Hacker News Reader

Comment by taeric | Hacker News Reader

My point would be to do whatever to improve interop. At the least, focus on ux. Intrinsic quality of the code does little to address this, as much as we developers want otherwise.

And as a student, I definitely used free options. I encourage that use. Just don't think you will make much inroads in the commercial world. Until you can use all of the commercial world.

View on HN

They are doing work on the UX, but that's work for a different kind of people than the folks who do security.

Security is an enormous concern with an office suite. The formats which need to be supported by an office suite are terrifyingly complex, and people open documents from untrusted sources on a regular basis. Opening malicious documents in office suites is a major vector for malware attacks, especially in the "commercial world".

and as soperj points out, LibreOffice often has better interop with word, especially older documents. It also correctly opens and saves ODF formats, generates better quality PDFs.

The UX is also kinda hard to objectively iterate on. The current iteration of the office ribbon generates some amount of positive feedback for new users, but experienced MSO and OO.o users often prefer the current style of LibreOffice. If iterating on the UX means anything, it probably means adding a second UI. Outside of UI, it would probably be good to get better cloud (or even collaborative) support. It's frankly not certain how that would be done, maybe federated synchronization servers.

This is an extremely fair criticism of my post. And looking back, I regret that I was a downer on the first post.

You are right that they can and are doing both.

It inter-operates better than Microsoft's own software. Try opening a document that was saved from an older version of word in the newest. LibreOffice is better.

I could have agreed with your claim 10 years ago, but not now. I've had such problems with MS software back in office97/2000 era, but not now - on the other hand, I have had interop issues with Libreoffice back when I used it last year, so it's currently not good enough (yet?) to be used safely.

To be clear on definitions, my understanding of "interoperability" involves the following workflow - if I open a complex document made by someone else in MS Excel or Word, change one word in it, save it (in MS format) and send it back, then I have to be sure that I haven't corrupted anything else in that document - layout nuances, styles,references, pagination, whatever. Currently it's not the case, and it makes it unusable, since I can't even verify if something looks different on my computer than it had looked in MS Office.

I mean, it's a simple test - if opening ("importing"? "converting"? technically that might be more appropriate than merely opening) a document and immediately saving it changes it somehow, it's not ok.

Use a open standard. Save it as an odt.

That's switching to another workflow, not interoperability with the existing one.

I had a bunch of old Word-documents from a late researcher using non-latin scripts and custom fonts that needed fixing for archiving (the only important step was to make these readable by humans). Don't remember what encoding. Current version of Word wouldn't even touch these, raising errors left and right.

In the end I changed the strings that specified the problematic/missing fonts in a hex editor (I naively assumed that would do the least damage...) to a compatible font I eventually found after some digital archaeology. LibreOffice accepted these files but didn't display them correctly (as in: according to what my eyes and brain expected - "machine readable" was in a different universe at this point). I saved the hex-edited documents in whatever version of docx LibreOffice exports to. Word accepted these, displayed them as well as could be hoped, and a little bit of modern history could be saved for future generations. Not a very technical approach, I guess.

I feel the 1990's in general did a lot of damage. It's not only Word. Competing, proprietary formats together with lacking text-encodings (and in my case custom fonts messing up character tables, due to academic/linguistic needs) has caused more information loss than the decades before and after for the data I've worked with.

Suffice to say I use plain text as much as possible (markdown, latex), as text-encoding alone is hard enough.

That's sounds like a proper programmer-archaeologist gig to make Vernor Vinge proud.

Reminds me of wine, which apparently has better luck with old programs than do new versions of Windows.

> My point would be to do whatever to improve interop. At the least, focus on ux. Intrinsic quality of the code does little to address this, as much as we developers want otherwise.

Intrinsic quality of the code allows the developers to change the code faster and with less risk of breakage. This helps them address everything else, including interop and ux.

This is the argument, but evidence is actually severely lacking that it is the case.

And let me be clear, I agree with the argument.

Fixing the sort of bugs found by fuzzers doesn't always lead to cleaner, easier to maintain code. It depends on what the bug is and how it's fixed. Sometimes you just have to write something ugly and inscrutable to deal with how your compiler handles undefined behavior, memory allocation, etc. (Or, in other words: in unsafe languages, there's a class of bugs that occur precisely because doing it the straightforward, idiomatic way leads to the bug.)

This doesn't make this bad or mean that you should leave bugs around, but fixing bugs found by a fuzzer is unlikely to lead to improvements in Office compatability, UX, etc.

Maybe I'm alone in this, but ever since Microsoft went to the ribbon-style UI I've not much liked the MSO user interface, it hides too much stuff. I find LibreOffice's menus far simpler to follow and find things in. Possibly just from having grown up on WordPerfect and Lotus 123.

Just a taste thing though I suppose.

Yeah, it's just a taste thing. I personally don't like the ribbon either, but back when I worked tech support, we transitioned our workstations from Office to LibreOffice. The biggest struggles were due to a difference in UI. I think having an optional ribbon UI would make a big difference to those familiar to Microsoft's products that are trying to switch.

I could never figure out the ribbon-style UI. Always went round and round looking for my frequently used menu options.

I'm presently half way between MSO and LO - I use MSO in a VM for legacy work, mainly in Excel. And I use LO for all my other work.

I hated the ribbon style of a lot at first because only about 20% of my time was in ribbon. Once my home and all school computers switched to ribbon, I picked it up quickly. I also like knowing the hot keys to get stuff done (Alt+h+b for bold, alt+h+m+c for merging cells, etc.).

> it hides too much stuff

The Ribbon was created to improve discoverability, in part because users were constantly requesting features that the Office products already had. User testing was performed to verify that the Ribbon improved discovery. I believe there is/was a series of blog posts on the Office blog that go into great detail about the design.

> Just a taste thing though I suppose.

Could be. Or perhaps the UI/UX folks in the Office organization did the testing incorrectly.

A ribbon-style UI will be optionally available not too long from now.

You can already try it out in LibreOffice 5.3 by ticking the "Enable experimental features (may be unstable)"-checkbox in the settings under LibreOffice->Advanced. Then in the menubar, select View->Toolbar Layout->Notebookbar.

https://www.youtube.com/watch?v=WWaLxFIVX1s

That's the reason I feel google's play for the classrooms is a great entry to break Microsoft office monopoly in the workplace. Though would have preferred an open source office suit to do so.

Reply on news.ycombinator.com