bert hubert 🇺🇦🇪🇺🇺🇦

Software developer, entrepreneur, former government regulator, current government advisor, amateur scientist.  https://berthub.eu/

Website: https://berthub.eu/
Orcid: https://orcid.org/0000-0001-6317-0670
Linkedin: https://www.linkedin.com/in/bert-hubert-b05452/

Public Key

npub1579yezsv64rkghcezalde636kasdazzpnehtl99wt4g9d62ac0aqu5sp9g

NIP-05 Address

[email protected]

Profile Code

nprofile1qqs20zjv3gxd23mytuv3wlkuagatwcx73pqeum4ljjh965zka9wu87spz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0ds9arp9e

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1579yezsv64rkghcezalde636kasdazzpnehtl99wt4g9d62ac0aqu5sp9g

Show more details

Last Notes

2025-05-14 10:21:37 CEST

- reply

Here is what the Galileo agency mailed round last week. Not easy to understand what they mean, and I guess this is by design:
https://cdn.fosstodon.org/media_attachments/files/114/505/211/114/152/336/original/a0537e05cef3a373.png

2025-05-14 09:21:32 CEST

The bot is making a lot of noise about Galileo, the EU GPS. This is because Galileo is being very weird and is clearly running on outdated orbital data. It might still be accurate enough though. This was announced very cryptically last week that there would be “deviations”. Galileo communications remains absolutely terrible.
From: [@gnss_changes](https://bot.country/@gnss_changes )
https://bot.country/@gnss_changes/statuses/01JV6RDHRHXPQ88WTNXJP4AXFM

2025-05-13 10:42:06 CEST

As I type away into the void, hoping to achieve something, I sometimes wonder if it is worth it. Do people read this stuff? There are now 222 articles on https://berthub.eu/articles/ and sometimes one of these gets read really well. But what about the "back catalog"? Turns out that a lot of traffic is in fact coming from the 200 older pieces:
https://cdn.fosstodon.org/media_attachments/files/114/499/615/087/350/437/original/ef2a27f9c669685a.png

2025-05-11 22:06:34 CEST

I too have an important announcement tomorrow! It is good news, so no worries.
https://cdn.fosstodon.org/media_attachments/files/114/490/995/238/878/605/original/045e694111b847b5.png

2025-05-10 09:49:09 CEST

- reply

hele matige journalistiek dit. Wat is het nou.

2025-05-08 14:57:17 CEST

- reply

I can recommend reading some of the articles I linked to. Most users are now relying on higher level services than you describe. “We” can’t just change that.

2025-05-08 13:09:23 CEST

I've written an unhealthy amount of words on "the cloud" and specifically Europe's woes. In the post below I tie many articles together into a hopefully useful overview. It may be good to know that nothing what I write on the cloud is novel or original, I mostly hope to report things as they are. Which is scary enough already! https://berthub.eu/draft-klad-tijdelijk-draft/cloud-overview.html

2025-05-07 20:32:47 CEST

- reply

the bandwidth cost often is only a tiny fraction of the rework cost to move to something else. Unless you are storing petabytes of course. But even 100s of thousands of euros/dollars is not a lot compared to the manual labour required..

2025-05-07 20:07:24 CEST

- reply

I have an upcoming post on this too. I've radicalized on the topic of the techno-illiteracy. But I'm waiting for the Dutch cabinet to fall before I launch my rant on this ;-) For now I can only function on the outside by writing very plain blog posts...

2025-05-07 09:34:20 CEST

Everyone agrees something needs to be done about our total dependence on US clouds. But what? And who should do it? Ample words have been written on the 'why', and we also have a lot of text on our values and that *someone* should do *something*. Here I elaborate on an earlier suggestion with a coherent strategy that is concrete enough to disagree with. This reads well with the latest Euro-Stack letter also: https://berthub.eu/articles/posts/a-coherent-non-us-cloud-strategy/

2025-05-06 22:18:15 CEST

- reply

so views were started because the number of processes you could run was limited. It is sort of a virtual machine in your nameserver. It makes everything pretty complicated. You might also get surprises if you mess it up even slightly. I'm not much of a fan.

2025-05-06 21:21:19 CEST

- reply

I’m mostly out of dns. So while I might have ideas they aren’t current. Sorry!

2025-05-06 16:45:36 CEST

- reply

indeed.

2025-05-05 14:47:34 CEST

- reply

It really is worth it to read the whole 1939 article. https://www.lehman.edu/faculty/rcarey/BRSQ/06may.orwell.htm

2025-05-05 11:57:15 CEST

"Underlying this is the idea that common sense always wins in the end. And yet the peculiar horror of the present moment is that we cannot be sure that this is so. It is quite possible that we are descending into an age in which two and two will make five when the Leader says so." - George Orwell in https://www.lehman.edu/faculty/rcarey/BRSQ/06may.orwell.htm

2025-05-03 10:28:59 CEST

- reply

I just die inside when that happens! That was the reason for writing this also.

2025-05-03 08:50:07 CEST

- reply

he’s very right. He could start his presentation like this because the audience was clearly already in awe of him. But he still needed a nuclear start to get everyone hooked!

2025-05-02 20:53:08 CEST

- reply

always works great yes ;-)

2025-05-02 20:48:41 CEST

- reply

Thanks for spreading the word! And also please know I appreciate your thoughtful and often touching posts on your life experiences. I often fondly recall similar things from my life based on what you write. Sometimes brings a tear to my eyes even.

2025-05-02 20:27:42 CEST

Quick note on giving talks. It is not enough to have great content. Expectations are low. When you start off, quickly establish some kind of bond with your audience. Once that is in place, your listeners will give your presentation a decent chance: https://berthub.eu/draft-klad-tijdelijk-draft/but-first-connect-with-your-audience.html

2025-05-01 21:33:59 CEST

- reply

in my world this is known as the screwdriver theory, that no computer is secure in the face of someone with time and a screwdriver. With very rare exceptions this is true.

2025-05-01 11:27:08 CEST

- reply

All these Microsoft assurances that it really won't be so bad with Trump remind me of this John Cleese sketch where the pilots tell the passengers that the wings are NOT on fire: https://youtu.be/XfLdFZ4my9g?t=35

2025-05-01 09:39:59 CEST

- reply

Thanks! There is a weird sentence in there "vaccine were legeoss protected". Otherwise great!

2025-05-01 09:06:39 CEST

- reply

Danke!

2025-05-01 07:43:57 CEST

«Die Cloud» ist nicht nur Server. «In die Cloud gehen» kann auch bedeuten, einen ewigen Dienstleistungsvertrag abzuschliessen, schreibt Bert Hubert.
https://dnip.ch/2025/05/01/die-cloud-der-unkuendbare-subunternehmer/

2025-05-01 07:38:47 CEST

I just read Microsoft's post on how much they love Europe & that they are not the same as Donald Trump. As far as I can tell there is nothing new in there, no new/useful protections, no guarantees. It rehashes already debunked things like the "Microsoft Lock Box". The story mostly serves as a reminder of how worried we should all be about basing our governments & societies in companies far away under questionable legal regimes: https://blogs.microsoft.com/on-the-issues/2025/04/30/european-digital-commitments/

2025-04-29 20:19:29 CEST

- reply

the next exciting thing is unfurling the scary antenna...https://www.bbc.co.uk/newsround/articles/c0jzy3g0zx2o

2025-04-29 17:37:48 CEST

- reply

last time round I wasn’t EU enough for them somehow.

2025-04-29 13:11:23 CEST

If I wanted to leave fosstodon.org, where should I go? I need a very boring server with no surprising moderators causing upheaval..

2025-04-29 12:16:38 CEST

- reply

but is this biologically plausible? Is it statistically likely that a single chromosome would end up being sequenced like this?

2025-04-29 10:59:24 CEST

In 15 minutes Europe will hopefully launch its next climate satellite. The launch can however only be watched via YouTube since we apparently can’t do that ourselves and have to put our government info next to the antivax promo. https://www.esa.int/ESA_Multimedia/ESA_Web_TV

2025-04-28 19:44:05 CEST

- reply

this is partially addressed in this article which is also linked from the post: https://berthub.eu/articles/posts/your-tech-my-tech/ - it is also true for large enterprises these days…

2025-04-28 17:21:44 CEST

- reply

and followed back!

2025-04-27 09:11:22 CEST

“Going to the cloud” can mean renting services/servers that you could get from anywhere. There’s little lock-in. The same four words “going to the cloud” might also mean locking your operations to a specific cloud provider, forever. This difference vital, yet often ignored: https://berthub.eu/articles/posts/beware-cloud-is-part-of-the-software/

2025-04-24 09:40:15 CEST

"Three of the four most exploited vulns were zero days, all were in cybersecurity products (Palo-Alto, Ivanti Connect Secure, Ivanti Policy Secure and Fortinet)." https://cyberplace.social/@GossiTheDog/114391474274944937

2025-04-23 12:26:56 CEST

With some luck, I'll be presenting tomorrow. And I'll again be talking about https://en.wikipedia.org/wiki/The_Two_Cultures "So the great edifice of modern physics goes up, and the majority of the cleverest people in the western world have about as much insight into it as their neolithic ancestors would have had"
https://cdn.fosstodon.org/media_attachments/files/114/386/791/205/456/433/original/fec14fa1265d9554.png

2025-04-23 10:14:51 CEST

- reply

Thank you everyone. The answer to 'does the Microsoft Customer Lockbox protect your data against the US government' is a resounding NO. According to Microsoft that is. https://learn.microsoft.com/en-us/purview/customer-lockbox-requests#does-customer-lockbox-protect-against-data-requests-from-law-enforcement-agencies-or-other-third-parties
https://cdn.fosstodon.org/media_attachments/files/114/386/275/480/267/391/original/ca3acda8f5793a03.png

2025-04-22 18:36:45 CEST

People, a Microsoft (re)seller is telling governments you can put your MS 365 email in a "Microsoft Lockbox", and that this will protect you form US government spying. Is there any truth to this? https://techcommunity.microsoft.com/blog/windows-itpro-blog/windows-365-customer-lockbox-generally-available/4077641 (reposts appreciated - it is somewhat time critical).

2025-04-20 22:01:41 CEST

- reply

DANK U!!

2025-04-20 12:54:04 CEST

- reply

I've found that the level of care on published genomes varies enormously. People are not always overly interested in hearing they/their former postdoc etc messed it up. Might even ruin a whole publication. Or career.

2025-04-20 09:45:23 CEST

"Learn the difference between 'helping' and being 'helpy'"
~ @wendynather at #BSidesSEA - this is very useful. Just pointing out that an article does not make a specific point you'd like to see is "helpy". Pointing out that many people already knew what an article states also does little for the world. Being actually helpful looks different & is more work. 1/2

2025-04-16 16:38:28 CEST

- reply

does it do AI suggestions already? Probably next.

2025-04-16 11:42:42 CEST

- reply

I haven't tried though!

2025-04-15 20:07:45 CEST

- reply

or cowards.

2025-04-15 19:59:53 CEST

- reply

also not true. https://berthub.eu/articles/posts/dear-hosting-providers-you-sell-lumber-and-not-furniture/
https://cdn.fosstodon.org/media_attachments/files/114/343/279/043/777/428/original/3661eaa81fc0cea8.jpeg

2025-04-15 18:49:21 CEST

- reply

"Are you going to the cloud? Or is the cloud going into you?" - the difference between renting compute/standardized services, OR, building your future on bundles of powerful but proprietary services, effectively making a specific cloud your eternal subcontractor.

2025-04-15 08:48:44 CEST

- reply

bit too tricky I think. But good inspiration!

2025-04-14 23:13:38 CEST

A fascinating read: https://www.theguardian.com/news/2025/apr/10/deep-cover-kgb-spy-recruited-son-peter-herrmann-illegals

2025-04-13 09:34:50 CEST

Regarding the previous post https://fosstodon.org/@bert_hubert/114326762794571784 I find it interesting that apparently there really is no distinctive name for "going deeper into the cloud than only using commodity services". PAAS has been mentioned but the various clouds use that term inconsistently.

2025-04-12 23:40:57 CEST

- reply

not so sure if this is a widely used definition. I've been reading what Google and Microsoft describe as PaaS it is not so clear cut. PaaS is described as being some kind of development environment. https://cloud.google.com/learn/paas-vs-iaas-vs-saas https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-paas

2025-04-12 21:59:17 CEST

"Going to the cloud' can mean anything from renting a VM, deploying containers, or building on top of advanced (hyperscaler) cloud services (queues, buckets, customer IAM etc). Is there a good term to describe the latter development style, going all in on cloud services? I've used 'cloud native' before but the meaning of that is not clear or useful. Better suggestions very welcome!

2025-04-12 18:20:30 CEST

- reply

have you actually tried this since these options do not appear on iPhone signal.

2025-04-12 16:41:38 CEST

- reply

same. And I don’t know what to do about this. Seems this is by design. I have to first uninstall all kinds of other things now and then update.

2025-04-06 17:19:34 CEST

From an upcoming blog post, in which I argue that cloud providers are actually more like proprietary software providers: "These days under the guise of ‘cloud as a commodity’ we are gluing ourselves to specific suppliers, without realizing that they are not in fact supplying commodity services."
https://cdn.fosstodon.org/media_attachments/files/114/291/683/857/796/128/original/8d6fcc7a4354a6a7.png

2025-04-02 11:35:11 CEST

With great fanfare the #NeoNephos foundation was announced just now over at #Kubecon. It is not that easy to figure out that this means, but it appears to be about Kubernetes-like infrastructure, which I rank as 'step 3' on the cloud scale, where AWS/Google/Azure are at 5.
https://neonephos.org/2025/03/31/the-linux-foundation-announces-the-launch-of-neonephos-to-advance-digital-autonomy-in-europe/ - and some projects are here https://neonephos.org/projects/ - further reading on the kinds of clouds: https://berthub.eu/articles/posts/the-european-cloud-ladder/

2025-03-15 17:50:10 CET

- reply

maybe not the best example (for me), but perhaps I know too much. I'm not putting that much stock in security barriers I can't see. The history is not that great. https://www.techradar.com/pro/amazon-pauses-usd1bn-microsoft-365-rollout-following-russian-security-concerns

2025-03-15 17:33:36 CET

- reply

well, I did not mean to endorse AWS etc, but I am pretty fed up with linguistic tricks used by at best "container" providers that they are in the same business as AWS/GCP etc.

2025-03-15 17:16:07 CET

No matter how much you want it, you can't use a clever definition of "cloud native" to pretend that you compete with the AWS/Azure/Google stack. And please don't try to fool people with a wonky definition, it will backfire eventually. "There is no cloud just other people's computers" means you don't understand what modern developers are doing with clouds. https://berthub.eu/articles/posts/the-european-cloud-ladder/

2025-03-09 13:15:42 CET

Back when COVID-19 hit, savvy observers were aware this was going to be Bad, even as we were surrounded by people who assumed this would not be a big deal. I vividly remember doing last minute shopping with folks thronging around bowls of free food. I knew then not to get anywhere near that. I get the same feeling now. We cling to normalcy but the world is heading to a crazy situation at a rapid clip. Yet our thinking is not yet there. "Surely this will pass?"

2025-03-09 11:26:20 CET

For the people speculating about "jailbreaking" the F35 so you could fly it without US support. The thing is so maintenance heavy it won't fly after a week if you stop supplying it proprietary parts and loving US vendor help. The software is only a tiny thing relatively speaking. This goes for a lot of high-end military kit.

2025-03-06 19:50:52 CET

So if you wanted to design a setup to lure as much (government) information as possible to US controlled servers, you'd get something that looks remarkably like MS 365 and Windows 11.

2025-03-06 12:56:48 CET

- reply

these machines need constant love and care to function. They don’t need a kill switch. However, there are lots of US components in the machines as well which limits the strategic options.

2025-03-03 11:03:17 CET

Previous post inspired by failing attempts to get people in charge of important things to understand anything about the important things they are in charge of.

2025-03-03 10:54:27 CET

A reminder, no matter how hard you try, there are things you can't *meaningfully* explain to people who do not (yet) have the background to understand it. There is only so much gap you can bridge in a single attempt. Some things truly require education and a lot of time. Here Feynman (in pain I think) explains the limits of explanation: https://www.youtube.com/watch?v=MO0r930Sn_8

2025-03-01 23:50:09 CET

- reply

- back up your computer, we're going to need it!
https://cdn.fosstodon.org/media_attachments/files/114/089/614/769/188/877/original/1e8f52707c7a7784.png

2025-02-26 11:14:28 CET

Regarding previous post, I have more sharing links available if you need them.

2025-02-23 20:50:32 CET

If you send me a Microsoft Teams link to talk about your open source policies, I'll respond with a Signal meeting link. And if you find it hard to install additional software to meet with me, I feel exactly the same way.

2025-02-23 12:53:22 CET

It is very hard to accept, but it is no longer safe to move EU governments & societies to US clouds. Not only is it dangerous to do so, it is also likely flat out illegal in the near feature. We're trading convenience for utter dependence on a mad king. It should stop.
https://berthub.eu/articles/posts/you-can-no-longer-base-your-government-and-society-on-us-clouds/

2025-02-23 10:41:49 CET

- reply

via hoger onderwijs persbureau in veel universiteitsbladen: https://www.cursor.tue.nl/achtergrond/2025/februari/week-3/waarom-mag-chatgpt-reclame-maken-bij-surf/

2025-02-22 22:35:02 CET

- reply

yes. The code out there has rarely or never seen really strange data. I do know there are heuristics in there that start with GPS. Ublox fundamentally changes its behaviour also if you tell it not to use GPS.

2025-02-22 21:40:45 CET

- reply

I've tried in many places to raise the issue of 'fuzzing' GNSS receivers how they'd react to truly odd data coming from one constellation. Most places did not quite understand the question. Many seemed to assume this would not happen/not be a problem. Only one (important) place assured me they had looked into it.

2025-02-19 21:37:24 CET

The legality (not wisdom) of putting European private data on US clouds hinges on the availability of the US Privacy and Civil Liberties Oversight Board. Trump neutered this board, and the European parliament has taken notice & asked the European Commission what they think:
https://cdn.fosstodon.org/media_attachments/files/114/032/467/413/708/900/original/f36b37e474c43380.png

2025-02-09 16:13:05 CET

- reply

perhaps reconsider that you need to be sure your outgoing email is actually OK before you can complain rightfully that it ends up in the spam folder.

2025-02-09 14:46:44 CET

I'm no fan of the email oligopoly of Google, Microsoft and Apple. But if you care about how Gmail feels about your email, I highly recommend enabling Google Postmaster Tools for your domain. This tells you how often people flag your email as spam, if your encryption is enabled, if your DKIM/DMARC is working & more.
https://cdn.fosstodon.org/media_attachments/files/113/974/229/717/399/463/original/e09f155592c29b71.png

2025-02-03 14:14:57 CET

- reply

I showed you an independent legal analysis. Do you also have a link to one that backs up your claims? That would be mighty interesting. And novel btw.

2025-02-03 13:44:51 CET

- reply

I recommend reading https://www.ncsc.nl/binaries/ncsc/documenten/publicaties/2022/augustus/16/cloud-act-memo/Cloud+Act+Memo+Final.pdf before continuing to opine on the security of EU data on Microsoft servers.

2025-02-02 14:38:29 CET

In Dutch, but translation works well. My call for the Dutch government to retain at least a core IT/communication/email/file capability that is independent of US clouds. Named after the iconic Radio Kootwijk which we built in response to the English cutting off our communications with Indonesia in 1916: https://berthub.eu/articles/posts/mailen-en-communiceren-zonder-musk-en-trump-cloud-kootwijk/

2025-02-01 22:25:35 CET

- reply

(looks like this is not happening for now btw, taken out of the agenda and proposal)

2025-01-30 17:10:17 CET

- reply

The SQLite FTS5 people feel the same way, and offer a mode where your index becomes *way* smaller ('contentless tables'), but it is quite the hassle. You also lose some functionality. In this case my code would've become a lot more tricker and I would have saved operators 120 euros. Or their far bigger cloud spend to get the same. So probably worth it. 2/2

2025-01-30 17:07:08 CET

Today I doubled the memory in a server to 64GB and everything is so much better now. This cost me 120 euros. This is for a full text index on >150GB of documents, and the index is 17GB. But it still feels like cheating to use that much RAM. However, if you can solve your problem for 120 euros that is also a big win. 1/

2025-01-29 17:43:12 CET

- reply

For how this works, developers are always worried no one is yet running their latest versions. You do them a great favour by deploying and trusting their latest and greatest. It helps everyone gain confidence in the new versions. And on the flip side, it is a lot of work for a developer to try to investigate your issue in years old releases. So everything is just Way better if first reproduce on something recent.

2025-01-29 17:38:38 CET

A pro-tip from the field - if you want to report a problem to an open source project, it helps TREMENDOUSLY if you can reproduce the issue with a very recent release of their work. In other news, my parliamentary data monitor https://opentk.nl/ is now running on a two week old SQLite version.

2025-01-29 09:32:28 CET

- reply

.. Hopefully this will then prompt actual projects in many of the niches we mention, but EuroStack is not a development project in itself. It's an opinion campaign and everyone is welcome to join." https://www.linkedin.com/feed/update/urn:li:activity:7270923211142262784?commentUrn=urn%3Ali%3Acomment%3A%28activity%3A7270923211142262784%2C7289300319555993602%29&replyUrn=urn%3Ali%3Acomment%3A%28activity%3A7270923211142262784%2C7289631793245265922%29&dashCommentUrn=urn%3Ali%3Afsd_comment%3A%287289300319555993602%2Curn%3Ali%3Aactivity%3A7270923211142262784%29&dashReplyUrn=urn%3Ali%3Afsd_comment%3A%287289631793245265922%2Curn%3Ali%3Aactivity%3A7270923211142262784%29

2025-01-28 09:00:14 CET

Always good to go to the source - even on modest hardware you can use Ollama to run different versions of DeepSeek-R1. It sure is long winded in its thinking! This is the 70b version:
https://cdn.fosstodon.org/media_attachments/files/113/904/920/928/636/308/original/b208a40fc4dce82f.png

2025-01-27 20:35:57 CET

- reply

We might have hoped for more immediate breakthroughs, and people were surely expecting those. These days we are working on 'pangenomes' which incorporate all human DNA variation, and these may in fact hold more 'smoking gun' results than the "universal genome" created by the HGP. But because we would not have gotten where we are today without first doing "a" HGP, it has by definition been a success. /end

2025-01-27 20:33:29 CET

- reply

yes hello, I do know a bit. The human genome project can now be reproduced in a single week. We now have a depth and breadth of DNA knowledge of all kinds of human beings the original project never dreamed of. But to get to where we are today the HGP had to happen first. One of the main discoveries may have been that our genome is far more nuanced and complicated than people thought. There is no chance we would not have done the project. It was inevitable. 1/

2025-01-27 09:22:17 CET

Interestingly, one person captured by big tech found that Microsoft was perfectly in their rights to POST to you "since we could have no proof a user wasn't behind initiating that POST". Another reader wanted Microsoft to be prosecuted for hacking. Others chimed in how this explained their subscribe/unsubscribe links being weird lately: https://berthub.eu/articles/posts/shifting-cyber-norms-microsoft-post/

2025-01-26 20:42:57 CET

"The second advantage of WAL-mode is that writers do not block readers and readers to do not block writers. This is mostly true" - coming to you live from the part where this is not true. #sqlite3 https://sqlite.org/wal.html

2025-01-23 16:56:33 CET

"So how long have you been into cyber" - apparently since October 1989. I must have been insufferable if I read this correctly! Using the PowerC compiler no less! http://www.mixsoftware.com/product/powerc.htm
https://cdn.fosstodon.org/media_attachments/files/113/878/483/024/662/490/original/59128f981f5aa6e1.jpg

2025-01-23 13:05:01 CET

- reply

I wonder what your thoughts are on the above - Microsoft running JavaScript on links found in email & sending out POSTs, and where this might be heading.

2025-01-23 12:50:33 CET

Yesterday a user told me they couldn't log in to my parliamentary monitoring site. On investigating, I found that Microsoft email security was logging in on behalf of my user by executing a POST. This broke the single-use sign-on link. Executing POSTs is usually considered unacceptable, and in this way Microsoft again transgresses an important norm. Here's how to deal with the specific POST problem & what might be done about these transgressions in general:
https://berthub.eu/articles/posts/shifting-cyber-norms-microsoft-post/

2025-01-23 00:53:32 CET

- reply

a neat trick is to use Math.random() and get it to POST the result. The Google crawler at least for ages always returned the same random number. Could also try the time perhaps.

2025-01-22 22:59:39 CET

The GPS augmented landing system in Europe appears to be degraded tonight. Only the parts in red are working well if I understand it correctly. Second image is what this normally looks like. https://egnos.gsc-europa.eu/aviation-portal/aviation-dashboard
https://cdn.fosstodon.org/media_attachments/files/113/874/238/641/140/834/original/c7c6a448fecae2d0.png
https://cdn.fosstodon.org/media_attachments/files/113/874/246/860/504/711/original/9b4bd23afee5b4e7.png

2025-01-22 20:34:11 CET

So I send out login links in emails. If the user clicks on the link, JavaScript in that webpage will *POST* to a URL to log in. The login token works only *once*. I now have a user that tells me they can't log in. From the logs, I see a Microsoft IP address POSTing that login link. Is some kind of MS email security product executing JavaScript and POSTING things? Is this a known phenomenon? I've now made this a login *button* which the user has to click again. I hope it helps.

2025-01-21 22:19:23 CET

George #Orwell’s 1940 review of Mein Kampf is an astounding read. It is very educational to study primary literature from just before historical events. In here we read that up to 1939 Hitler was regarded rather favorably by many, for reasons we might recognize today. Frightening: https://bookmarks.reviews/george-orwells-1940-review-of-mein-kampf/

2025-01-20 20:36:28 CET

Wishing all my American friends strength today & for the future. Please hang in there!

2025-01-19 23:08:58 CET

So I run this parliamentary monitoring site, and most bots are welcome to trawl it. I spent a lot of (OCR) time making those documents accessible, so have at it. But what really gets my goat is if you try to anonymously crawl the whole site while pretending (badly) to be a normal internet user. As we used to say on usenet *plonk*.
https://cdn.fosstodon.org/media_attachments/files/113/857/294/548/932/062/original/42a9def3a132697d.png

2025-01-19 17:25:17 CET

Also for your mental health, I recommend never reading the source code for the various implementations of mktime() and strptime(). The authors really try their best, but on reading that, you worry how anything ever works. https://github.com/bminor/glibc/blob/master/time/mktime.c
https://cdn.fosstodon.org/media_attachments/files/113/855/943/400/422/277/original/4b2b402e88072ece.png

2025-01-19 16:15:28 CET

So how hard could it be to convert `Fri, 17 Jan 2025 06:07:07` in UTC into a UNIX epoch time_t timestamp, in C or C++? Quite hard. Many things that you'd think would work actually don't. Here I present solutions (with running code), and a little tour through 'struct tm', 'time_t', 'mktime' and 'strptime':
https://berthub.eu/articles/posts/how-to-get-a-unix-epoch-from-a-utc-date-time-string/

2025-01-18 17:07:26 CET

- reply

This is indeed good stuff. C++ will eventually get there with https://en.cppreference.com/w/cpp/utility/expected#Example