The number of weekly downloads has simply been increasing over time, due to growing popularity. Calling each new high "record-breaking" is a bit meaningless.
Unix timestamp counter breaks record after writing this comment.
It's the tried and true headline method of "event 2 happened after event 1", implying a causal effect between the two, despite no evidence of such effect existing. Technically the headline is correct, in the same way that "Canada Tariffs $20b in US Goods After Amazon Plane Crash" is also correct.
There has to be a much stronger correlation between good press for a software release and the subsequent downloads of it, even if a certain number of downloads were inevitable.
It actually happened. The US GNIS system was updated, and thus the apple maps and google maps have updated their maps to reflect
Other map providers will have to make a decision, but these sorts of things have historically been fraught with politics, I think most maps will go with whatever the official source they use has declared
It is actually no correct. A record by definition is an unsurpassed accomplishment or statistic. It needs to be an accomplishment before it can be called a record. The passage of time does not qualify as a record. Also, a statistical certainty cannot be called a record.
The year 2025 progressed to 2026. Never in the history of the Gregorian calendar have we EVER made it to 2026. That is an accomplishment. That is a record.
We do consider the oldest person alive and the oldest recorded person ever, to be records of a sort. So if that counts, why wouldn't a personal best count?
There is a semantic difference between capturing data (recording) and putting somethng on record (capturing/recording an achievement). One is the process of documenting and the other is the documented result that acts as a benchmark.
When you record an achievement, there is no need to record another one until you surpass it. You can record attempts, but they are now just data that makes the actual record more meaningful the longer it exists without being surpassed.
But still nice to hear. Really timely, too, since my 83-year-old mom called me the other day to help her E-mail a document she prepared in LibreOffice. I installed it on her (seldom-used) computer years ago. So I talked her through E-mailing it with Thunderbird.
And I say "talked," because the jagoffs at Microsoft REMOVED Remote Assistance from every version of Windows used by people who are likely to need it. And MS Office is trash; the more people abandon it the better.
If she's anything like my mom, i.e. someone who has been using computers for a while now but never had the confidence to experiment with it, any small change will freak her out and there really is a long tail of those. It's not a matter of Linux not being good enough, being different is already a real problem.
>because the jagoffs at Microsoft REMOVED Remote Assistance from every version of Windows used by people who are likely to need it.
when did they remove quick assist? i thought i used it recently (within a month or two), but maybe im misremembering. a quick google still shows it as available for home installs.
To continue further down the IT family support software, I use DWService because it seems like it's almost impossible for my sweet friends and relatives to accidentally break it. It's like a cockroach once it's in there and i've been very thankful for that
>And I say "talked," because the jagoffs at Microsoft REMOVED Remote Assistance from every version of Windows used by people who are likely to need it.
But also, you'll notice that giant banner on the linked page. It's because the people "most likely to need it" are also those most likely to fall victim to the increasing number of remote assistance scams.
Their point was more that, if every 'weekly download count after a new release' (or whatever record) is consistently being broken on each new release due to generally increasing popularity, then it doesn't make sense to imply causality between the messaging around this particular release and its already-expected record-breaking status.
But that is perfectly consistent with the downloads incrementally going up all the time. It doesn't say anything about whether the recent policy changes increased that rate.
Excuse you? What I originally said was "I don't understand your comment". They posted a clarification, and the clarification still doesn't make sense. No, I haven't read the download numbers. How is that going "well akshually"?
No, it doesn't. If the claim is just that download numbers are not consistently trending upwards, they should have just said that. I would have no trouble understanding that. What they actually said is "Libreoffice didn't say the numbers were going up, they claimed a specific record." Well, so what? Nobody said Libreoffice said that! The OP said that, because if it's true, then it's not surprising that the latest stat (release or not) is the highest ever. Nobody in the thread so far has said anything that contradicts "the numbers were going up anyway, therefore the record might not be to do with the AI policy".
The numbers graph is spiky and I can imagine you could argue either way that they are generally trending up or not. But that's irrelevant!
This objection is completely meaningless unless the download numbers follow a consistent upwards trend. You are simplifying the new record to a boolean, but most people still find meaning in extraordinary events happening, even if they follow a general trend.
So I fail to see this discussion as anything but sophistry until anyone demonstrates that the download numbers are totally ordinary. It's typical HN contrarianism without any value, because it's enough to argue for the possibility of something, even when the data for checking it is easily accessible.
I think the confusion is that robrain somehow understood smokel's objection as being "all-time total downloads obviously increase over time, so this is meaningless", rather than what smokel actually said which was "the number of weekly downloads has simply been increasing over time". Which is why robrain quoted LibreOffice's "This is the highest number of downloads in the first week”, to counter the idea (which nobody suggested) that the record was just about all-time cumulative downloads going up.
That's then what tempfile was questioning how it contradicted smokel's claim (because it doesn't, just an imagined claim). Your points, that if the number of weekly downloads isn't actually consistently increasing then this could still be meaningful, are valid - but irrelevant to tempfile who was wasn't arguing that this is meaningless; you should argue that to smokel instead.
Their point is that weekly downloads may have been on a constant rise for years due to a constant rise in popularity. It's not as constant as absolute number of downloads, but the point is that "breaking record" is still something that's likely happening frequently and doesn't need to be related to any statement they gave. They don't share any numbers, at least in the post.
> In one week, the installer was downloaded more than 1 million times — not counting updates via Linux distribution repositories.
Either there's 1m people downloading new software "because no ai", or there's another popular thing that installs this, and is not counted in the "updates via linux repos". One is much much more probable than the other.
Disliking clunky autocomplete is very different from going out of your way to replace Word with Libre Office. The average office worker will tolerate ten bad assistants before they voluntarily deal with ODF formatting quirks in a corporate environment
Yeah, the ones that know obvious slop and won't ship it are the ones who will win. Unfortunately it seems like I'm stuck with the other half at my current job.
Unless they're is a negative force, such as people being put off by the lack of AI features. While a lack of such a response might not justify the headline used here, I think it is certainly worth noting.
Yes. But in this case as there has been a recent announcement that could be significant, the trend continuing has some significance because it means the announcement likely⁰ had no detrimental effect¹.
We don't know absolutely nothing here: there is a new variable in play and this result gives us some evidence of the effect of that.
--------
[0] It _could_ have done, but something else countered it, though this is less likely according to Occam's razor.
[1] As opposed to reducing interest in the tools. The figures² apparently don't look to support an increase in interest though.
[2] Going by what others here are saying. The linked text doesn't state any numbers, nor does the announcement from the LibreOffice project, and I don't have time to go digging further ATM.
Something isn't adding up. Here's what the Document Foundation says about 26.8 [1]:
> 1,031,162 downloads
> This is the highest number of downloads in the first week of any LibreOffice release! These are just stats from our official downloads page, of course – many Linux users will have acquired the new release via their distribution’s package repositories.
But their weekly stats page [2] does not show anywhere near that many downloads. It shows a small number in the first week, then a bit over 700k next week, then almost a million in the third week. Even if we combine the first two weeks on the chart under the assumption that release weeks might not align with the weeks using in the chart it is nowhere near a million.
Also the mouseover popup on the stats page is showing 26.8 several weeks before release.
Update. Looking at the stats in day view instead of week view, the downloads for their release week were actually around [1] 1.259 million, using 2026-08-26 as the release date. That's the date in the release notes.
However, if we take the release date as 2026-08-25 then the 7 day total is exactly what their blog post said.
Also by comparing the week numbers to the daily numbers, we can determine that their week 35 is August 23-29 (Sunday to Saturday). They are using US week numbering, not ISO week numbering. In ISO it would be 24-30 (Monday to Sunday).
The Document Foundation is based in Germany so I kind of expected they would use ISO weeks.
[1] "around" rather than the exact number because this originally was an update to the original comment, but the edit window closed while I was writing it. When you submit in those circumstances you don't get an error and your editing area is undisturbed. You simply lose the editing area. I'm not going to go and add up everything again, so am going from memory of the number.
the stock market goes up and down daily, but with a functioning economy it will overall be up. from a distance you are right, something that keeps going up keeps achieving new highs.
but, if you follow the market on a daily basis, you will be paying attention to the market swooning for periods of time, maybe with a small recovery and then another swoon. New "all time highs" do provide convenient notifications that the economy has now grown from before. Without new highs, your investments are treading water.
I'm planning to make a killing when somebody unretires me to fix all the issues.. Or maybe not, I hope my retirement accounts are so large I can laugh at their "piddly $500/hr offer".
The hourly pay will be $1200, no less. Thanks to inflation this amount will cover just about the cost of two second-hand teabags and a mouldy lump of bread.
Banks are known for running mainframes which handle dates differently and will not be rolling over. The ones that use unix are likely to 64 bit systems that won't roll over.
Switching to the monthly view, the downloads are actually mostly flat going back to 2019
I can believe that the downloads record has been broken in the same way that the stock market hits record highs all the time because the long term trend is upward.
I don’t see anything to suggest that LibreOffice is seeing a sudden surge in popularity or that lack of AI features is driving a surge of downloads.
If anything, it would be seeing a surge because of AI. LibreOffice cli is the go to method for doing office documents -> pdf -> image rendering for agent doc rendering feedback.
Not sure what is "bold" about that claim, soffice is used in a ton of "document generation skills" and as soon as those tools start making into people's CI pipelines there will inevitably be spikes in download activity.
Tbh I wonder if it's related to AI at all, or just a consequence of Microsoft being institutionally unable to ship working software anymore and people are finally tired of spending so much money for such garbage products.
Buddy of mine tried to use some Exchange tools to troubleshoot a mail flow issue this morning. Puts in his query parameters (start/end date, and a user's email address). 1774 results, and the table below that shows "No data available." It'd be fucking comical if his company wasn't spending tens of thousands per month on their 365 infrastructure that's seemingly broken in new and exciting ways every. single. fucking. day.
I'd hardly consider fluctuations between 2.2 million and 3.3 million dowloads "flat".
Also, the metric used in the article was "downloads in the first week after a new release", which apparently broke a record. Regardless of the reason behind that record breaking number, switching to monthly dowloads to see if there is a sudden surge in popularity seems like the wrong metric.
Standard Linux software is often obtained through package management. Is LibreOffice self-updating? Is an update counted as a "download"? What about the torrents, Snap, Microsoft Store...?
Companies are totally blind to the fact that A.I. is a complete turn-off to many customers. Not having / using A.I. may even become a unique selling point in the future.
The thing is, all they really have to do is add an off switch for AI features so they’re not taking up space in the UI and aren’t consuming any extra resources. This toggle should also hide aggressive AI promotions in places like launch screens.
That alone would be enough to make most with a distaste for AI happy. It doesn’t have to be binary, but no, it gets shoved down everybody’s throat whether it’s wanted or not, and that’s how you induce fatigue.
They shouldn't and they won't. That will introduce a new metric -- %user disabling AI, and it will definitely be non-zero, which threatens the "everyone loves AI" narrative, which is the only thing thats floating all the tech-related valuation right now.
How can you be so disconnected from reality? The valuations are based on the immense value of LLMs, which exists regardless of how consumers - who are very much not where the revenue is coming from - feel about it.
I can see how much we are spending on Claude and codex at meta. Believe me, the AI company valuations are not coming from Microsoft word ai assistant.
What immense value? Has it solved world hunger? Has it provided abundant clean energy that no one needs to worry about it? Has it taken care of people's basic needs so that they can be free of work and focus on what's rewarding for them?
Although, some services have started labeling their old natural language processing functionality “AI,” and released additional “AI” features, so an off switch isn’t always possible. The options seem to be to lose previously existing functionality or gain current anti-features.
In their defense “AI” is a pretty nebulous concept.
Really what I want is a “give me 2020 software” toggle. But if we’re wishing, a 2010 toggle would also probably be nice…
> Really what I want is a “give me 2020 software” toggle. But if we’re wishing, a 2010 toggle would also probably be nice…
No kidding. It feels like by some time around 2015 the majority of the commercial software development world gave up entirely on trying to improve value for users. Given the choice, there’s only a handful of software that I wouldn’t prefer the 2010-2015 version of.
That's why for productivity software I vastly prefer open source. Sure it can be janky sometimes but at least I don't feel like I'm being fleeced and sold to the highest bidder.
That's a blanket statement. People like AI when it's actually useful, like photo editing tools that can remove unwanted objects from the background. Or how in iOS 27 they added AI to the Shortcuts app so you can describe a shortcut and it'll build it for you. It's turned the app into something I practically never used to something that's actually got me automating repetitive tasks on my phone.
The issue is the past few years of AI being shoved into everything with no taste being applied because it impresses shareholders. If you implement it in a way where it's actually useful, people will like it.
Why not just make it opt-in? Seems bad to leave the desires of nearly half your users unaddressed. As long as you're not shoehorning in features or making them unavoidable or the central focus, the only people who'd take issue would be zealot types who want to control other people's choices.
For enterprise software it's a feature. For most consumer software it's an anti-feature. Each of these crowds are on opposite ends of the same axis: the executive class coalesces around the mantra that LLMs bring automation benefits that will positively affect the org's bottom line. The populist belief is that LLMs and generative models are at best annoying and at worst anti-human.
At the risk of sounding trite there is a kernel of truth to both of these camps and things will eventually shake out. As LLM capabilities plateau and people broadly realize that they excel in domains that don't require discretion or taste the market will settle someplace in the middle.
I'm flexing my PM muscles by working on small personal projects and am constantly asking myself 1) what problem is what I'm building solving, 2) which parts of solving this problem are largely janitorial/annoying, then 3) could an LLM be leveraged (not used, leveraged) to reduce struggle. My golf app and fitness tracker telling me "lol u 3-putted too many holes" or "bro yr sleep is cooked - have you tried breathing slowly" do not fall into this category. Feeding an apartment lease into an LLM and having it enumerate the ways that my landlord is trying to push the boundaries of tenant law does.
Most people don't care that much. When I read some comments on Mastodon I get the impression that people absolutely hate it, then I'm reading some comments on X and people love it and build things with it. Then I'm talking to friends and they are excited about what ChatGPT can do for them and don't even think about the bigger picture outside of a "AI will take our jobs haha" joke.
How many, though? It looks to me like the anti-AI crowd is not all that big, but it’s a bubble so they think everyone thinks that way. Especially if we’re talking about people who are so rabidly against it that they’ll get upset about optional AI features that are completely ignorable.
That's true of any feature you don't use. And while it's nothing new for people to get upset about work being put into features they don't care about, I've never seen anything like this level of vitriol. In particular I never saw anyone say they won't use a piece of software at all if it contains a particular optional feature, until LLMs came along.
I have an app with several AI features that I will not mentioning use AI at all because its completely irrelevant and the users will hate it for no reason.
So you recognize that your users will dislike these features. And your solution is to try to hide it, rather than removing the features you know your users won't like? You do you, but that doesn't seem like good product design to me.
im using it for mundane text processing. categorizing and merging recipe ingredients for shopping lists. the users like the feature, unless i tell them its AI. i dont feel any ethical need to disclose that. its just software
So if I was black and racists were using my app that would stop if I put that prominently in the app, i should feel ethically compelled to disclose that?
> Companies are totally blind to the fact that A.I. is a complete turn-off to many customers.
In a normal office environment I would guess most people actually want more AI features.
Most people aren’t passionate about hand writing their best TPS reports with carefully crafted prose. They want to get their work done as easily as possible at their job so they can go home on time.
> I would guess most people actually want more AI features.
Not if they spend an unreasonable amount of time fighting the AI features. The people I know who are forced to use AI for their work spend a disproportional amount of the time fighting it.
Experience has shown that bad prose, grammar and orthograph has been accepted in business setting for the last 3 decades, basically since people stopped writing cards/letters and started communicating with sms and later instant messaging apps. Writing skills globally went downhill. The thing is: virtually nobody cares anymore.
What I see is rather management forcing employees to use AI, menacing them of bad yearly review if metric show they don't use them.
The TPS reports are written with the sender's AI and then read and summarized with the recipient's AI. At this point it would make more sense to just write the TPS reports in MCP.
Conversely, a no-AI ethos is one that will very likely need to be walked back at some point in the future for tech products. AI isn't going anywhere, so finding ways to use it constructively with your product is going to be a far better long term strategy. For now while it's new and scary and there's a lack of regulation and governance and ethical considerations - I can see why an anti-AI stance is popular, but it's not durable long term.
I'm a big supporter of LibreOffice, Codeberg, SourceHut, etc. on what they do, tell and defend, and I'm with them.
On the other hand, I'm not anti-AI, but I don't want it to be shoved down my throat in any and every app I have. I don't want that sparkly "Enhance this slide" button, "WhatsApp AI" in the corner, automatically generated "AI summary" in Google/YouTube, etc.
However, I'm fine with an unobtrusive "Summary" button in my search results (Kagi), or a tab which I can use with some models (Notion), "use ML" checkbox while resizing or "enhancing" images (Pixelmator Pro), or many small touches Apple enables with on-device models which doesn't have AI written all over it.
In short, I want consent, control and dignity in my workflows. Yes, I'm very against how these LLMs are trained, built and how the datacenters they run on harm the environment we live in. However, I'm not against the tech. I want responsible development of it, for the customer instead of their owners which wants to milk their users like content and cash cows.
I really appreciate this comment since it very effectively communicates how I feel about the subject, since I have been struggling with putting it into such clear language.
The internet has been around for quite a while now, and is absolutely critical for a lot of applications. Most people would never give up the internet.
Yet there are still many applications without explicit mandatory internet tie-ins like cloud sync. In fact, "stores data locally" and "won't disappear with zero notice" are valuable features to many people!
Just because AI is a no-brainer must-have for some apps doesn't mean it is a required feature for all apps - let alone one shoved in your face every few minutes.
I feel like the equivalent for the rise of the internet would be an explicitly anti-internet stance -- no website, no downloads, no online customer support, everything routed through snail mail, fax, and telephone (but no VoIP!) as an intentional ideological point.
Which products that sell this are broadly popular? I can see some people on HN wanting it, but people I know who are outside that bubble don't even think of it as an issue.
> The criteria we have listed do not amount to a definitive rejection. The AI sector is evolving rapidly; on-premises inference is becoming manageable on standard hardware; and open models are improving much faster than most of us had anticipated. Should an approach emerge that meets all the conditions, we will evaluate it very carefully.
That list seems pretty easy to meet today. I’m not sure why they’re waiting for an approach to emerge instead of building it now, unless there’s something else they haven’t listed.
"User-controlled execution." Seems easy enough to add settings so this is configurable.
"No content may leave the computer without authorisation." OK, so don't do that.
"No telemetry of any kind." There are plentiful API providers with zero data retention. If this means no data of any kind, even just records of having accessed the API, then local inference is a thing.
"No dependence on a single provider." Trivial. The OpenAI API is a de facto standard implemented by tons of providers. If you really want to go broad, it's easy to implement clients for other APIs too.
"No compromises on format." I don't even understand how this would be a problem in the first place. Text goes in, text comes out. The client can put it into whatever format it wants.
"Entirely optional." Trivial to add an off switch. Or an on switch, so it's off by default. If it must be completely removable and not just turned off, it can be implemented as an extension.
I don't think there's enough information accurately assess this. Anti-AI movements might grow much larger than today, given AIs potential to replace jobs faster than we have the ability to create them. The movement could even result in AI going somewhere.
> is one that will very likely need to be walked back at some point in the future for tech products
No doubt, but that future is when companies have figured out proper fit for AI in their product context and preferably with local AI models.
It should not be with the current slap AI anywhere and everywhere approach that uploads and trains on all available user data, which is what everyone is opposed to.
I hate products that want me to find a use-case for their feature that they wanted to develop. It's great that some people love AI features, but for me, they are rarely solving an actual problem I have. But now I have to relearn the software because the company wants to go all-in on AI? Why should I? One example of that is Notion. I've loved it for years, and I'm still using it, but it's starting to lean heavier and heavier into an AI-first workflow and I don't like it to the point where I consider switching to something else regularly.
This is awesome! I have used LibreOffice (and Open Office) extensively for decades now. It is my real office suite I use for real work and real business accounting tasks. Yes, I am very well aware that Calc is different than Excel but it can get nearly all the same tasks done as long as you are not tied deep into a corporate environment passing Excel-specific files around. I have also made great business uses of it in Docker too, generating perfect PDFS from templates, and normalizing document formats for analysis. I donate a modest $19/month personally to the foundation, much rather pay them than Microsoft!
Normalize into flat xml that I could then parse in Nokogiri.
Edit for clarity: I am a Rubyist at heart so I reach for Ruby for almost all of my automation, ML, and even now AI stuff. Fro a very long time, I have been able to use CLI LibreOffice to do amazing things. It can read everything, even obscure formats by today's standards, going back into the 1990s. And then once in OpenOffice XML you have one target to parse in your scripting language of choice to do document analysis, data loss prevention detection, and more. All with a clean open source license with no server usage restrictions, unlike MS Office.
Bundling in codex isn’t going to increase the download count on their website. Unless you’re assuming that every codex install downloads a fresh non-cached copy.
They have absolutely zero regard for system resources. I understand why, but such things were once considered unfeasible - and very recently were also considered undesirable.
I still use a laptop with a 100 GiB hard drive - only 80 of which are allocated to the /home/ directory. This would cut into the useable space significantly on that machine. How about Chromebooks?
The article says LibreOffice is almost 20 years old. Well, before it was LibreOffice it was OpenOffice in 2000 and before it was OpenOffice it was StarOffice and StarWriter originates in 1985. So you could say it (or the Writer component) is 41 years old.
Let's use AI to test the validity of an AI generated clickbait article.
Fable 5.1 was presented with the following prompt to do research.
> https://manualdousuario.net/en/libreoffice-download-record-n... states an interesting causality. Generate an artifact that tests this claim. Is it causal or coincidental. ie if libreoffice downloads have been increasing every release then it's possible to claim any recent feature change is responsible for the download. You need to get access to the download stats and then see what other major features could be responsible
Didn't people find it bundled with a bunch of other things in a single runtime? Downloaded as part of a whole runtime would suggest it wasn't downloaded directly from the source.
i think that's a perfectly coherent and reasonable way for both these apps to work.
people don't want AI features leaking into their non-ai apps. I'm not an AI hater - I want my AI app to be able to read and write word docs, but when i decide to open up libreoffice i'm doing that because i want to write something myself, not use AI. i don't need my word processor to bundle an inferior version of what the ChatGPT can do for me.
I think it's not baked into codex, but codex has either been instructed to download it when asked to edit word docs/excel spreadsheets, or it picked it up during training.
I've been using Libreoffice to store my diary for years -- IIRC they use blowfish on the crypto side of things? I had a weird issue a while back where I'm not sure if I forgot my diary's password, or the file got corrupted.
I've had that happen a few times, each with a long space in between I keep things on paper.
It's a great project that doesn't get enough love.
Why Libreoffice? Does your diary contain advanced formatting, tables, images, or other involved content? Or just for the encryption?
I ask because I also keep a diary and have refined how I store it and secure it, over 25+ years. Only recently did I find something that I like better than paper.
Here's hoping increased usage will lead to improvements in the "table" functionality. Like when you can define a table, and hit tab at the last column will put you at the next row first column. Missed that from MS office.
If anyone else wants to be rid of it (courtesy of astra):
Set NotebookbarColorTransparency to 255. This removes the colored tint while keeping the ribbon.
Go to Extras → Optionen → LibreOffice → Erweitert → Experteneinstellungen.
Search for NotebookbarColorTransparency.
Double-click the matching row under org.openoffice.Office.Common → Misc.
Change 150 to 255.
Confirm with OK, close the settings, and restart LibreOffice.
Both Codex and Claude (on the web) use LibreOffice to edit docs, so not sure you can really conclude it broke download records because end users didn't want AI in Office and went with LibreOffice instead
I don't buy the AI argument. I'd bet on Linux replacing Windows for a good chunk of desktop users. For example, I installed LibreOffice on my parents' PC last month, because I replaced their Windows with Linux Mint.
But they support AI features, like LanguageTool. I always bought Softmaker but since they don't offer this feature in their One-time purchase version, I have started using LibreOffice only.
Maybe it's just a secret cabal of rogue agents downloading LibreOffice for no particualr reason whatsoever? I mean there is a precedent already: those fellas hijacked some poor wiki.
I came here to say that this is extraordinarily uninteresting. But some useful thought came out of it: all "AI" features are moving into the generalized agent harness anyways, and good software often maintains a focus. Take Windows, for example. It has a rather broad job being an OS, but as a product it's a good example of loosening focus. it's losing consistency in its opinions about where updates should happen, how settings should apply, which menus are in charge, and even what the design philosophy is. And that bleeds out into the user surface too. Windows 11 is a scattered mess that's increasingly just a surface for finding new ways to pull your attention into ads.
Maybe software just needs to remain straightforward and focused, and we let the ChatGPT and Claude apps handle the AI part of things.
Or people outside the US distrusting US technology? There is a confluence of trends pointing away from US based tech. It will take a crazy amount of work to restore trust.
Another reason why it might be hitting records is many companies getting rid of Microsoft 365 downloads for desktop and only allowing use of the web version. And apparently it's the first thing that comes up when you try to download Microsoft 365? People don't try to pirate it anymore, they just download LibreOffice
I switched to LibreOffice just recently, but only because i’d been happy with OpenOffice for years and only recently decided to see if there was an update and found out it had been discontinued in 2011.
It was still working fine for me tbh, but i thought better to update to something newer.
Offtopic blogging tip: Americans will read dates like "8/9/2026" as "August 9, 2026." I'm pretty sure the author is using the (more sensible) convention of day/month/year but expressing dates as e.g. "08-Sep-2026" removes the ambiguity.
this is temporary marketing towards a direction that they may change their mind on in the future, which would likely anger people who downloaded it because of this preference
not a call I'd make early on, a preference to disable AI features like Firefox added is maybe a more reasonable call
all we know today is that AI is being used in annoying ways, this won't be forever
It has no AI built in, which is a feature to a growing set of users. No AI buttons all over the place and sending your personal data off to who knows where. And if you want AI, you can add your choice of AI via plugins if you want AI in your word processor.
Sure, but local LLMs are getting smaller all the time... so it could in the not so distant future ship with its own locally-hosted LLM embedded (no data leaving your device).
I guess it could always be relegated to plugins, but I suspect over the next 5 years the line is going to get a lot blurrier.
Awesome. Now can we fix the bugs and make it more usable? I’m longing for a Microsoft/Google alternative but they are all still clunky as hell and lack the integration of Outlook. Im currently using Planner as an excel replacement and it is painful but less painful than LibreOffice.
Hopefully LibreOffice is one of the projects that gets free Pro Max accounts from A and O.
Corporations have been trying Star/Open/LibreOffice migrations for literally 30 years at this point. And, at the pace it’s been going, it will probably take another 30 for it to reach Word 95 polish.
You’re wrong. Despite its usefulness in our profession, AI is deeply unpopular with the general public and products that don’t include it are rapidly gaining in popularity. People are tired of AI being shoved down their throats without consent.
Are you speaking from personal experience or do you have data to back that up? I honestly would love to see some study around this. I haven't done my research but --anecdotally-- everyone around me is ok with it or even enjoying AI to some degree. No strong stances either way tbh.
Here you go: https://archive.ph/9erfx bask in that 20% support of "more good than bad when it comes to news" and the 70% thinking it's destroying jobs.
Regardless of the fact that the 'record download' levels is probably due to some other factor, the world is already becoming conscious that the so-called AI is not all it's cracked up to be.
Very annoying that AI is this new virtue signaling topic. I guess it's a step up from a few years ago when it was about naming branches, but it's annoying.
Libreoffice is free software that's the big point of it. You can mod it and stip what's you dislike. I'd be glad if people developed AI integration with it. One could do that with AI itself. I've successfully modded free software with AI in many ways, adding features to Blender, for example.
This binary pro of pro and anti AI buzz is just annoying. Privacy is important. But AI is treated by some people as if it was some nuclear waste. In fact it's very similar to the anti nuclear green movement. It's an emotional aesthetic response.
The actual story below the headline is that AI is allowed in LibreOffice via plugins, they just aren't going to do it themselves since it violates some of their principles about privacy and local data storage.
They aren't against AI, they just have certain promises made around the product that would be hard not to break if they added first party AI.
Not a fan of pushing LLMs into everything, but this is one of those things where it would be great to have a powerful(or specialized) local llm to help you
Write a formula, make a macros, filter stuff, formatting. I'd rather not to Google it. Especially considering SEO made it impossible to find information how to do stuff in editor X of version Y, at best you'd get thrown with crap for other version and have to figure out how to adapt it for your version
You can still ask a general model to do that and it will probably have a nearly full success rate at this point. Either a good open weights or something like Astra.
Yeah, but it would in many cases remove you as a meat proxy when your relative or a friend or your boss ask you how to make X when they have a box where they can write to a genie "I want X" and it would put everything in place for them
I don't see not having AI as a feature, but I'd prefer for them to take the approach of optional, easy to implement interfaces that increase the ability for the suite to be 'scripted' rather than implement AI directly. My motivation is that the field of AI is moving so fast I think its a better long term plan to make your software something that can be automated with AI than to try to fight the speed of that innovation curve. I get that even that can be difficult because of security concerns, but it seems more straightforward than implementing models directly into your application.
That is, in fact, exactly what they explain as their approach in the article, as well as their reasoning behind it.
It isn't that they disallowed AI, its that they just leave it up to plugins and have no native interface since it violates many of their principles around privacy and local data.
I absolutely see a lack of soulless clankers to be a feature.
I completely agree that focusing on "scripting" would be great. Clanker proxies can allow their clanker to use it while grumpy bastards like me can stick to Bash to automate it.
The number of weekly downloads has simply been increasing over time, due to growing popularity. Calling each new high "record-breaking" is a bit meaningless.
Unix timestamp counter breaks record after writing this comment.
It's the tried and true headline method of "event 2 happened after event 1", implying a causal effect between the two, despite no evidence of such effect existing. Technically the headline is correct, in the same way that "Canada Tariffs $20b in US Goods After Amazon Plane Crash" is also correct.
"it's our most powerful X ever" - every Apple event ever
God, I'm hoping that this phraseology changes with the new CEO, I am so tired of that overused statement haha.
Didn’t Steve Jobs use that formula as well? The most powerful thing we released compared to last year’s thing we released?
He loved the term “revolutionary”
A woman can dream!
Post hoc ergo propter hoc.
The only reason I know this is the West Wing. I guess some TV applies in the real world too...
David Hume would argue that the causal effect is never seen between 2 events so what's the difference?
There has to be a much stronger correlation between good press for a software release and the subsequent downloads of it, even if a certain number of downloads were inevitable.
LibreOffice breaks download records after Trump renames Lake Ontario to Lake America
Did that actually happen as something that appears on maps or did it just get ignored as senile rambling?
It actually happened. The US GNIS system was updated, and thus the apple maps and google maps have updated their maps to reflect
Other map providers will have to make a decision, but these sorts of things have historically been fraught with politics, I think most maps will go with whatever the official source they use has declared
Did that actually happen as something that appears on maps or did it just get ignored as senile rambling?
Both Google Maps and Apple Maps changed it for users in the United States.
MapQuest publicly refused, causing literally dozens of people to ask, "MapQuest is still a thing?"
you joke but number of mapquest downloads jumped 400% (from 1 to 5 weekly downloads)
David Hume reads the news
It is actually no correct. A record by definition is an unsurpassed accomplishment or statistic. It needs to be an accomplishment before it can be called a record. The passage of time does not qualify as a record. Also, a statistical certainty cannot be called a record.
The year 2025 progressed to 2026. Never in the history of the Gregorian calendar have we EVER made it to 2026. That is an accomplishment. That is a record.
We do consider the oldest person alive and the oldest recorded person ever, to be records of a sort. So if that counts, why wouldn't a personal best count?
I certainly celebrate mine every time.
It's not an accomplishment. It's obviously been recorded and is therefore a record.
There is a semantic difference between capturing data (recording) and putting somethng on record (capturing/recording an achievement). One is the process of documenting and the other is the documented result that acts as a benchmark.
When you record an achievement, there is no need to record another one until you surpass it. You can record attempts, but they are now just data that makes the actual record more meaningful the longer it exists without being surpassed.
But still nice to hear. Really timely, too, since my 83-year-old mom called me the other day to help her E-mail a document she prepared in LibreOffice. I installed it on her (seldom-used) computer years ago. So I talked her through E-mailing it with Thunderbird.
And I say "talked," because the jagoffs at Microsoft REMOVED Remote Assistance from every version of Windows used by people who are likely to need it. And MS Office is trash; the more people abandon it the better.
If I can get my mom on Linux... hey now!
What's the issue of getting her on Linux? I bet she wouldn't see a difference, especially if its GUI is similar to Windows, like Linux Mint.
If she's anything like my mom, i.e. someone who has been using computers for a while now but never had the confidence to experiment with it, any small change will freak her out and there really is a long tail of those. It's not a matter of Linux not being good enough, being different is already a real problem.
>because the jagoffs at Microsoft REMOVED Remote Assistance from every version of Windows used by people who are likely to need it.
when did they remove quick assist? i thought i used it recently (within a month or two), but maybe im misremembering. a quick google still shows it as available for home installs.
I use Google remote desktop with my mother. It's not perfect but it's adequate.
To continue further down the IT family support software, I use DWService because it seems like it's almost impossible for my sweet friends and relatives to accidentally break it. It's like a cockroach once it's in there and i've been very thankful for that
RustDesk can be self-hosted.
>And I say "talked," because the jagoffs at Microsoft REMOVED Remote Assistance from every version of Windows used by people who are likely to need it.
https://rustdesk.com/
But also, you'll notice that giant banner on the linked page. It's because the people "most likely to need it" are also those most likely to fall victim to the increasing number of remote assistance scams.
From Libre’s announcement (linked in the article):
“This is the highest number of downloads in the first week of any LibreOffice release!”
I don't understand this comment. You don't seem to be disagreeing with OP?
There is a specific record being claimed.
They're not just saying the absolute downloads are going up, as the OP suggests.
Their point was more that, if every 'weekly download count after a new release' (or whatever record) is consistently being broken on each new release due to generally increasing popularity, then it doesn't make sense to imply causality between the messaging around this particular release and its already-expected record-breaking status.
And that's the case? Weekly download numbers have been consistently trending upwards?
They're saying the weekly numbers grow a little bit every week, so every week is technically record breaking.
The absolute number of lifetime downloads can of course only go up (but that's even more vacuous).
Their last sentence:
“Unix timestamp counter breaks record after writing this comment”
makes that vacuous connection, hence my original comment.
But that is perfectly consistent with the downloads incrementally going up all the time. It doesn't say anything about whether the recent policy changes increased that rate.
I'm sure you've actually checked the download numbers and are not just going "well akshually", right?
Excuse you? What I originally said was "I don't understand your comment". They posted a clarification, and the clarification still doesn't make sense. No, I haven't read the download numbers. How is that going "well akshually"?
The clarification makes sense, unless you presuppose that download numbers are consistently trending upwards.
No, it doesn't. If the claim is just that download numbers are not consistently trending upwards, they should have just said that. I would have no trouble understanding that. What they actually said is "Libreoffice didn't say the numbers were going up, they claimed a specific record." Well, so what? Nobody said Libreoffice said that! The OP said that, because if it's true, then it's not surprising that the latest stat (release or not) is the highest ever. Nobody in the thread so far has said anything that contradicts "the numbers were going up anyway, therefore the record might not be to do with the AI policy".
The numbers graph is spiky and I can imagine you could argue either way that they are generally trending up or not. But that's irrelevant!
This objection is completely meaningless unless the download numbers follow a consistent upwards trend. You are simplifying the new record to a boolean, but most people still find meaning in extraordinary events happening, even if they follow a general trend.
So I fail to see this discussion as anything but sophistry until anyone demonstrates that the download numbers are totally ordinary. It's typical HN contrarianism without any value, because it's enough to argue for the possibility of something, even when the data for checking it is easily accessible.
I think the confusion is that robrain somehow understood smokel's objection as being "all-time total downloads obviously increase over time, so this is meaningless", rather than what smokel actually said which was "the number of weekly downloads has simply been increasing over time". Which is why robrain quoted LibreOffice's "This is the highest number of downloads in the first week”, to counter the idea (which nobody suggested) that the record was just about all-time cumulative downloads going up.
That's then what tempfile was questioning how it contradicted smokel's claim (because it doesn't, just an imagined claim). Your points, that if the number of weekly downloads isn't actually consistently increasing then this could still be meaningful, are valid - but irrelevant to tempfile who was wasn't arguing that this is meaningless; you should argue that to smokel instead.
Absolute number of downloads inevitably goes up. Weekly downloads number doesn't.
> Weekly downloads number doesn't.
The ChatGPT app downloads a copy of LibreOffice when you ask it to make pptx files.
So I've downloaded LibreOffice 3 times this week without really paying attention to it.
Ironic since the headline implies that people are downloading it specifically to avoid AI features being shoved into our document editors.
Their point is that weekly downloads may have been on a constant rise for years due to a constant rise in popularity. It's not as constant as absolute number of downloads, but the point is that "breaking record" is still something that's likely happening frequently and doesn't need to be related to any statement they gave. They don't share any numbers, at least in the post.
> In one week, the installer was downloaded more than 1 million times — not counting updates via Linux distribution repositories.
Either there's 1m people downloading new software "because no ai", or there's another popular thing that installs this, and is not counted in the "updates via linux repos". One is much much more probable than the other.
> or there's another popular thing that installs this
Yeah, an AI tool:
The ChatGPT/Codex app bundles a full copy of LibreOffice
https://news.ycombinator.com/item?id=49527396
and the bundled copies will contribute to installer download statistics?
You underestimate the public's hatred of AI-in-everything. HN is a bubble of AI.
Disliking clunky autocomplete is very different from going out of your way to replace Word with Libre Office. The average office worker will tolerate ten bad assistants before they voluntarily deal with ODF formatting quirks in a corporate environment
I fully expect the bubble is the people in tech who have strong feelings about the goodness or badness of AI one way or another.
Yeah, the ones that know obvious slop and won't ship it are the ones who will win. Unfortunately it seems like I'm stuck with the other half at my current job.
Global population is still rising. In general you should expect numbers to go up.
Unless they're is a negative force, such as people being put off by the lack of AI features. While a lack of such a response might not justify the headline used here, I think it is certainly worth noting.
If you know absolutely nothing, though: expect usage of something used by humans to scale with the number of humans.
Yes. But in this case as there has been a recent announcement that could be significant, the trend continuing has some significance because it means the announcement likely⁰ had no detrimental effect¹.
We don't know absolutely nothing here: there is a new variable in play and this result gives us some evidence of the effect of that.
--------
[0] It _could_ have done, but something else countered it, though this is less likely according to Occam's razor.
[1] As opposed to reducing interest in the tools. The figures² apparently don't look to support an increase in interest though.
[2] Going by what others here are saying. The linked text doesn't state any numbers, nor does the announcement from the LibreOffice project, and I don't have time to go digging further ATM.
Congratulation to your new unix timestamp record! (and yes, you too, nth reader!)
I can tell you wrote this comment when you were younger.
First!
Yes, here's the graph of weekly downloads: https://stats.documentfoundation.org/downloads#week,version The 26.8 peak is within the normal range of week-to-week variation.
> Calling each new high "record-breaking" is a bit meaningless
It is the first to break a million in a week. Even if it has been steadily going up most people would consider passing that number interesting.
That would make sense, but according to the stats posted in a sibling comment they crossed the million weekly downloads in week 31 already.
Something isn't adding up. Here's what the Document Foundation says about 26.8 [1]:
> 1,031,162 downloads
> This is the highest number of downloads in the first week of any LibreOffice release! These are just stats from our official downloads page, of course – many Linux users will have acquired the new release via their distribution’s package repositories.
But their weekly stats page [2] does not show anywhere near that many downloads. It shows a small number in the first week, then a bit over 700k next week, then almost a million in the third week. Even if we combine the first two weeks on the chart under the assumption that release weeks might not align with the weeks using in the chart it is nowhere near a million.
Also the mouseover popup on the stats page is showing 26.8 several weeks before release.
[1] https://blog.documentfoundation.org/blog/2026/09/02/one-week...
[2] https://stats.documentfoundation.org/downloads#week,version
Update. Looking at the stats in day view instead of week view, the downloads for their release week were actually around [1] 1.259 million, using 2026-08-26 as the release date. That's the date in the release notes.
However, if we take the release date as 2026-08-25 then the 7 day total is exactly what their blog post said.
Also by comparing the week numbers to the daily numbers, we can determine that their week 35 is August 23-29 (Sunday to Saturday). They are using US week numbering, not ISO week numbering. In ISO it would be 24-30 (Monday to Sunday).
The Document Foundation is based in Germany so I kind of expected they would use ISO weeks.
[1] "around" rather than the exact number because this originally was an update to the original comment, but the edit window closed while I was writing it. When you submit in those circumstances you don't get an error and your editing area is undisturbed. You simply lose the editing area. I'm not going to go and add up everything again, so am going from memory of the number.
A graph to confirm it: https://stats.documentfoundation.org/downloads#week,version
true, however this event seems like a significant daily spike: https://stats.documentfoundation.org/downloads#day,version
Isn't it also a dep of some AI tools? Or maybe they use their own binary
I feel this every time the stock market hits a new “all time high”
the stock market goes up and down daily, but with a functioning economy it will overall be up. from a distance you are right, something that keeps going up keeps achieving new highs.
but, if you follow the market on a daily basis, you will be paying attention to the market swooning for periods of time, maybe with a small recovery and then another swoon. New "all time highs" do provide convenient notifications that the economy has now grown from before. Without new highs, your investments are treading water.
Hell, it took like 5 years for the market to get back to its peak after 2008.
is the trend of increasing weekly downloads not itself meaningful?
clickbait gonna clickbait. But you cannot blame it because in today's attention economy it really HAS to since everything else is too
Calling each new high "record-breaking" is a bit meaningless.
Like how Hollywood is constantly saying each week's ticket sales are record-breaking, because they don't factor for inflation or population growth.
See also, every new iPhone/Pixel being the best they've ever release. You'd hope so!
What? A Unix timestamp literally cannot ever be lower. Its a completely assinine comparison isn't it?
This is comparative.
False, in January 2038 at least some systems will go back in time
I'm planning to make a killing when somebody unretires me to fix all the issues.. Or maybe not, I hope my retirement accounts are so large I can laugh at their "piddly $500/hr offer".
The hourly pay will be $1200, no less. Thanks to inflation this amount will cover just about the cost of two second-hand teabags and a mouldy lump of bread.
just make sure you get paid in cash before it rolls over.
Banks are known for running mainframes which handle dates differently and will not be rolling over. The ones that use unix are likely to 64 bit systems that won't roll over.
thats_the_joke.jpg
Just wait til 2038 ;)
I found this graph of LibreOffice downloads https://stats.documentfoundation.org/downloads
Switching to the monthly view, the downloads are actually mostly flat going back to 2019
I can believe that the downloads record has been broken in the same way that the stock market hits record highs all the time because the long term trend is upward.
I don’t see anything to suggest that LibreOffice is seeing a sudden surge in popularity or that lack of AI features is driving a surge of downloads.
If anything, it would be seeing a surge because of AI. LibreOffice cli is the go to method for doing office documents -> pdf -> image rendering for agent doc rendering feedback.
Yep. I use it locally for rendering PDFs via LLM tool use.
how does it compare to using pandoc, which is what I currently use for rendering PDFs via LLMs?
Came here to say this as well. It's the obvious all-encompassing free office document processing suite and LLMs know how to utilize it well.
Codex bundles it with the desktop app as well for the same purpose.
Highly doubt this. Do you have a source or numbers on that bold claim?
https://simonwillison.net/2026/Sep/1/codex-libreoffice/ - https://news.ycombinator.com/item?id=49527396
(and I can also just add that it's my go-to tool for that exact use-case, and used it in a dozen different client projects)
Not sure what is "bold" about that claim, soffice is used in a ton of "document generation skills" and as soon as those tools start making into people's CI pipelines there will inevitably be spikes in download activity.
https://github.com/anthropics/skills/blob/main/skills/docx/S...
Tbh I wonder if it's related to AI at all, or just a consequence of Microsoft being institutionally unable to ship working software anymore and people are finally tired of spending so much money for such garbage products.
Buddy of mine tried to use some Exchange tools to troubleshoot a mail flow issue this morning. Puts in his query parameters (start/end date, and a user's email address). 1774 results, and the table below that shows "No data available." It'd be fucking comical if his company wasn't spending tens of thousands per month on their 365 infrastructure that's seemingly broken in new and exciting ways every. single. fucking. day.
> Switching to the monthly view
I'd hardly consider fluctuations between 2.2 million and 3.3 million dowloads "flat".
Also, the metric used in the article was "downloads in the first week after a new release", which apparently broke a record. Regardless of the reason behind that record breaking number, switching to monthly dowloads to see if there is a sudden surge in popularity seems like the wrong metric.
I wouldn't say that's flat at all, if you switch to the "architecture" graph it gives a clearer picture without the between-version artefacts of dips that don't exist: https://stats.documentfoundation.org/downloads#month,arch
What matters more than the peaks are the floors, which are clearly increasing.
What methods do these "downloads" include?
Standard Linux software is often obtained through package management. Is LibreOffice self-updating? Is an update counted as a "download"? What about the torrents, Snap, Microsoft Store...?
https://www.libreoffice.org/download-other/
Companies are totally blind to the fact that A.I. is a complete turn-off to many customers. Not having / using A.I. may even become a unique selling point in the future.
The thing is, all they really have to do is add an off switch for AI features so they’re not taking up space in the UI and aren’t consuming any extra resources. This toggle should also hide aggressive AI promotions in places like launch screens.
That alone would be enough to make most with a distaste for AI happy. It doesn’t have to be binary, but no, it gets shoved down everybody’s throat whether it’s wanted or not, and that’s how you induce fatigue.
They shouldn't and they won't. That will introduce a new metric -- %user disabling AI, and it will definitely be non-zero, which threatens the "everyone loves AI" narrative, which is the only thing thats floating all the tech-related valuation right now.
They will interpret that metric as "they don't love AI YET,so lets throw more at them so they can see how wonderful it is" .....
How can you be so disconnected from reality? The valuations are based on the immense value of LLMs, which exists regardless of how consumers - who are very much not where the revenue is coming from - feel about it.
I can see how much we are spending on Claude and codex at meta. Believe me, the AI company valuations are not coming from Microsoft word ai assistant.
What immense value? Has it solved world hunger? Has it provided abundant clean energy that no one needs to worry about it? Has it taken care of people's basic needs so that they can be free of work and focus on what's rewarding for them?
Although, some services have started labeling their old natural language processing functionality “AI,” and released additional “AI” features, so an off switch isn’t always possible. The options seem to be to lose previously existing functionality or gain current anti-features.
In their defense “AI” is a pretty nebulous concept.
Really what I want is a “give me 2020 software” toggle. But if we’re wishing, a 2010 toggle would also probably be nice…
> Really what I want is a “give me 2020 software” toggle. But if we’re wishing, a 2010 toggle would also probably be nice…
No kidding. It feels like by some time around 2015 the majority of the commercial software development world gave up entirely on trying to improve value for users. Given the choice, there’s only a handful of software that I wouldn’t prefer the 2010-2015 version of.
That's why for productivity software I vastly prefer open source. Sure it can be janky sometimes but at least I don't feel like I'm being fleeced and sold to the highest bidder.
Yes, it needs to not be shoved everywhere and try to rewrite your workflows constantly and that would be enough
That's a blanket statement. People like AI when it's actually useful, like photo editing tools that can remove unwanted objects from the background. Or how in iOS 27 they added AI to the Shortcuts app so you can describe a shortcut and it'll build it for you. It's turned the app into something I practically never used to something that's actually got me automating repetitive tasks on my phone.
The issue is the past few years of AI being shoved into everything with no taste being applied because it impresses shareholders. If you implement it in a way where it's actually useful, people will like it.
For my text editor, KeenWrite[1], 42% of users answered that they'd use AI. Since the majority don't want it, no AI has been added.
[1]: https://keenwrite.com/
Why not just make it opt-in? Seems bad to leave the desires of nearly half your users unaddressed. As long as you're not shoehorning in features or making them unavoidable or the central focus, the only people who'd take issue would be zealot types who want to control other people's choices.
I would love to have codex plugins into libreoffice like m365 office.
The issue with M365 and MSFTs implementation is they did clipping all over again. They force you in the most useless ways.
For enterprise software it's a feature. For most consumer software it's an anti-feature. Each of these crowds are on opposite ends of the same axis: the executive class coalesces around the mantra that LLMs bring automation benefits that will positively affect the org's bottom line. The populist belief is that LLMs and generative models are at best annoying and at worst anti-human.
At the risk of sounding trite there is a kernel of truth to both of these camps and things will eventually shake out. As LLM capabilities plateau and people broadly realize that they excel in domains that don't require discretion or taste the market will settle someplace in the middle.
I'm flexing my PM muscles by working on small personal projects and am constantly asking myself 1) what problem is what I'm building solving, 2) which parts of solving this problem are largely janitorial/annoying, then 3) could an LLM be leveraged (not used, leveraged) to reduce struggle. My golf app and fitness tracker telling me "lol u 3-putted too many holes" or "bro yr sleep is cooked - have you tried breathing slowly" do not fall into this category. Feeding an apartment lease into an LLM and having it enumerate the ways that my landlord is trying to push the boundaries of tenant law does.
Most people don't care that much. When I read some comments on Mastodon I get the impression that people absolutely hate it, then I'm reading some comments on X and people love it and build things with it. Then I'm talking to friends and they are excited about what ChatGPT can do for them and don't even think about the bigger picture outside of a "AI will take our jobs haha" joke.
How many, though? It looks to me like the anti-AI crowd is not all that big, but it’s a bubble so they think everyone thinks that way. Especially if we’re talking about people who are so rabidly against it that they’ll get upset about optional AI features that are completely ignorable.
Even if the features are ignorable, they have consumed development time and attention to get there.
That's true of any feature you don't use. And while it's nothing new for people to get upset about work being put into features they don't care about, I've never seen anything like this level of vitriol. In particular I never saw anyone say they won't use a piece of software at all if it contains a particular optional feature, until LLMs came along.
I have an app with several AI features that I will not mentioning use AI at all because its completely irrelevant and the users will hate it for no reason.
So you recognize that your users will dislike these features. And your solution is to try to hide it, rather than removing the features you know your users won't like? You do you, but that doesn't seem like good product design to me.
im using it for mundane text processing. categorizing and merging recipe ingredients for shopping lists. the users like the feature, unless i tell them its AI. i dont feel any ethical need to disclose that. its just software
If you know someone would act different if something would be disclosed there is always and ethical need to disclose it.
So if I was black and racists were using my app that would stop if I put that prominently in the app, i should feel ethically compelled to disclose that?
> Companies are totally blind to the fact that A.I. is a complete turn-off to many customers.
In a normal office environment I would guess most people actually want more AI features.
Most people aren’t passionate about hand writing their best TPS reports with carefully crafted prose. They want to get their work done as easily as possible at their job so they can go home on time.
> I would guess most people actually want more AI features.
Not if they spend an unreasonable amount of time fighting the AI features. The people I know who are forced to use AI for their work spend a disproportional amount of the time fighting it.
Experience has shown that bad prose, grammar and orthograph has been accepted in business setting for the last 3 decades, basically since people stopped writing cards/letters and started communicating with sms and later instant messaging apps. Writing skills globally went downhill. The thing is: virtually nobody cares anymore.
What I see is rather management forcing employees to use AI, menacing them of bad yearly review if metric show they don't use them.
Why bother with people then, if the AI is going to do their work.
The TPS reports are written with the sender's AI and then read and summarized with the recipient's AI. At this point it would make more sense to just write the TPS reports in MCP.
Conversely, a no-AI ethos is one that will very likely need to be walked back at some point in the future for tech products. AI isn't going anywhere, so finding ways to use it constructively with your product is going to be a far better long term strategy. For now while it's new and scary and there's a lack of regulation and governance and ethical considerations - I can see why an anti-AI stance is popular, but it's not durable long term.
I'm a big supporter of LibreOffice, Codeberg, SourceHut, etc. on what they do, tell and defend, and I'm with them.
On the other hand, I'm not anti-AI, but I don't want it to be shoved down my throat in any and every app I have. I don't want that sparkly "Enhance this slide" button, "WhatsApp AI" in the corner, automatically generated "AI summary" in Google/YouTube, etc.
However, I'm fine with an unobtrusive "Summary" button in my search results (Kagi), or a tab which I can use with some models (Notion), "use ML" checkbox while resizing or "enhancing" images (Pixelmator Pro), or many small touches Apple enables with on-device models which doesn't have AI written all over it.
In short, I want consent, control and dignity in my workflows. Yes, I'm very against how these LLMs are trained, built and how the datacenters they run on harm the environment we live in. However, I'm not against the tech. I want responsible development of it, for the customer instead of their owners which wants to milk their users like content and cash cows.
I really appreciate this comment since it very effectively communicates how I feel about the subject, since I have been struggling with putting it into such clear language.
Thank you very much.
The internet has been around for quite a while now, and is absolutely critical for a lot of applications. Most people would never give up the internet.
Yet there are still many applications without explicit mandatory internet tie-ins like cloud sync. In fact, "stores data locally" and "won't disappear with zero notice" are valuable features to many people!
Just because AI is a no-brainer must-have for some apps doesn't mean it is a required feature for all apps - let alone one shoved in your face every few minutes.
I feel like the equivalent for the rise of the internet would be an explicitly anti-internet stance -- no website, no downloads, no online customer support, everything routed through snail mail, fax, and telephone (but no VoIP!) as an intentional ideological point.
> valuable features to many people!
Which products that sell this are broadly popular? I can see some people on HN wanting it, but people I know who are outside that bubble don't even think of it as an issue.
> finding ways to use it constructively with your product is going to be a far better long term strategy
It's worth pointing out that the original Document Foundation blog post (https://blog.documentfoundation.org/blog/2026/09/03/yes-no-a...) seems to be doing exactly this:
> The criteria we have listed do not amount to a definitive rejection. The AI sector is evolving rapidly; on-premises inference is becoming manageable on standard hardware; and open models are improving much faster than most of us had anticipated. Should an approach emerge that meets all the conditions, we will evaluate it very carefully.
That list seems pretty easy to meet today. I’m not sure why they’re waiting for an approach to emerge instead of building it now, unless there’s something else they haven’t listed.
What features do you think meet the list right now?
I'm not sure where the supposed difficulty lies.
"User-controlled execution." Seems easy enough to add settings so this is configurable.
"No content may leave the computer without authorisation." OK, so don't do that.
"No telemetry of any kind." There are plentiful API providers with zero data retention. If this means no data of any kind, even just records of having accessed the API, then local inference is a thing.
"No dependence on a single provider." Trivial. The OpenAI API is a de facto standard implemented by tons of providers. If you really want to go broad, it's easy to implement clients for other APIs too.
"No compromises on format." I don't even understand how this would be a problem in the first place. Text goes in, text comes out. The client can put it into whatever format it wants.
"Entirely optional." Trivial to add an off switch. Or an on switch, so it's off by default. If it must be completely removable and not just turned off, it can be implemented as an extension.
I don't think there's enough information accurately assess this. Anti-AI movements might grow much larger than today, given AIs potential to replace jobs faster than we have the ability to create them. The movement could even result in AI going somewhere.
> is one that will very likely need to be walked back at some point in the future for tech products
No doubt, but that future is when companies have figured out proper fit for AI in their product context and preferably with local AI models.
It should not be with the current slap AI anywhere and everywhere approach that uploads and trains on all available user data, which is what everyone is opposed to.
Nobody should be "finding ways to use" any technology. We should be solving problems first and foremost.
I hate products that want me to find a use-case for their feature that they wanted to develop. It's great that some people love AI features, but for me, they are rarely solving an actual problem I have. But now I have to relearn the software because the company wants to go all-in on AI? Why should I? One example of that is Notion. I've loved it for years, and I'm still using it, but it's starting to lean heavier and heavier into an AI-first workflow and I don't like it to the point where I consider switching to something else regularly.
This is awesome! I have used LibreOffice (and Open Office) extensively for decades now. It is my real office suite I use for real work and real business accounting tasks. Yes, I am very well aware that Calc is different than Excel but it can get nearly all the same tasks done as long as you are not tied deep into a corporate environment passing Excel-specific files around. I have also made great business uses of it in Docker too, generating perfect PDFS from templates, and normalizing document formats for analysis. I donate a modest $19/month personally to the foundation, much rather pay them than Microsoft!
> normalizing document formats for analysis
What do you mean here? Normalizing it into ODF format? Does that limit which tools/libraries you can use to analyze it?
Normalize into flat xml that I could then parse in Nokogiri.
Edit for clarity: I am a Rubyist at heart so I reach for Ruby for almost all of my automation, ML, and even now AI stuff. Fro a very long time, I have been able to use CLI LibreOffice to do amazing things. It can read everything, even obscure formats by today's standards, going back into the 1990s. And then once in OpenOffice XML you have one target to parse in your scripting language of choice to do document analysis, data loss prevention detection, and more. All with a clean open source license with no server usage restrictions, unlike MS Office.
Thanks, that makes a lot of sense. I wonder if that's how Codex uses LibreOffice.
Could also be due to the fact that every codex install includes a copy of libreoffice
Bundling in codex isn’t going to increase the download count on their website. Unless you’re assuming that every codex install downloads a fresh non-cached copy.
GH copilot installed it for me when trying to read some docs lol
what? why?
https://simonwillison.net/2026/Sep/1/codex-libreoffice/
To better parse office file formats
They have absolutely zero regard for system resources. I understand why, but such things were once considered unfeasible - and very recently were also considered undesirable.
I still use a laptop with a 100 GiB hard drive - only 80 of which are allocated to the /home/ directory. This would cut into the useable space significantly on that machine. How about Chromebooks?
The article says LibreOffice is almost 20 years old. Well, before it was LibreOffice it was OpenOffice in 2000 and before it was OpenOffice it was StarOffice and StarWriter originates in 1985. So you could say it (or the Writer component) is 41 years old.
Let's use AI to test the validity of an AI generated clickbait article.
Fable 5.1 was presented with the following prompt to do research.
> https://manualdousuario.net/en/libreoffice-download-record-n... states an interesting causality. Generate an artifact that tests this claim. Is it causal or coincidental. ie if libreoffice downloads have been increasing every release then it's possible to claim any recent feature change is responsible for the download. You need to get access to the download stats and then see what other major features could be responsible
https://funny-taffy-f8d6c0.netlify.app/
Conclusion is that the claim is speculative.
(This post was meat written. The report is AI generated content. )
It's also used as a headless office document editor by openai models so... does that impact their metrics?
I wonder how much of this is also driven by Microsofts current downward trajectory, I and others have switched to linux this year.
Can you control it agentically? If so it’s has all the ai features I would want. I typically don’t care for built in ai features.
This. I don’t want it to have AI features, but often find myself asking some other AI, “how do I change the paper size from A4 to US Letter,” etc.
Presumably. The headless CLI part of libreoffice is even included in Codex and Claude Code.
It’s also been baked into Codex. Go figure
It's open source software - it can be baked into pretty much anything given they follow the license terms.
The question is whether the 10 million active users of Codex / ChatGPT Work count towards the download count.
That would contradict the cause implied in the title here.
Explains the download records as well
Not really. One entity embedding the software into another will only account for at most a handful of downloads.
Unless Codex is vibe coded and it downloads it directly on every installation…
Didn't people find it bundled with a bunch of other things in a single runtime? Downloaded as part of a whole runtime would suggest it wasn't downloaded directly from the source.
i think that's a perfectly coherent and reasonable way for both these apps to work.
people don't want AI features leaking into their non-ai apps. I'm not an AI hater - I want my AI app to be able to read and write word docs, but when i decide to open up libreoffice i'm doing that because i want to write something myself, not use AI. i don't need my word processor to bundle an inferior version of what the ChatGPT can do for me.
I think it's not baked into codex, but codex has either been instructed to download it when asked to edit word docs/excel spreadsheets, or it picked it up during training.
from the thread last week, it seems to be installed by default on every mac install, possibly not on windows installs.
https://news.ycombinator.com/item?id=49528079
Nah, it is probably due to AI agents' love for Libre Office. Codex prefers Libre Office for docs. Other agents may do this as well.
People who have never known about Libre Office are now downloading and using Libre Office (often without realizing it).
https://news.ycombinator.com/item?id=49527396
I've been using Libreoffice to store my diary for years -- IIRC they use blowfish on the crypto side of things? I had a weird issue a while back where I'm not sure if I forgot my diary's password, or the file got corrupted.
I've had that happen a few times, each with a long space in between I keep things on paper.
It's a great project that doesn't get enough love.
Why Libreoffice? Does your diary contain advanced formatting, tables, images, or other involved content? Or just for the encryption?
I ask because I also keep a diary and have refined how I store it and secure it, over 25+ years. Only recently did I find something that I like better than paper.
Been using libreoffice for years.
Here's hoping increased usage will lead to improvements in the "table" functionality. Like when you can define a table, and hit tab at the last column will put you at the next row first column. Missed that from MS office.
I do love LibreOffice, but who thought this colouring is a good idea? https://i.imgur.com/cUzW97S.png
If anyone else wants to be rid of it (courtesy of astra): Set NotebookbarColorTransparency to 255. This removes the colored tint while keeping the ribbon.
Go to Extras → Optionen → LibreOffice → Erweitert → Experteneinstellungen. Search for NotebookbarColorTransparency. Double-click the matching row under org.openoffice.Office.Common → Misc. Change 150 to 255. Confirm with OK, close the settings, and restart LibreOffice.
Both Codex and Claude (on the web) use LibreOffice to edit docs, so not sure you can really conclude it broke download records because end users didn't want AI in Office and went with LibreOffice instead
Suggested titles (same logic):
LibreOffice breaks download records after GPT-6 Astra is available to the public.
LibreOffice breaks download records after DaVinci Resolve 21.1 is released.
The only thing I'd need and want from my office suite for AI are for it to stay in sync with a disk file.
Then I can do stuff to the XLSX with Claude, and work on the spreadsheet in my spreadsheet editor.
What I DONT want is for my office suite to have any limiting AI features built in.
No AI features is the first feature in years that shipped on time and worked on the first try.
I don't buy the AI argument. I'd bet on Linux replacing Windows for a good chunk of desktop users. For example, I installed LibreOffice on my parents' PC last month, because I replaced their Windows with Linux Mint.
UNO is for programmatic access. StarOffice ftw.
https://en.wikipedia.org/wiki/Universal_Network_Objects
But they support AI features, like LanguageTool. I always bought Softmaker but since they don't offer this feature in their One-time purchase version, I have started using LibreOffice only.
Maybe it's just a secret cabal of rogue agents downloading LibreOffice for no particualr reason whatsoever? I mean there is a precedent already: those fellas hijacked some poor wiki.
Why is this website asking me to enable notifications?!?
Kind of telling that "We don't have AI" is now a feature. Maybe people just want software that works.
Kind of telling that "We don't have AI" is now a feature. Maybe people just want software that works.
I broke a weight record after decimating a Mars bar. Weight's been about the same for a year.
What is the role of tools like libreoffice in the age of ai ? Is it only viewer ?
I was fighting alignment on the doc created manually, after 3 mins struggle I typed in Gemini tab of Google Docs , and it did everything perfectly.
Same with presentations , tools like bento.page generates you perfect html .
I really do not want to go and edit things manually anymore
today I broke my personal age record – it was a tough ride, but with grit and determination, I think maybe, just maybe, I can do it again tomorrow.
Oh, and I also have no AI features.
Hate to say it but it is because of AI, not in spite of it. Codex recently started including the entire libreoffice package for MS office compatibility. https://simonwillison.net/2026/Sep/1/codex-libreoffice/
I came here to say that this is extraordinarily uninteresting. But some useful thought came out of it: all "AI" features are moving into the generalized agent harness anyways, and good software often maintains a focus. Take Windows, for example. It has a rather broad job being an OS, but as a product it's a good example of loosening focus. it's losing consistency in its opinions about where updates should happen, how settings should apply, which menus are in charge, and even what the design philosophy is. And that bleeds out into the user surface too. Windows 11 is a scattered mess that's increasingly just a surface for finding new ways to pull your attention into ads.
Maybe software just needs to remain straightforward and focused, and we let the ChatGPT and Claude apps handle the AI part of things.
Do nothing and win
Isn’t OpenOffice called by llms to do tasks?
This is the way!
Hand rolled code is worth more.
I’m using Pages on my macbook neo and it seems it doesn’t really have ai features, and if it does, they are very well hidden
Are you sure it wasn't Microsoft doing a rug-pull October 13 on offline Office 2021 pro lifetime license installations for MacOS.
A lot of people are not happy about that stunt. =3
Or people outside the US distrusting US technology? There is a confluence of trends pointing away from US based tech. It will take a crazy amount of work to restore trust.
Another reason why it might be hitting records is many companies getting rid of Microsoft 365 downloads for desktop and only allowing use of the web version. And apparently it's the first thing that comes up when you try to download Microsoft 365? People don't try to pirate it anymore, they just download LibreOffice
false claim of causality (post hoc, ergo propter hoc).
Clickbait title…
I switched to LibreOffice just recently, but only because i’d been happy with OpenOffice for years and only recently decided to see if there was an update and found out it had been discontinued in 2011.
It was still working fine for me tbh, but i thought better to update to something newer.
If only it were not dog slow in performance metrics. Still not as slow as Open Office but man it's close.
And that's on a newer Dell Pro Max with an Ultra 7 processor, 96GB RAM, and dual GPUs (Arc and RTX Blackwell.)
Basic tools don't need AI features, but facilitating AI access from external agents is extremely useful. We used to call these things APIs.
The last time I tried to use LibreOffice, I made three hours worth of comments on a Word document.
As soon as I saved the document, they all vanished.
Then the developers mocked me for complaining.
Now I just use real MS Office in a snapshotted virtual machine, because it works every time and loads just as fast.
LibreOffice is downloaded by several LLM harnesses like codex so...
Offtopic blogging tip: Americans will read dates like "8/9/2026" as "August 9, 2026." I'm pretty sure the author is using the (more sensible) convention of day/month/year but expressing dates as e.g. "08-Sep-2026" removes the ambiguity.
Highly disagree that either is more sensible than the other, especially given how each culture actually uses dates in their speech.
Or we could follow Japan and just use ISO 8601 directly and write 2026-08-09.
https://gcdnb.pbrd.co/images/BzKdTMyDkDyU.png
I use this for everything except where I write the month out or use an abbreviation. Much easier.
The only real sensible date format is yyyy-mm-dd, makes it easy to sort.
Agreed. You can leave out the hyphens to make sorting even easier with almost no loss of readability.
Related:
Yes, no (built-in) AI is now a feature
https://news.ycombinator.com/item?id=49553350
So related it's in the blog post.
I mean not having voice transcription in a document editor is kind of dumb
But not having some shitty clippy-like help system is very nice
this is temporary marketing towards a direction that they may change their mind on in the future, which would likely anger people who downloaded it because of this preference
not a call I'd make early on, a preference to disable AI features like Firefox added is maybe a more reasonable call
all we know today is that AI is being used in annoying ways, this won't be forever
It has no AI built in, which is a feature to a growing set of users. No AI buttons all over the place and sending your personal data off to who knows where. And if you want AI, you can add your choice of AI via plugins if you want AI in your word processor.
Sure, but local LLMs are getting smaller all the time... so it could in the not so distant future ship with its own locally-hosted LLM embedded (no data leaving your device).
I guess it could always be relegated to plugins, but I suspect over the next 5 years the line is going to get a lot blurrier.
Strange, when office suites are the one software category most people are anticipating better AI integration for unlike everything else.
Nobody wants AI in browsers, but an AI that can make me the Powerpoint I want without me having to drag cumbersome items by hand? Yes please.
Awesome. Now can we fix the bugs and make it more usable? I’m longing for a Microsoft/Google alternative but they are all still clunky as hell and lack the integration of Outlook. Im currently using Planner as an excel replacement and it is painful but less painful than LibreOffice.
Hopefully LibreOffice is one of the projects that gets free Pro Max accounts from A and O.
Corporations have been trying Star/Open/LibreOffice migrations for literally 30 years at this point. And, at the pace it’s been going, it will probably take another 30 for it to reach Word 95 polish.
Tell me you haven't used LibreOffice without telling me you haven't used LibreOffice.
It's actually a pretty great software suite.
enough with this anti ai noise resistance is futile it's getting popular due to improved ui
You’re wrong. Despite its usefulness in our profession, AI is deeply unpopular with the general public and products that don’t include it are rapidly gaining in popularity. People are tired of AI being shoved down their throats without consent.
Are you speaking from personal experience or do you have data to back that up? I honestly would love to see some study around this. I haven't done my research but --anecdotally-- everyone around me is ok with it or even enjoying AI to some degree. No strong stances either way tbh.
Here you go: https://archive.ph/9erfx bask in that 20% support of "more good than bad when it comes to news" and the 70% thinking it's destroying jobs.
There's also this, quite related one, showing that the BigAI CEOs are very disliked by below 34's https://futurism.com/artificial-intelligence/young-people-ai...
I'm not anti-AI (I have a ChatGPT subscription), I just don't like bloat and nagging.
Enough with this pro-ai noise.
Nonsense. I can (and do) resist a lot of things. Society in general may well be screwed, but I'll at least hold onto a bit of self-respect.
Regardless of the fact that the 'record download' levels is probably due to some other factor, the world is already becoming conscious that the so-called AI is not all it's cracked up to be.
Very annoying that AI is this new virtue signaling topic. I guess it's a step up from a few years ago when it was about naming branches, but it's annoying.
Libreoffice is free software that's the big point of it. You can mod it and stip what's you dislike. I'd be glad if people developed AI integration with it. One could do that with AI itself. I've successfully modded free software with AI in many ways, adding features to Blender, for example.
This binary pro of pro and anti AI buzz is just annoying. Privacy is important. But AI is treated by some people as if it was some nuclear waste. In fact it's very similar to the anti nuclear green movement. It's an emotional aesthetic response.
The actual story below the headline is that AI is allowed in LibreOffice via plugins, they just aren't going to do it themselves since it violates some of their principles about privacy and local data storage.
They aren't against AI, they just have certain promises made around the product that would be hard not to break if they added first party AI.
Not a fan of pushing LLMs into everything, but this is one of those things where it would be great to have a powerful(or specialized) local llm to help you
Write a formula, make a macros, filter stuff, formatting. I'd rather not to Google it. Especially considering SEO made it impossible to find information how to do stuff in editor X of version Y, at best you'd get thrown with crap for other version and have to figure out how to adapt it for your version
You can still ask a general model to do that and it will probably have a nearly full success rate at this point. Either a good open weights or something like Astra.
Yeah, but it would in many cases remove you as a meat proxy when your relative or a friend or your boss ask you how to make X when they have a box where they can write to a genie "I want X" and it would put everything in place for them
I don't see not having AI as a feature, but I'd prefer for them to take the approach of optional, easy to implement interfaces that increase the ability for the suite to be 'scripted' rather than implement AI directly. My motivation is that the field of AI is moving so fast I think its a better long term plan to make your software something that can be automated with AI than to try to fight the speed of that innovation curve. I get that even that can be difficult because of security concerns, but it seems more straightforward than implementing models directly into your application.
That is, in fact, exactly what they explain as their approach in the article, as well as their reasoning behind it.
It isn't that they disallowed AI, its that they just leave it up to plugins and have no native interface since it violates many of their principles around privacy and local data.
I absolutely see a lack of soulless clankers to be a feature.
I completely agree that focusing on "scripting" would be great. Clanker proxies can allow their clanker to use it while grumpy bastards like me can stick to Bash to automate it.