppImage

I have no idea. Honestly, it seems stupid not to mention the version numbers in the file names, but at least I'd download the Fresh release because I'd be sure I have the latest version.

[caligula]

Example, you have to declare encoding type, it is not UTF-8 by default

I don't know how it works internally, but the more "modern" luatex and xetex engines have supported UTF-8 for a long time, with UTF-8 being the default encoding. I wrote a paper using LaTeX years ago - even back then the LaTeX distributions did support Unicode to some extent. It seems the whole ecosystem is switching to UTF-8: https://www.texdev.net/2018/03/25/la...8-as-standard/

to use simple common things like hyperlinks you have to import packages

This is hardly an issue. Keep in mind some publishers don't want documents with hyperlinks.

for somethings you have to compile the document twice for it to work, I don't think you can directly compile from .tex to .pdf without first compiling to PostScript, etc.

The modern engines compile directly to PDF, but they still require 2+ runs due to the way hyperlinks, labels, and counters work.

[timofonic]

I would like to see more love given to Base. It's a nice software, but it's almost the same as when it was Open Office (I event saw some days ago a label that was saying "OpenOffice" instead of "LibreOffice".

I agree. Base socks!

[uid313]

I don't know how it works internally, but the more "modern" luatex and xetex engines have supported UTF-8 for a long time, with UTF-8 being the default encoding. I wrote a paper using LaTeX years ago - even back then the LaTeX distributions did support Unicode to some extent. It seems the whole ecosystem is switching to UTF-8: https://www.texdev.net/2018/03/25/la...8-as-standard/


This is hardly an issue. Keep in mind some publishers don't want documents with hyperlinks.


The modern engines compile directly to PDF, but they still require 2+ runs due to the way hyperlinks, labels, and counters work.

Oh, it is a nice improvement that it these days use UTF-8 by default.
I think hyperlinks are part of the document and if a publisher does not want it, you should be able to compile the document without editing out the hyperlinks, I think there should be a --no-hyperlinks or --without-hyperlinks option to the compiler.

It's also nice that it can compile directly to PDF without an intermediate DVI format these days.
Yeah, that thing requiring 2+ runes really sucks, and is confusing, and wish it wasn't like that.

[danmcgrew]

It's been TEN YEARS, FOLKS, and The Document Foundation still can't make a tool that's compatible with, whether you hate it/them or not, Microsoft's OFFICE--the industry standard for sending documents back and forth.

And no, I'm not NOT a Microsoft shill; I am Linux through and through, but one canNOT deny the absolute facts. And the facts are these: there are simply very important people and organizations to whom you had BETTER send MS-Office-generated documents to, or just don't bother sending them at all.
Back to my REAL point--

What in the bloody hell is wrong with The Document Foundation?

If you're foolish enough to contribute to these non-performers, you need to re-think your priorities.

[foolishgrunt]

It's been TEN YEARS, FOLKS, and The Document Foundation still can't make a tool that's compatible with, whether you hate it/them or not, Microsoft's OFFICE--the industry standard for sending documents back and forth.

And no, I'm not NOT a Microsoft shill; I am Linux through and through, but one canNOT deny the absolute facts. And the facts are these: there are simply very important people and organizations to whom you had BETTER send MS-Office-generated documents to, or just don't bother sending them at all.
Back to my REAL point--

What in the bloody hell is wrong with The Document Foundation?

If you're foolish enough to contribute to these non-performers, you need to re-think your priorities.

You seem to be confusing interoperability with interchangeability. You can export a document from Writer and import it into Word, and yes, the layout will change a little - but if that's your standard for success you're setting yourself up for disappointment. If layout is crucial, send a pdf - if the focus is content, the interoperability is already there.