I've had the opposite experience, it's been a long time listening to people going "It's really good now" before it developed to a permutation that was actually worth the time to use it.
ChatGPT 3.5/4 (2023-2024): The chat interface was verbose and clunky and it was just... wrong... like 70+% of the time. Not worth using.
CoPilot autocomplete and Gitlab Duo and Junie (late 2024-early 2025): Wayyy too aggressive at guessing exactly what I wasn't doing and hijacked my tab complete when pre-LLM type-tetris autocomplete was just more reliable.
Copilot Edit/early Cursor (early 2025): Ok, I can sort of see uses here but god is picking the right files all the time such a pain as it really means I need to have figured out what I wanted to do in such detail already that what was even the point? Also the models at that time just quickly descended into incoherency after like three prompts, if it went off track good luck ever correcting it.
Copilot Agent mode / Cursor (late 2025): Ok, great, if the scope is narrowly scoped, and I'm either going to write the tests for it or it's refactoring existing code it could do something. Like something mechanical like the library has a migration where we need to replace the use of methods A/B/C and replace them with a different combination of X/Y/Z. great, it can do that. Or like CRUD controller #341. I mean, sure, if my boss is going to pay for it, but not life changing.
Zed Agent mode / Cursor agent mode / Claude code (early 2026): Finally something where I can like describe the architecture and requirements of a feature, let it code, review that code, give it written instructions on how to clean it up / refactor / missing tests, and iterate.
But that was like 2 years of "really it's better and revolutionary now" before it actually got there. Now maybe in some languages or problem domains, it was useful for people earlier but I can understand people who don't care about "but it works now" when they're hearing it for the sixth time.
And I mean, what one hand gives the other takes away. I have a decent amount of new work dealing with MRs from my coworkers where they just grabbed the requirements from a stakeholder, shoved it into Claude or Cursor and it passed the existing tests and it's shipped without much understanding. When they wrote them themselves, they tested it more and were more prepared to support it in production...
Which is equal parts praise and damnation. Claude Code does do a lot of nice things that people just kind of don't bother for time cost / reward when writing TUIs that they've probably only done because they're using AI heavily, but equally it has a lot of underbaked edges (like accidentally shadowing the user's shell configuration when it tries to install terminal bindings for shift-enter even though the terminal it's configuring already sends a distinct shift-enter result), and bugs (have you ever noticed it just stop, unfinished?).
> - What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music?
Starcraft and Factorio are exactly what it is not. Starcraft has a loooot of micro involved at any level beyond mid level play, despite all the "pro macros and beats gold league with mass queens" meme videos. I guess it could be like Factorio if you're playing it by plugging together blueprint books from other people but I don't think that's how most people play.
At that level of abstraction, it's more like grand strategy if you're to compare it to any video game? You're controlling high level pushes and then the units "do stuff" and then you react to the results.
I think the StarCraft analogy is fine, you have to compare it not to macro and micro RTS play, but to INDIVIDUAL UNITS. For your whole career until now, you have been a single Zergling or Probe. Now you are the Commander.
It's like the Victoria 3 combat system. You just send an army and a general to a given front and let them get to work with no micro. Easy! But of course some percentage of the time they do something crazy like deciding to redeploy from your existential Franco-Prussian war front to a minor colonial uprising...
Note of course that if you’re using gnome or kde then nautilus and dolphin are both good ftp/scp/sftp clients, so not fully sure why you’d want to these days
Fair point, outside my rose coloured memories of Windows 2000, it was likely never a beacon of stability. This is all purely subjective, but in my, frankly not always very reliable memory, I still have the distinct feeling that what has changed is the "in version progression" for lack of a better term.
A fresh install of a later Service Pack Windows XP or Vista did, again purely in my recollection, behaved a lot more stably on the same system to a fresh install of an earlier instance.
8.1 also is of particular note (unpopular UX not withstanding), it worked incredibly solidly on a Netbook with a big colourful sticker proudly proclaiming an entire Gigabyte of memory back in the day, even when using it for image editing via GIMP, for what it's worth.
So I was actually planning on upgrading from a Pixel 7 Pro to a Pixel 10 around the time this announcement came out last year, but have put it on hold as I wait to see what form these changes take.
Like if it was "you need to do the developer 7-tap of the version label in settings", it'd be like "whatever". But given how long this process has taken, I suspect that is not what they've planned - it wouldn't take this long to develop, it certainly wouldn't take this long to explain.
So I suspect that we're actually in the "Maybe Later" phase of "Google wants to control which apps you install: [ ] Yes [ ] Maybe Later". And I mean, if their proposed solution turns out to be "Me and 25 of my closest friends can install apps I make by phoning home to Google servers", then like, I can do that on iOS too. And if I'm not going to have meaningfully more control of my Android device, I may as well just go to iOS where Apple at least have a better privacy record and don't seem to have have an all-encompassing goal of "where can we put AI features to drive AI usage metrics up the most?"
Honestly just install grapheneos on your Pixel, that is what I did and bought a Pixel for that reason alone. I use all Google play services and it works great, only payment with phone doesn't work.
Yes I agree: if you already have a Pixel, try GrapheneOS on it. Then if it can wait (Pixel 7 is still supported for a while, isn't it?), GrapheneOS may support a non-Google phone in 2026, so it may be worth waiting.
There are five options in my country, 3 of which require app push based 2FA to log into the web interface and 2 of which only have an app interfere.
Maybe I could get a EU bank from another EU country but my employer will not accept an out of country account for salary deposits because it makes their tax life difficult and my mortgage provider doesn't trust foreign accounts either.
> It's called a debit/credit card
Since about two years ago, activating a card requires the app.
> Maybe I could get a EU bank from another EU country but my employer will not accept an out of country account for salary deposits because it makes their tax life difficult and my mortgage provider doesn't trust foreign accounts either.
I do not doubt this is happening, but it is forbidden under SEPA. All IBANs, no matter from which member country, must be treated equally. Unfortunately, "IBAN discrimination" happens quite frequently still. The European commission recommends filing a complaint with your national governing body.
It's not just tax obligations, no? Employers in many countries have an obligation to ensure that your salary reflects on the X day of the month (or whatever frequency you're paid). Banks in my country have a payroll payment system for this reason, where funds will clear on the day they're made despite the destination bank (in the same country).
If my employer has to use SWIFT to pay me, on whom does this obligation to ensure I'm paid on time fall? I've had a salary payment from a foreign employer fail to be delivered for 2 weeks a few times. We'd have to go back and forth with my bank, their bank, their payroll vendor. That's an exception because they hired me as a foreign employee. Despite paying their local employees on time, I always received my salary at least 4 days 'late', as long as their payroll system reflected that I was paid on the X day, it wasn't their problem.
so Eire has 5 significant banks, and 15 'less significant'. There are also 276 Credit Unions, I don't know if they are useful. (I had a Credit Union account in the past, could send/receive online but no payment card)
(I don't know their suitability, but there are more than 5 options in your country)
Of the "significant banks" listed, only AIB and Bank of Ireland do consumer bank accounts. I suspect the presence of the others is more to do with wanting an EU entity for targeting larger EU markets than the Irish domestic market. For example, Citibank only expanded from "large tech multinationals" to also "mid sized businesses that are planning to scale internationally" in 2023 [1]
Also on that Wikipedia page are Dell's private bank, Danske Bank (closed their Irish retail business in 2013), Klarna (sort of banking-adjacent, but they're not giving you a current account), etc.
The 5 banks offering retail consumer accounts nationwide are AIB, permanent TSB, Bank of Ireland, Revolut and N26. The first 3 are the surviving brick and mortar banks and the latter 2 are recent-ish neobank entries.
Credit unions are limited to serving customers in their local area. The one credit union who's catchment area I'm in also requires app based 2FA.
(Side note: The name of the country in English is Ireland, the name in Irish is Éire - using the accent-less Irish name in English was promoted by the UK government and BBC because they didn't want to recognise the name of the country prior to the GFA in 1998. Most people will also accept Republic of Ireland if you need to distinguish from Northern Ireland, even though that's technically not the name)
Both of the headline sentences on the home pages tell me they’re self-hosted.
They both also have cloud options, one focussing more on large scale and features I don’t care about, the other others some sort of hosted instance that’s private.
It does, and I've even had Gitlab as the primary repo for some time. But if your projects pick up any steam, github mirrors are going to pop up whether you run them or not - I've had people mirror my projects onto github because it means less questions for them when they want to package them for their organisation or minor packaging system than pulling source from "not-Github". Of course, the license allows them to do that, and they're upfront why they're doing it, but if there's going to be a github mirror anyway, may as well have it official.
Also if we're being honest, despite Gitlab being the #2 platform, you're going to get less contributions than on Github as people just aren't going to want to sign into a second service. Now most of my public projects are like "I made this, I put it here to show off, and use it if you like" so if people _don't_ use it, it's no big deal for me, but if you're in it for revenue or clout or just like seeing usage numbers going up, it's clearly not the optimal choice.
ChatGPT 3.5/4 (2023-2024): The chat interface was verbose and clunky and it was just... wrong... like 70+% of the time. Not worth using.
CoPilot autocomplete and Gitlab Duo and Junie (late 2024-early 2025): Wayyy too aggressive at guessing exactly what I wasn't doing and hijacked my tab complete when pre-LLM type-tetris autocomplete was just more reliable.
Copilot Edit/early Cursor (early 2025): Ok, I can sort of see uses here but god is picking the right files all the time such a pain as it really means I need to have figured out what I wanted to do in such detail already that what was even the point? Also the models at that time just quickly descended into incoherency after like three prompts, if it went off track good luck ever correcting it.
Copilot Agent mode / Cursor (late 2025): Ok, great, if the scope is narrowly scoped, and I'm either going to write the tests for it or it's refactoring existing code it could do something. Like something mechanical like the library has a migration where we need to replace the use of methods A/B/C and replace them with a different combination of X/Y/Z. great, it can do that. Or like CRUD controller #341. I mean, sure, if my boss is going to pay for it, but not life changing.
Zed Agent mode / Cursor agent mode / Claude code (early 2026): Finally something where I can like describe the architecture and requirements of a feature, let it code, review that code, give it written instructions on how to clean it up / refactor / missing tests, and iterate.
But that was like 2 years of "really it's better and revolutionary now" before it actually got there. Now maybe in some languages or problem domains, it was useful for people earlier but I can understand people who don't care about "but it works now" when they're hearing it for the sixth time.
And I mean, what one hand gives the other takes away. I have a decent amount of new work dealing with MRs from my coworkers where they just grabbed the requirements from a stakeholder, shoved it into Claude or Cursor and it passed the existing tests and it's shipped without much understanding. When they wrote them themselves, they tested it more and were more prepared to support it in production...
reply