That's incredibly dismissive, and if you think that about any of JetBrain's software, I'm unsure about your qualifications to judge anyone's productivity.
I’m a big user of both Vim and Emacs and agree that they offer huge productivity boosts as editors. The speed that modal editing (Vim) gives you or the extensibility of Emacs is unparalleled.
But at the same time I believe it’s naive to underestimate the productivity offered by IDEs such as IntelliJ or Pycharm. While news tools such as LSP are wonderful additions to Vil/Emacs, the experience doesn’t come close to what dedicated IDEs offer.
My personal flow is to use Vim for true editing, which gives me the power of modal editing, and switch to a full IDE when I need to do extensive refactoring or need code insights or need powerful debugging offered by an IDE.
IdeaVim works fine. I can't use vim sandwich and that's about my only complaint.
I was firmly resistant to IDEs like OP and quite in love with neovim for years. Tried PyCharm because of work, never looking back.
I could never find just a simple debugger for vim that was easy and intuitive to set up. Always had trouble.
And now, the in-IDE Jupyter Notebooks that allow me to use vim keybindings (albeit with an annoying bug that at least has a workaround https://youtrack.jetbrains.com/issue/VIM-2504) has completely changed the way I work.
Are there occasionally show stopper bugs? Yes. Is it worth it for amazing debugger integration? Absolutely.
Having once been a heavy user of emacs and then switched to IDEA, I'm not sure emacs actually is more extensible. The IDEA API is enormous and very powerful. Plugins don't seem particularly limited in what they can do.
I wish there was a modern terminal-first IDE that has the same amount of backing as VSCode or Jetbrains IDEs. While I enjoy vim's speed and modal editing, the number of barely working plugins that break if you so much as look at it wrong is what keeps driving me back to VSCode + Vim emulator.
For every release of neovim we're getting closer and closer to that vision.
I've been using NVChad (just neovim with a preset of mappings and plug-ins all pre-configured for you) for a while now. And it's been working pretty decent.
But I do get what you're saying. It would be nice to have a single (maybe) entity which delivers all needed plug-ins and support which VS code has today.
Here to remind you that real men don’t need compilers. What, you can’t remember every CPU extension? On the latest intel chip from 2000.
What’s that, newer chips offer more features out of the box? How lazy of you.
I spent four years in college studying CS. 2.5 of those years building massively parallel real-time fpga/dsp systems from almost scratch. Even had to build our own boards. Except I didn’t use Emacs. I’m so retarded.
I coax my garden into growing such a way as to condition the butterflies to do the needful. Only amateurs get caught up actually programming bugs themselves.
This is so off base and irrationally elitist, you really just don't see how wrong it is.
These tools do SO MUCH more than just give you a GUI for a coding system. Code analysis, preventative suggestions, and a common look and feel for people switching languages and stacks.
Yes, a lot of it can be cobbled together with open source tooling and scripts and blah blah but I want to spend my time getting shit done, not writing code to be able to write code. The 100$ a year was well worth it for the suite I had, and even double that would still be an irrelevant amount of money to me with my engineering salary.
Time is money, saving time - not just my own, but the time of others on my team using similar tooling - is worth spending a little money and dropping snobbish elitism based on nothing stronger than "I must be smart because I know how something works".
Can vim or eMacs provide auto completion when they detect an SQL query within a string in any language, based on an actual database schema that is connected?
On the corp side, the bump is enough that this becomes problematic. Going from $650 to $780 for that first year is going to cause all sorts of "but what about free" conversations next week. Argh.
I have yet to meet a company that cares about $130 (or even $1,300) in expenses for a productivity tool like this. Maybe that’s because I wouldn’t work for one that did.
Well when 80% of all US Dollars were "printed" / created in the last 2-3 years massive inflation will follow
What we are seeing is the predictable result of the classic monetary models, and a rejection of Modern Monetary Theory which linked limits of government spending to the availability of real resources, like workers. Which is why you see people like The Treasury Sec saying the Economy is good because of employment numbers
Classic Model of economics predicted this, MMT ignores the realities of economics
>> What we are seeing is the predictable result of the classic monetary models
Respectfully, models that are known to be wrong, as all models are - they are models after all.
Austerity - where has that ever worked? Post-war Germany and... ??? I mean it's constantly tried everywhere (Greece is a good recent example of its futility). Classic monetary models clearly predict it should work, we've tried it hundreds of times in different countries over the past ~100 years, and it's worked once. Maybe we just need to try extra super duper hard the next time?
My skepticism rises when i hear people confidently proclaim their economoic model predicts anything at the macro level. What we're seeing is a lazy attempt to ignore the models are not useful. If prices stay low for 5 more years than the models predict then we are told by the over-confident types that we should just wait a little longer to see the effect but they're coming. If prices go up then "AHA! Justified, i told you so". Prices can ONLY go down, up or stay the same - if you wait long enough of course they go up!
There's a big wooly gap between alleged cause and effect that we're just supposed to not question when people confidently tell us their model predicted the outcome.
Prices are not rising because people have too much money and are over purchasing food and energy just to show off to their neighbours.
If that were the case then there are things that would work. Increase taxes on income, unearned income, wealth and excess profits to take money out of the economy (MMT theory would support this if the issue was genuinely too much money) Introduce price controls and rationing on food and basic services including energy and housing to stop profiteering.
>>Prices are not rising because people have too much money and are over purchasing food and energy just to show off to their neighbours.
I dont believe I said that in any statement, that is red herring and strawman in one. Good job...
Most of the expansion of the monetary supply has been doled out in the form to debt not cash. However buying goods (Cars Home, etc) with massive amounts of debt still causes inflation
Of course Supply chain issues account for some inflation as well, but not all of it.
>> Introduce price controls and rationing on food and basic services
In the entire history of the world, that has never worked, nor will it ever work
I guess? But if there are 5 times more dollars to buy why do they become more expensive (for me, outside the US)? That doesn’t seem intuitively correct to me.
>Well when 80% of all US Dollars were "printed" / created in the last 2-3 years massive inflation will follow
Be careful. This theory tracks for regular currencies, but gets distorted (note: I didn’t not say that it doesn’t apply, only that it isn’t as strong as you say) when the currency in question is the reserve currency for the world.
Classic Model didn’t predict this, because “this” is a totally bonkers scenario. Inflation is running hot (additionally) because supply chains have been torched (China and Russia) just at the time when demand came roaring back (the world minus China gave up on lockdowns).
You’re not wrong, but I encourage you to resist the temptation to simplify the story down to “classic” vs “MMT.” It’s tempting, but the world is very very complex and it ain’t getting any simpler.
MMT also predicted this. It says you're supposed to stop spending when resource constraints are reached otherwise you will see inflation. It's actually a much better description of how money actually flows
The problem was that everyone remembered the GFC and went way too far with spending in an already supply constrained environment.
Hahaha, good to know the fact checkers are there to tell me that no, 80% of all hard currency has not literally passed through a printing press in the last “two months.” Need a good laugh after seeing [1]. And yeah, I know most of the increase is due to an accounting change.
your link proves why a Disinformation Governance Board is a TERRIBLE idea.
Notice the comment says "printed" / created, as I was using "printed" in the Figurative meaning not literal meaning
The Fed can create new money with out requesting that money be physically printed by the BEP. This is why a "run on the banks" is a fear, because no bank in the world has enough physical currency to cover all "dollars" in accounts. Only a fraction of dollars exist as a physical item
If you're selling a product today, why wouldn't you increase your prices? What force would stop you from looking around and saying to yourself, everyone else is jacking up prices so i might as well. You could release a little press release that says "blaa blaa blaa inflation blaa blaa" and everyone would more or less accept it.
Say you made cars for America and through no fault of your own, dealer stock evaporated due to supply issues upstream. Let's say you could try hard to increase production once that supply issue had been ironed out - but you could equally choose to take it easy, cars are fetching well over sticker price, so maybe be very slow at ramping up supply again to maintain that. Obviously if you colluded with your competitors to do that it would be illegal, but, if there was some kind of global event that put you all in the same basket...
I see the talk of increasing interest rates to solve inflation, i just can't see how that stops price gouging inflation. Furthermore, we don't even appear to have a language for this idea, it's not supply side inflation i'm talking about (although that clearly exists in some products), it's not demand side inflation (although that clearly exists for toilet paper for a while - not sure if it affects much today?)
It is a three headed dragon, and we see two of them already.
* Massive increase in money supply (more money chasing goods) - CHECK. We printed and our sins do not go away as quickly as we wish they would.
* Supply issues resulting in fewer goods to buy (same or more money chasing fewer goods) - CHECK.
* Wage-Price spiral wherein employees demand higher wages to account for their costs, which increase the cost to the employer, which has to increase prices, which leads to employees demanding higher wages. - POTENTIALLY NOT HERE YET.
Wage-Price spirals are the things of nightmares and when it happens, the only way its ever been broken is via war or a prolonged recession.
People at the low end of the pay scale working in travel, food service, hospitality, cleaning, etc, got an opportunity to not grind themselves down for a pittance.
It was already happening throughout the 2010s, although wages in tech were rising sufficiently fast to easily offset the increase in prices. My conjecture is the bottom 0% to 40% are making gains and they are coming from the 40% to 90%.
With a lower and lower labor participation rate (aging population), coupled with the opportunities for remote work, I would expect those in person, customer service or jobs where you get your hands dirty to demand even higher prices. I.e. expect to pay more for travel/restaurants/hotels/retail/etc and expect to receive lower quality services/products.
USD is becoming worthless, the fed kept printing infinite money and people are sick of receiving toilet paper
Also the fact that they had to fire their entire Russian R&D department, and instead rely on Westerners, they are more expensive, so they are forced to increase their price overall, that's my guess
Germany is much worse than the US. Huge inflation with a microscopic salary increase and unlike FED, European central bank doesn't do anything at all to fight inflation (except increasing interest rates to laughable 0.25% - 7x lower than in US). Basically, ECB want to devaluate Euro, so that our salaries can be even lower than in China.
Inflation in Germany is different than inflation in the US. We have huge supply side issues because of the war in Ukraine
The US just printed a lot of money.
They are both inflation but very different kind.
In the case of US the money supply changed A LOT, while in the case of germany the cost of goods changed.
You realize the ECB has probably "created" just as much money through their sovereign bond purchases than the FED? If anything, the ECB has been much much less conservative about it's monetary policy than the FED, and that has been true for years. Just look at the rates.
I understand the frustration but surely you can see the causality link here? Jetbrains employees have to eat too.
The individual solution is sadly for you to ask for a 20% raise and screw people down the line.
Global solution would be enforced price controls, we know it works because France did it after the war, the government imposed lower prices and started selling good directly itself and France got out with a much lower inflation than other countries.
Price controls don’t work; the US figured this out during WW2 when it just led to shortages, had spotty enforcement, and needed a gigantic bureaucracy to actually determine prices.
It’s easy to see why price controls fail to work: if there’s 10 guys willing to pay 10 bucks for 10 apples, setting the maximum price of an apple to $1 isn’t gonna change the fact that more people want apples than there are apples. This either means queues so those most willing to queue gets things (the soviet model), black markets where things are bought at the true value (nonsensical official exchange rates see this), or a mixture of both (WW2 US).
We currently have inflation because of a few things, mainly that there’s too much money chasing too few goods. So the equilibrium is that raised prices lead to people buying fewer things because they can’t afford as many, and a new supply/demand equilibrium is established.
But of course this can’t happen if wages are also going up, as people just try to but the same always amount of things with their newly increased salary, and we’re back to the problem of not enough good again.
The only way this gets solved is a mixture of increased supply of good, decreased demand through either decreasing real salaries, increasing real prices, and quantitative tightening.
People don’t like hearing this, because hearing that you’ll meet to have less purchasing power in order for supply to meet demand means that things that you could afford previously can no longer be bought. Saying that wages shouldn’t go up is a non-starter, and rightfully so, which is how inflation-wage spirals get started.
Basically, the only way out that’s politically feasible (I mean politics in the general sense of people interacting with each other, not government) is to trigger a recession and then everybody is forced to buy fewer things and nobody has to say “we’re not raising your salary because that just exacerbated the problem”.
Blaming price control as the issue or really any one exonomic argument for Argentina won’t work. Argentina has been a failure in so many ways. Not all of it only the countries fault either.
We can see having tons of natural resources can be great. Canada and Norway are good examples. There are specific reasons why most other resource rich countries in Africa, ME, South America, have not fared as well.
> Global solution would be enforced price controls, we know it works because France did it after the war, the government imposed lower prices and started selling good directly itself and France got out with a much lower inflation than other countries.
That was way more complicated than you described. War-related price controls are very different because of many things, and even in countries whose territory wasn't directly affected people experienced shortages constantly.
Well I get upset when IT guys complain about stuff being not free and price hikes.
It is a pet peeve with the software developers whingeing about the cost of tools.
I guess I overreacted to your statement. I remember spending over $300 in todays money for Delphi when it first come out in 1995, and in my view it was well spent.
It wasn't even in employment related then. I considered the tool useful for my personal desires.
So seeing professional people complain about the prices of tools which are extremely low-priced when adjusted for inflation kind of bugs me. As the guys said their prices haven't gone up for seven years.
If you consider JetBrains subscriptions expensive, try Delphi, or the commercial versions of Lisp and Smalltalk.
The parent poster says they notice a 20% increase in prices, and you reply with paragraphs on paragraphs of "you have a mental problem" and "stop whining". A bit ironic, if you ask me.
You are reading way too much into the parents comment.
>This is the first time in my life (34 yo) that price changes have been so large that they’re apparent to me.
We lived through a unique time in history, and that unique time has lasted for our entire adult lives. I’m also 34. I graduated college right when the low interest rates were juicing the economy. But the party had to come to an end sometime.
Ask your parents what it was like in the 1970s. Oil shocks, interest rates at 15%. History rhymes.
Good point. While the interest rates are “high” now (prime rate is 4.75?) that is actually a regression towards the mean, and definitely not 1970s hyperinflation territory.
I wonder though — given that we can’t go to that 15% rate (because then servicing public debt would be too expensive), what other tools do we have left to tame inflation?
We deliberately raised the gas prices to punish Putin. Gas prices affect the price of food due to shipping. Food prices affect the price of software due to eating. Don't worry, it's all under control. We will soon fix this by printing more money, prohibiting gasoline cars and making people eat insects for protein. Meanwhile, we suggest you weather the storm by investing most of your money into cryptocurrencies[1].
With 10 USD, I can buy 1 weeks worth of food in my country. I wish SASS, software providers like Jetbrains adjusts their price based on country per capita.
People in rich countries hate it when companies do that. There's also the additional hassle of verifying addresses. All so you can get a little extra money from some poor countries. I can see why they wouldn't bother.
Steam does it very well for games, pricing games differentially such that they are much cheaper in 3rd world countries. JetBrains could do something similar too.
Steam basically turned it on for every developer until a lot of them noticed how little are they getting from some countries.
If iPhones and TV cost the same I'm not sure why software products should.
Not the parent - it is definitely a better pay than other salaries in the country, but for example in Hungary a junior gets around like 2000 USD (per month) gross (and we have quite huge income tax here as well). Around 2600 USD it would be considered a nice senior salary.
Most JetBrains products are super slow, and resource hogs. I get red mist in front of my eyes whenever I see a DataGrip icon as it tends to consume all your CPU and memory for complex databases regardless of your machine spec.
These days I am mostly using VsCode for development and TablePlus/Postico for database development.
I thought I was the only one. For me it was GoLand, not to mention the constant feeling that im being watched. VSCode without telemetry is what I moved to.
JetBrains IDEs ask you whether you want to submit usage data on the first startup, and honor your choice. You are not watched if you don't give an explicit, opt-in permission for that.
I have a MacBook Pro with 16GB of memory and i5 processor. It is sufficient for VsCode, however for some more complex Jetbrains products I sometimes struggle to copy and paste from one file to the other. On an 8GB machine VsCode is likely struggling as well, so a resource hog as well, just less so than some JetBrains products.
> Most JetBrains products are super slow, and resource hogs
Yeah, every time i've tried CLion or Rider, the heap reaches its maximum size (4-8gb) like 10-20 minutes after starting the editor and then it becomes extremely slow and sluggish. My cpu goes crazy any time i start an analysis or any sort of task. I've tried to disable on the fly analysis, plugins and whatever i could find that i don't always need to be enabled, but it's barely better.
On the other hand, I've used Intellij for years and it's never been that bad. It has much lower cpu/ram consumption even while having almost everything enabled. I've used the same configuration for years though, so it's possible that at some point I disabled whatever causes such resource usage and completely forgot about it. Or maybe the editors are actually different, though from my understanding they all share the same core.
The only reason they take up more resources than necessary is that the default VM settings are bad. They use an older GC algorithm with very low RAM setting, but all it does is make the GC run much more than it would with a decent size.
If you happen to try it out again, I recommend setting the heap size to at least 4GB and removing the obsolete GC setting (defaulting to the G1 collector) or even better, changing to ZGC. Hopefully they will fix it in a newer version soon.
I don’t have much to add but what I do have is a question. What happens when people are priced out of economic activity. Most recent memory of this much economic turbulence is ‘08 but the conditions between then and now are incomparable.
By the way I think it would be nice to also have a hobbyist license option for those who barely use it and don't earn any money this way. The actual price seems almost free in terms of how much a full-time developer would earn using it yet slightly expensive for something [awesome] I only launch on some occasions a year.
They do if you contribute to an open source project. There's also Pycharm Community edition which is available through the package manager of most Linux distros.
We support OSS independently of their popularity. If you have had issues with your project being supported (due to maybe some automated checks we've put in place to help us), please feel free to reach out to me with the details. My email is hadi at jetbrains dot com.
Yeah, I saw this coming when they sent around a survey a couple of months back asking what people thought of their current pricing and what price increases they would stomach before ditching it.
An extra £16 per year for IntelliJ seems OK as rises go.
@jetbrains: Please add support for your JetBrains 'space' as a task server in your IDEs. Yes, there's already YouTrack tickets for this, from months ago.
In general I think a price increase is reasonable but Jetbrains also had to fire most of its staff because of the war and development has slowed to a crawl.
I still pay because I want to support them and their stance but paying more for less leaves a bad aftertaste.
They didn't fire most of their staff. They closed the Russian offices and offered relocation to everyone in Russia (and many took it).
I mean they were practically forced to close their Russian offices with the current political situation and war going on.
The relocation process is very expensive as well as opening new offices in Europe and rehiring there as CoL and salaries in Europe are much higher.
I know they were forced. I know they did it the most humane way possible. This is precisely why I support them financially.
Doesn't change the fact that they had to let go of a huge portion of their staff. Call it firing, or special unemployment operation, the result is still the same; and I can't help but notice that updates are now very, very sparse and pretty minor ( maybe its a pure coincidence because they are busy with Fleet and the Java update who knows ).
All in all it’s understandable and not a big deal but when faced, like right now, with whether I should prepay 3 years or not it's hard not to think about the future of the company.
I get so much value out of JetBrains products that if a price increase is what is needed or them to stay on business I can accept that. (The products are paid by my employer so this only affects indirectly)
It might be an unpopular opinion but I appreciate the transparency of the post. There are other companies which products I use that I’m afraid are not financially sustainable and thus I’m afraid for an email saying that the products has been discontinued.
There’s always the possibility of open source, home brewed or whatever, but when it comes to getting stuff done sometimes paying is cheaper than the time/effort of an alternative.
Agreed. I also get a license from my employer, but continue to pay for the all products pack separately because they make such nice stuff, and I'm happy to support them in this small, mostly symbolic way.
Same. For me, it’s the JavaScript development experience that really sets Jetbrains apart.
I’m sure other IDEs can offer parity with the compiled languages, but if I’m using an IDE in the first place, I’m probably working on a web app, and getting first class support for both the frontend and backend in the same window is huge for productivity.
Nicely communicated. Everything that JetBrains does around their pricing and subscription model always feels really fair and classy to me. (perpetual fallback licensing, continuity discounts).
In this case, being completely upfront in the post title that this is a price increase, no "Updated pricing" or "Revised pricing" etc. Reasonable notice period with an option to renew for multiple years at the current pricing before the change comes into effect.
On https://account.jetbrains.com/licenses if you click "Renew" for the appropriate product then on the next page, under the Subscription dropdown there should also be a "Purchase Period" dropdown. It's probably set to "Default" rather than "1 year", but can be changed to "2 years" or "3 years". Does that option not appear for you?
Sort of. You can keep using the product version from when you last paid (the beginning of the sub not the end). So if you were annual then you actually have to downgrade to a 1 year old version. It's a really wonky system. I was pretty upset when they switched to it but I love their products enough that I gave in.
No, its like a landlord doing renovations and improvements to keep up with the local property market and then putting rent up to match those improvements and maintain parity with the market. (with notice and options to pay old level for period upfront)
More like a landlord doing renovation, feeling the budget crunch, and looking at the calendar and realizing "Oops! I forgot to raise rent in the past 7 years!"
The price increase they're asking for is approximately equal to inflation from 2015-2021. They don't even try to include the last year's inflation.
The heavier IDEs are in real trouble given the impressive state of VS Code lately. I haven't found the need to reach for a bigger IDE than Code in a while and I've been perfectly productive in JS, C, Rust and Python projects lately.
Do yourself a favour and try out Code for your next project.
After using vscode for years, I've tried nvim with lunarvim starter pack and am using it more. Using vscode without mouse is always a problem even with vim plugin. On the other hand, mouse is almost useless in nvim.
Not sure why you'd say a mouse in Neovim is useless. A mouse can be very useful in both Vim and Neovim when set up correctly, it will even work across SSH.
Even just using two fingered trackpad scrolling to move the view port is something I do almost everyday without thinking about it.
If you wanted to you could also highlight text, show balloon help, select windows and tab and interact with file explorers.
Of course you can do all that very quickly with keyboard shortcuts, which is what I mostly do, but they option of the mouse is still there.
The problem wasn't Eclipse - it was Google's plugins that were buggy. I hated the transition over to Android Studio.
Eclipse's Java core is by far the best of all Java IDEs: refactoring, auto-complete, editing, etc. IntelliJ had earlier support for Maven and Gradle, but it was always a less-polished IDE experience using Swing vs Eclipse's native windowing toolkit.
Maybe it's improved in the past 3ish years, but the last time I had to use Eclipse for a project I couldn't stand it. It may have been lighter weight, but it was buggy and had many fewer helpful features (refactorings, inline hints, navigation). I couldn't wait to get back to intellij.
I don't think IntelliJ has added support for running non-compiling Java code yet, which was a major productivity booster for me.
Eclipse's ability to debug and run a single failing unit test were massive as well. IntelliJ was always hit or miss as to whether it would allow you to run a subset.
IntelliJ refactorings also were a lot less smart IMO - often trying to refactor unrelated comments and sometimes code that was unrelated but happened to be named the same thing.
I don't mean to sound rude here, but that just sounds like nonsense to me. How are you supposed to run code that won't compile? Does eclipse just generate random bytecodes for the party that doesn't compile?
> debug and run a single failing unit test
I'm not sure where the issue is for you, but this has always been available since I have used IntelliJ, going back 15 years.
> An incremental Java compiler. Implemented as an Eclipse builder, it is based on technology evolved from VisualAge for Java compiler. In particular, it allows to run and debug code which still contains unresolved errors.
True, I had to switch to intellij about six years ago because Eclipse just wasn't able to handle projects I was working on (gradle, spring boot) anymore. It just made a mess of the gradle support and I gave up on it. Never saw that working properly in Eclispe. These days with Kotlin, it's just not an option. The Kotlin support is just not really usable.
Features I still miss:
- Real incremental compilation. Real as in, lift your finger and it's ready to run before you manage to hit the key-combination that actually launches the test. Intellij under the best of circumstances does not get closer than five to ten seconds (full seconds). Eclipse could run the same test dozens of time in the same time span. It was that fast.
- An IDE that does not lie about the compilation state of my project. Intellij is perpetually confused about this. It will tell me everything is fine when it is not. And the opposite. I can never trust it until I do a full rebuild. Which is slow. Eclipse was much more consistent with this. Make typo, everything goes red instantly. Fix the typo, the red disappears. It mostly stayed consistent throughout a working day. Occasionally it would need a little nudge and then it would fix itself and get back to being super responsive. Typically when messing with dependencies.
Those two are connected of course.
Intellij is two orders of magnitudes slower than Eclipse for this. That's a 100x. I'm talking ms vs. ten seconds plus. I could make a case for 3 orders of magnitudes for some things. I deal with scenarios every day where it needs half a minute (or worse) to boil the oceans where I know from a decade plus of experience that Eclipse would have caught up nearly instantly and it would not have slowed me down.
Unfortunately, Intellij would have to be redesigned from the ground up to get close to this level of performance. It's simply designed wrong to ever get that good. The fact that it relies on gradle (or maven) to do a lot of things means it's going to be bottleneck. Eclipse had a real incremental compiler properly integrated into the IDE and a good enough integration with maven that it could import classpath dependencies from there without needing to run it all the time.
It even did partial compilation when part of your code was wrong/broken (e.g. because you had not finished typing yet). It would still be able to run the non broken parts of it. And it would accurately tell which parts were broken. Intellij freaks out if that happens and quite often needs to be restarted to get rid of all the red squiggly lines for phantom problems. It will spawn lots of errors that aren't errors until you fix the actual error. There's a reason it has synchronize buttons, invalidate caches, and other functionality to force it to re-assess the state of things: it loses track of this all the time. Their issue tracker is full of issues related to this. People have been reporting them for decades (just in case somebody unhelpfully suggests I create an issue for this).
So, yes, I miss that. Kotlin particularly is a pain in the ass. As much as I like the language, I really hate it's compiler speed (or lack thereof).
While I wish it were faster, I actually appreciate that Intellij can defer as much as it does to the underlying build tooling. The native ability of the IDE to interpret the full contextualized classpath of a given project is a double-edged sword. While it _can_ improve iteration speed, it also introduces a lot of room for IDE-specific environmental issues where something works in the IDE, but not in the build (and vice-versa).
Pros and Cons, but my experience spending hours or even days solving IDE classpath hell in large projects is a huge annoyance to me, especially when the same issues don't actually exist in the 'real' build. So much that I'm willing to pay a few more seconds for every task/build initiation in large projects if that's the cost of having an accurate representation of the classpath in the IDE.
I also really enjoy Kotlin, but find the weird configuration and versions of the kotlin sdk bleeding between the IDE, build tool (gradle) and actual dependencies to be a source of confusion. This is especially true when building/publishing gradle plugins written in kotlin.
Thanks, but no thanks. I mainly code in PHP and VS Code even with proper plugins and Intelephense can not come near PHPStorm's utility. I've tried everything for PHP nowadays, Vim + LSP, Sublime + LSP, VSCode. It's PHPStorm all the way.
I feel the opposite way. VSCode feels clunky and disjointed compared to the experience I get from JetBrains IDEs.
Also, both have 3rd party plug-ins but JetBrains seems to contain a lot more of the features I want as first class integrations whereas I generally have to find a plugin for VSCode and hope it's well maintained.
The first class integrations are key. I think VSCode can probably do just about everything GoLand can, but you need to configure a half dozen plugins to get there.
I will say that writing a VSCode plugin seems much easier than an IntelliJ plugin. So if you need that level of customization, VSCode might be the preferred tool.
I prefer Vscode to WebStorm, the plugin ecosystem is just a lot more mature, and I feel like Vscode is visually less cluttered and organized better.
But I would die before I switched away from Rider or Intellij. The auto suggestions, code smells, auto refactoring, ease of debugging, I could go on. It removes so much friction from developing that I feel like I have a hand tied behind my back when I move to anything else.
I often wonder if these people who recommend vscode as a "for everything" IDE have enough experience with the IDEs to know what they are missing by strong arming vscode into working for them.
Unfortunately Microsoft forces .Net developers to use Visual Studio since they don't really offer a compelling .Net/C# answer in VSCode. OmniSharp exists, but outside of toying it is worse than Visual Studio Community.
A lot of developers looking for that VSCode experience are actually switching to Rider, even those who already have a Visual Studio license/MSDN via their employer. Ultimately Visual Studio is just a laggy, micro-stuttery, monolith, with a UI designed to create Forms for Visual Basic back in 1995. The x64 migration has helped, but it feels like a band-aid.
If you're doing Web Development in Asp.net Core in 2022 you don't really feel at home, not least of all because you may be using VSCode already to do TS/SASS/CSS/JS/etc.
It's a fine lightweight editor but if you want to refactor or do anything complex you want something like IDEA in your pocket. It comes pretty much "Batteries included" whereas you have to build your own IDE with VSCode which I don't have the time or effort for maintaining. I get that IDEs can look scary or seem bloated but once moving first to PHPStorm then IDEA I realized a lot of my "facts" about IDEs were either completely wrong and/or the heavier weight was actually worth it.
I'll be interested to see if JetBrain's new Fleet will eat into VS Code's market (it's a similar tool).
Big problem for me is that they are using proceeds from well-selling products like ReSharper to fund their strategic but under-performing products like Rider, all while neglecting the former.
Over last few years there were no big features added in ReSharper whatsoever and it became slower.
Where do you get information about JetBrains's sales and engineering expenses to draw this conclusion?
One reason Rider will pull ahead of Resharper is because Rider builds on the same base IDE as most of their products, so it benefits to a large extent from investment in any of those IDEs. I'm assuming that Resharper can only benefit from direct investment.
This comes as no surprise and was in fact expected by Jetbrains. Each time they increase prices, VSCode becomes more attractive to developers. On top of that, Microsoft is doing a very clever pricing scam as I said before [0]; by making their base tools free and charging for exquisite features in VSCode like Copilot.
On other editors, either you use a barebones editor and pay for Copilot or you pay for both copilot and the editor subscriptions. Microsoft wins both ways and as evidently shown here, JetBrains conceded to make a subscription price rise.
Jetbrains doesn’t get my money because of how much of a confusing mess their product lineup is. They have a bunch of different IDEs sold separately that ultimately seem to be identical except for language support. Why isn’t language support implemented as plugins if all the IDEs share a common framework? Last I checked, it seemed like they do have an “ultimate” version with multiple language support… except that one doesn’t support C/C++.
I was interested in buying CLion, but also needed to be able to work with Python in the same project. I couldn’t tell if that means I have to also buy the PyCharm IDE and deal with two different IDEs running at once.
It also doesn’t help that they have a subscription licensing plan. Even though they do have perpetual fallback, it doesn’t seem like a giant bloated IDE like this will hold up very long without updates.
So in the end, I just use Sublime Text. It’s far from perfect, but I can make it do nearly everything I need it to without having to strain my brain at the checkout screen.
> It also doesn’t help that they have a subscription licensing plan. Even though they do have perpetual fallback, it doesn’t seem like a giant bloated IDE like this will hold up very long without updates.
FWIW, I stopped updating paying for IDEA in 2019 and it hasn't bothered me. I don't do JVM stuff every day anymore, that's why I stopped, but when I have to go hack on something, it's just fine.
Agreed on the multiple IDEs thing. I don't write C++ regularly but that'd make me tear my hair out.
IntelliJ has two totally different meanings, by itself the name doesn't really identify a product.
IntelliJ Community is pretty much just a Java (and other JVM language) IDE.
IntelliJ Ultimate works with all the languages that work with their other IDEs (Python, Go, JavaScript, ...). This is the one you're talking about. C++ isn't listed even for this though [1]
It's definitely confusing, even if Ultimate did support C++, because IntelliJ is primarily billed as a Java IDE. And with PyCharm, for example, the free vs paid difference is many useful extra features but still primarily aimed at the same language. Maybe they should've just called it JetBrains Ultimate IDE or something, rather than mentioning their Java IDE.
Sublime is great. Clion supports Python (and you can edit any file from any of their products.) The easiest way to figure out if you like their apps before purchase is to just run the EAP version of one if your evaluation period expired.
I do not want my product checked constantly -- I want to pay for a year and then leave me alone ! The trend is towards constant monitoring and associated disconnect immediately with a monthly payment. I object to that model.
I pay for a large jetBrains product right now. Not a problem to raise the price a bit. Excellent company and products in my experience.
Before I clicked the link I assumed this was a sensationalised headline. It's interesting how direct the communication is here.
Normally companies will try to justify their price increases in a lengthy blog post focusing on the "awesome new features" they've been working on, but not here. Just a few paragraphs summarised by, "we are at the point where we need to increase our subscription prices". Fair enough.
As a user obviously price increases are never welcome, but within reason I think we have to be understanding. JetBrains offers a huge amount of value to many companies and individuals. It might not be worth the price to everyone, but their pricing doesn't seem unreasonable to me and quality software is never going to be cheap.
I've been a paying customer since 2006. It's worth every penny and more and while I don't love paying more, I'm content to pay the cost for great dev tools.
I almost always have at least one subscription with Jetbrains. This all sounds reasonable, especially with competition from VSCode ecosystem as well as LSP support improving.
I see you point. Some people will buy their products without too much regards to cost. I am sure they work hard to optimize profit vs. product pricing.
Immediately paid for 3 more years without a second thought though the pricing is completely reasonable and I wouldn't have had an issue paying the new price.
Also I give JetBrains massive respect for including the old/new pricing on their pricing page versus what so many companies do which is "our prices have changed" without making it clear and leaving people to using the Wayback Machine to figure it out. Also allowing people to renew up to 3 years at the old price is a move that inspries a lot of loyalty (as shown by my immediate purchase/extension).
I understand if you don't program enough to make it worth paying for a tool but if you develop professionally then the JetBrains products are a steal. I really only use IDEA (PHP, TS, Java) and DataGrip (databases) but I pay for the full product pack in case I ever need another tool. I recently used AppCode and was very pleasantly surprised by it as I quite dislike Xcode for actual coding.
180 comments
[ 0.23 ms ] story [ 226 ms ] threadGreat idea, I bet many will take advantage of the deal.
But at the same time I believe it’s naive to underestimate the productivity offered by IDEs such as IntelliJ or Pycharm. While news tools such as LSP are wonderful additions to Vil/Emacs, the experience doesn’t come close to what dedicated IDEs offer.
My personal flow is to use Vim for true editing, which gives me the power of modal editing, and switch to a full IDE when I need to do extensive refactoring or need code insights or need powerful debugging offered by an IDE.
I was firmly resistant to IDEs like OP and quite in love with neovim for years. Tried PyCharm because of work, never looking back.
I could never find just a simple debugger for vim that was easy and intuitive to set up. Always had trouble.
And now, the in-IDE Jupyter Notebooks that allow me to use vim keybindings (albeit with an annoying bug that at least has a workaround https://youtrack.jetbrains.com/issue/VIM-2504) has completely changed the way I work.
Are there occasionally show stopper bugs? Yes. Is it worth it for amazing debugger integration? Absolutely.
I've been using NVChad (just neovim with a preset of mappings and plug-ins all pre-configured for you) for a while now. And it's been working pretty decent.
But I do get what you're saying. It would be nice to have a single (maybe) entity which delivers all needed plug-ins and support which VS code has today.
What’s that, newer chips offer more features out of the box? How lazy of you.
I spent four years in college studying CS. 2.5 of those years building massively parallel real-time fpga/dsp systems from almost scratch. Even had to build our own boards. Except I didn’t use Emacs. I’m so retarded.
These tools do SO MUCH more than just give you a GUI for a coding system. Code analysis, preventative suggestions, and a common look and feel for people switching languages and stacks.
Yes, a lot of it can be cobbled together with open source tooling and scripts and blah blah but I want to spend my time getting shit done, not writing code to be able to write code. The 100$ a year was well worth it for the suite I had, and even double that would still be an irrelevant amount of money to me with my engineering salary.
Time is money, saving time - not just my own, but the time of others on my team using similar tooling - is worth spending a little money and dropping snobbish elitism based on nothing stronger than "I must be smart because I know how something works".
Everything I pay in dollars became 20% more expensive.
My cereal became 20% more expensive.
Jetbrains becomes 20% more expensive.
This is the first time in my life (34 yo) that price changes have been so large that they’re apparent to me.
And even with a dev salary the combined amount is high enough that it has an impact on my behavior.
What we are seeing is the predictable result of the classic monetary models, and a rejection of Modern Monetary Theory which linked limits of government spending to the availability of real resources, like workers. Which is why you see people like The Treasury Sec saying the Economy is good because of employment numbers
Classic Model of economics predicted this, MMT ignores the realities of economics
Respectfully, models that are known to be wrong, as all models are - they are models after all.
Austerity - where has that ever worked? Post-war Germany and... ??? I mean it's constantly tried everywhere (Greece is a good recent example of its futility). Classic monetary models clearly predict it should work, we've tried it hundreds of times in different countries over the past ~100 years, and it's worked once. Maybe we just need to try extra super duper hard the next time?
My skepticism rises when i hear people confidently proclaim their economoic model predicts anything at the macro level. What we're seeing is a lazy attempt to ignore the models are not useful. If prices stay low for 5 more years than the models predict then we are told by the over-confident types that we should just wait a little longer to see the effect but they're coming. If prices go up then "AHA! Justified, i told you so". Prices can ONLY go down, up or stay the same - if you wait long enough of course they go up!
There's a big wooly gap between alleged cause and effect that we're just supposed to not question when people confidently tell us their model predicted the outcome.
If that were the case then there are things that would work. Increase taxes on income, unearned income, wealth and excess profits to take money out of the economy (MMT theory would support this if the issue was genuinely too much money) Introduce price controls and rationing on food and basic services including energy and housing to stop profiteering.
I dont believe I said that in any statement, that is red herring and strawman in one. Good job...
Most of the expansion of the monetary supply has been doled out in the form to debt not cash. However buying goods (Cars Home, etc) with massive amounts of debt still causes inflation
Of course Supply chain issues account for some inflation as well, but not all of it.
>> Introduce price controls and rationing on food and basic services
In the entire history of the world, that has never worked, nor will it ever work
Here are 100 roubles for you, comrade.
Which is why you see China, Russia and other nation using the current opportunity to attempt to parallel economy to the one based on US Dollar
Be careful. This theory tracks for regular currencies, but gets distorted (note: I didn’t not say that it doesn’t apply, only that it isn’t as strong as you say) when the currency in question is the reserve currency for the world.
Classic Model didn’t predict this, because “this” is a totally bonkers scenario. Inflation is running hot (additionally) because supply chains have been torched (China and Russia) just at the time when demand came roaring back (the world minus China gave up on lockdowns).
You’re not wrong, but I encourage you to resist the temptation to simplify the story down to “classic” vs “MMT.” It’s tempting, but the world is very very complex and it ain’t getting any simpler.
The problem was that everyone remembered the GFC and went way too far with spending in an already supply constrained environment.
I've heard this is a completely debunked conspiracy theory:
https://leadstories.com/hoax-alert/2022/04/fact-check-80-per...
You see, this is why Disinformation Governance Board was a good idea.
[1]: https://fred.stlouisfed.org/series/M1SL
Notice the comment says "printed" / created, as I was using "printed" in the Figurative meaning not literal meaning
The Fed can create new money with out requesting that money be physically printed by the BEP. This is why a "run on the banks" is a fear, because no bank in the world has enough physical currency to cover all "dollars" in accounts. Only a fraction of dollars exist as a physical item
Say you made cars for America and through no fault of your own, dealer stock evaporated due to supply issues upstream. Let's say you could try hard to increase production once that supply issue had been ironed out - but you could equally choose to take it easy, cars are fetching well over sticker price, so maybe be very slow at ramping up supply again to maintain that. Obviously if you colluded with your competitors to do that it would be illegal, but, if there was some kind of global event that put you all in the same basket...
I see the talk of increasing interest rates to solve inflation, i just can't see how that stops price gouging inflation. Furthermore, we don't even appear to have a language for this idea, it's not supply side inflation i'm talking about (although that clearly exists in some products), it's not demand side inflation (although that clearly exists for toilet paper for a while - not sure if it affects much today?)
It is a three headed dragon, and we see two of them already.
* Massive increase in money supply (more money chasing goods) - CHECK. We printed and our sins do not go away as quickly as we wish they would.
* Supply issues resulting in fewer goods to buy (same or more money chasing fewer goods) - CHECK.
* Wage-Price spiral wherein employees demand higher wages to account for their costs, which increase the cost to the employer, which has to increase prices, which leads to employees demanding higher wages. - POTENTIALLY NOT HERE YET.
Wage-Price spirals are the things of nightmares and when it happens, the only way its ever been broken is via war or a prolonged recession.
Hold onto your hats if that happens.
It was already happening throughout the 2010s, although wages in tech were rising sufficiently fast to easily offset the increase in prices. My conjecture is the bottom 0% to 40% are making gains and they are coming from the 40% to 90%.
With a lower and lower labor participation rate (aging population), coupled with the opportunities for remote work, I would expect those in person, customer service or jobs where you get your hands dirty to demand even higher prices. I.e. expect to pay more for travel/restaurants/hotels/retail/etc and expect to receive lower quality services/products.
Also the fact that they had to fire their entire Russian R&D department, and instead rely on Westerners, they are more expensive, so they are forced to increase their price overall, that's my guess
> Jetbrains becomes 20% more expensive.
I understand the frustration but surely you can see the causality link here? Jetbrains employees have to eat too.
The individual solution is sadly for you to ask for a 20% raise and screw people down the line.
Global solution would be enforced price controls, we know it works because France did it after the war, the government imposed lower prices and started selling good directly itself and France got out with a much lower inflation than other countries.
Also using an example from 70 years ago doesn’t seem enough to justify it.
Haha, no, I didn’t, but now that you point it out it seems obvious ;)
That said, I somehow doubt Jetbrains employees will see these raised prices reflected in their salaries.
We can see having tons of natural resources can be great. Canada and Norway are good examples. There are specific reasons why most other resource rich countries in Africa, ME, South America, have not fared as well.
Jetbrains products were always cheap, and now they are more expensive. I never said it’s now a bad deal.
I’ve just never seen prices rise so fast. Even during the whole pandemic when China shipping basically stopped, rising prices were unnoticable.
Now prices rise more, and it surprises me, that’s all.
It is a pet peeve with the software developers whingeing about the cost of tools.
I guess I overreacted to your statement. I remember spending over $300 in todays money for Delphi when it first come out in 1995, and in my view it was well spent. It wasn't even in employment related then. I considered the tool useful for my personal desires.
So seeing professional people complain about the prices of tools which are extremely low-priced when adjusted for inflation kind of bugs me. As the guys said their prices haven't gone up for seven years.
If you consider JetBrains subscriptions expensive, try Delphi, or the commercial versions of Lisp and Smalltalk.
You are reading way too much into the parents comment.
We lived through a unique time in history, and that unique time has lasted for our entire adult lives. I’m also 34. I graduated college right when the low interest rates were juicing the economy. But the party had to come to an end sometime.
Ask your parents what it was like in the 1970s. Oil shocks, interest rates at 15%. History rhymes.
Except the G-20 average public debt in the 70s was like 30% of the GDP.
We cannot go to interest rates at 15% with countries having 100%+ public debt (as is the case nearly everywhere in Europe for example).
I wonder though — given that we can’t go to that 15% rate (because then servicing public debt would be too expensive), what other tools do we have left to tame inflation?
We deliberately raised the gas prices to punish Putin. Gas prices affect the price of food due to shipping. Food prices affect the price of software due to eating. Don't worry, it's all under control. We will soon fix this by printing more money, prohibiting gasoline cars and making people eat insects for protein. Meanwhile, we suggest you weather the storm by investing most of your money into cryptocurrencies[1].
[1] https://www.nasdaq.com/articles/cryptocurrency-and-the-esg-i...
Developer wages too vary from country to country.
These days I am mostly using VsCode for development and TablePlus/Postico for database development.
> These days I am mostly using VsCode for development
Please qualify the difference in speed between JetBrains and VS Code. From my perspective, they are both “super slow” and “resource hogs”.
Yeah, every time i've tried CLion or Rider, the heap reaches its maximum size (4-8gb) like 10-20 minutes after starting the editor and then it becomes extremely slow and sluggish. My cpu goes crazy any time i start an analysis or any sort of task. I've tried to disable on the fly analysis, plugins and whatever i could find that i don't always need to be enabled, but it's barely better.
On the other hand, I've used Intellij for years and it's never been that bad. It has much lower cpu/ram consumption even while having almost everything enabled. I've used the same configuration for years though, so it's possible that at some point I disabled whatever causes such resource usage and completely forgot about it. Or maybe the editors are actually different, though from my understanding they all share the same core.
If you happen to try it out again, I recommend setting the heap size to at least 4GB and removing the obsolete GC setting (defaulting to the G1 collector) or even better, changing to ZGC. Hopefully they will fix it in a newer version soon.
By the way I think it would be nice to also have a hobbyist license option for those who barely use it and don't earn any money this way. The actual price seems almost free in terms of how much a full-time developer would earn using it yet slightly expensive for something [awesome] I only launch on some occasions a year.
https://www.jetbrains.com/community/opensource/#support
Are you sure you don't mean substantial contributions to a very popular open source project?
Thanks.
The process to renew was super easy too, although I ended up just buying it so I could use it at work too.
An extra £16 per year for IntelliJ seems OK as rises go.
I still pay because I want to support them and their stance but paying more for less leaves a bad aftertaste.
All in all it’s understandable and not a big deal but when faced, like right now, with whether I should prepay 3 years or not it's hard not to think about the future of the company.
It might be an unpopular opinion but I appreciate the transparency of the post. There are other companies which products I use that I’m afraid are not financially sustainable and thus I’m afraid for an email saying that the products has been discontinued.
There’s always the possibility of open source, home brewed or whatever, but when it comes to getting stuff done sometimes paying is cheaper than the time/effort of an alternative.
Best of luck jetbrains!
I’m sure other IDEs can offer parity with the compiled languages, but if I’m using an IDE in the first place, I’m probably working on a web app, and getting first class support for both the frontend and backend in the same window is huge for productivity.
In this case, being completely upfront in the post title that this is a price increase, no "Updated pricing" or "Revised pricing" etc. Reasonable notice period with an option to renew for multiple years at the current pricing before the change comes into effect.
It's like a landlord doing unrequested renovation and then sending a bill.
https://sales.jetbrains.com/hc/en-gb/articles/206544679-What...
I can stop paying and carry on using the product indefinitely. Try that with your landlord regardless of renovations!
The price increase they're asking for is approximately equal to inflation from 2015-2021. They don't even try to include the last year's inflation.
Do yourself a favour and try out Code for your next project.
Even just using two fingered trackpad scrolling to move the view port is something I do almost everyday without thinking about it.
If you wanted to you could also highlight text, show balloon help, select windows and tab and interact with file explorers.
Of course you can do all that very quickly with keyboard shortcuts, which is what I mostly do, but they option of the mouse is still there.
IntelliJ was always heavier and buggier whenever I tried using it.
Google switched from Eclipse to IntelliJ for Android Studio mostly because everyone hated how heavy and buggy Eclipse was.
Eclipse's Java core is by far the best of all Java IDEs: refactoring, auto-complete, editing, etc. IntelliJ had earlier support for Maven and Gradle, but it was always a less-polished IDE experience using Swing vs Eclipse's native windowing toolkit.
Eclipse's ability to debug and run a single failing unit test were massive as well. IntelliJ was always hit or miss as to whether it would allow you to run a subset.
IntelliJ refactorings also were a lot less smart IMO - often trying to refactor unrelated comments and sometimes code that was unrelated but happened to be named the same thing.
I don't mean to sound rude here, but that just sounds like nonsense to me. How are you supposed to run code that won't compile? Does eclipse just generate random bytecodes for the party that doesn't compile?
> debug and run a single failing unit test
I'm not sure where the issue is for you, but this has always been available since I have used IntelliJ, going back 15 years.
https://www.eclipse.org/jdt/core/
> An incremental Java compiler. Implemented as an Eclipse builder, it is based on technology evolved from VisualAge for Java compiler. In particular, it allows to run and debug code which still contains unresolved errors.
Features I still miss:
- Real incremental compilation. Real as in, lift your finger and it's ready to run before you manage to hit the key-combination that actually launches the test. Intellij under the best of circumstances does not get closer than five to ten seconds (full seconds). Eclipse could run the same test dozens of time in the same time span. It was that fast.
- An IDE that does not lie about the compilation state of my project. Intellij is perpetually confused about this. It will tell me everything is fine when it is not. And the opposite. I can never trust it until I do a full rebuild. Which is slow. Eclipse was much more consistent with this. Make typo, everything goes red instantly. Fix the typo, the red disappears. It mostly stayed consistent throughout a working day. Occasionally it would need a little nudge and then it would fix itself and get back to being super responsive. Typically when messing with dependencies.
Those two are connected of course.
Intellij is two orders of magnitudes slower than Eclipse for this. That's a 100x. I'm talking ms vs. ten seconds plus. I could make a case for 3 orders of magnitudes for some things. I deal with scenarios every day where it needs half a minute (or worse) to boil the oceans where I know from a decade plus of experience that Eclipse would have caught up nearly instantly and it would not have slowed me down.
Unfortunately, Intellij would have to be redesigned from the ground up to get close to this level of performance. It's simply designed wrong to ever get that good. The fact that it relies on gradle (or maven) to do a lot of things means it's going to be bottleneck. Eclipse had a real incremental compiler properly integrated into the IDE and a good enough integration with maven that it could import classpath dependencies from there without needing to run it all the time.
It even did partial compilation when part of your code was wrong/broken (e.g. because you had not finished typing yet). It would still be able to run the non broken parts of it. And it would accurately tell which parts were broken. Intellij freaks out if that happens and quite often needs to be restarted to get rid of all the red squiggly lines for phantom problems. It will spawn lots of errors that aren't errors until you fix the actual error. There's a reason it has synchronize buttons, invalidate caches, and other functionality to force it to re-assess the state of things: it loses track of this all the time. Their issue tracker is full of issues related to this. People have been reporting them for decades (just in case somebody unhelpfully suggests I create an issue for this).
So, yes, I miss that. Kotlin particularly is a pain in the ass. As much as I like the language, I really hate it's compiler speed (or lack thereof).
Pros and Cons, but my experience spending hours or even days solving IDE classpath hell in large projects is a huge annoyance to me, especially when the same issues don't actually exist in the 'real' build. So much that I'm willing to pay a few more seconds for every task/build initiation in large projects if that's the cost of having an accurate representation of the classpath in the IDE.
I also really enjoy Kotlin, but find the weird configuration and versions of the kotlin sdk bleeding between the IDE, build tool (gradle) and actual dependencies to be a source of confusion. This is especially true when building/publishing gradle plugins written in kotlin.
Also, both have 3rd party plug-ins but JetBrains seems to contain a lot more of the features I want as first class integrations whereas I generally have to find a plugin for VSCode and hope it's well maintained.
I will say that writing a VSCode plugin seems much easier than an IntelliJ plugin. So if you need that level of customization, VSCode might be the preferred tool.
But I would die before I switched away from Rider or Intellij. The auto suggestions, code smells, auto refactoring, ease of debugging, I could go on. It removes so much friction from developing that I feel like I have a hand tied behind my back when I move to anything else.
I often wonder if these people who recommend vscode as a "for everything" IDE have enough experience with the IDEs to know what they are missing by strong arming vscode into working for them.
A lot of developers looking for that VSCode experience are actually switching to Rider, even those who already have a Visual Studio license/MSDN via their employer. Ultimately Visual Studio is just a laggy, micro-stuttery, monolith, with a UI designed to create Forms for Visual Basic back in 1995. The x64 migration has helped, but it feels like a band-aid.
If you're doing Web Development in Asp.net Core in 2022 you don't really feel at home, not least of all because you may be using VSCode already to do TS/SASS/CSS/JS/etc.
It's a fine lightweight editor but if you want to refactor or do anything complex you want something like IDEA in your pocket. It comes pretty much "Batteries included" whereas you have to build your own IDE with VSCode which I don't have the time or effort for maintaining. I get that IDEs can look scary or seem bloated but once moving first to PHPStorm then IDEA I realized a lot of my "facts" about IDEs were either completely wrong and/or the heavier weight was actually worth it.
I'll be interested to see if JetBrain's new Fleet will eat into VS Code's market (it's a similar tool).
Over last few years there were no big features added in ReSharper whatsoever and it became slower.
One reason Rider will pull ahead of Resharper is because Rider builds on the same base IDE as most of their products, so it benefits to a large extent from investment in any of those IDEs. I'm assuming that Resharper can only benefit from direct investment.
On other editors, either you use a barebones editor and pay for Copilot or you pay for both copilot and the editor subscriptions. Microsoft wins both ways and as evidently shown here, JetBrains conceded to make a subscription price rise.
Microsoft knows you can't beat free.
[0] https://news.ycombinator.com/item?id=31826190
They have never increased subscription prices. They still charge the same amount they did 7 years ago, when subscriptions were introduced.
That doesn't explain why Windows and Office (both paid products) have edge over desktop Linux and LibreOffice (free).
I was interested in buying CLion, but also needed to be able to work with Python in the same project. I couldn’t tell if that means I have to also buy the PyCharm IDE and deal with two different IDEs running at once.
It also doesn’t help that they have a subscription licensing plan. Even though they do have perpetual fallback, it doesn’t seem like a giant bloated IDE like this will hold up very long without updates.
So in the end, I just use Sublime Text. It’s far from perfect, but I can make it do nearly everything I need it to without having to strain my brain at the checkout screen.
FWIW, I stopped updating paying for IDEA in 2019 and it hasn't bothered me. I don't do JVM stuff every day anymore, that's why I stopped, but when I have to go hack on something, it's just fine.
Agreed on the multiple IDEs thing. I don't write C++ regularly but that'd make me tear my hair out.
IntelliJ Community is pretty much just a Java (and other JVM language) IDE.
IntelliJ Ultimate works with all the languages that work with their other IDEs (Python, Go, JavaScript, ...). This is the one you're talking about. C++ isn't listed even for this though [1]
It's definitely confusing, even if Ultimate did support C++, because IntelliJ is primarily billed as a Java IDE. And with PyCharm, for example, the free vs paid difference is many useful extra features but still primarily aimed at the same language. Maybe they should've just called it JetBrains Ultimate IDE or something, rather than mentioning their Java IDE.
[1] https://www.jetbrains.com/products/compare/?product=idea&pro...
https://www.jetbrains.com/products/compare/?product=clion&pr...
But yes, it's simultaneously weird and great that the most cross-application languages (Python/Javascript) are supported in every IDE.
I pay for a large jetBrains product right now. Not a problem to raise the price a bit. Excellent company and products in my experience.
It's an awesome product, I recommend everybody to use it. I personally use it with its amazing Scala plug-in, also friendly licensed.
Normally companies will try to justify their price increases in a lengthy blog post focusing on the "awesome new features" they've been working on, but not here. Just a few paragraphs summarised by, "we are at the point where we need to increase our subscription prices". Fair enough.
As a user obviously price increases are never welcome, but within reason I think we have to be understanding. JetBrains offers a huge amount of value to many companies and individuals. It might not be worth the price to everyone, but their pricing doesn't seem unreasonable to me and quality software is never going to be cheap.
Competing with a free product by raising prices? Might work with corporate accounts.
Also I give JetBrains massive respect for including the old/new pricing on their pricing page versus what so many companies do which is "our prices have changed" without making it clear and leaving people to using the Wayback Machine to figure it out. Also allowing people to renew up to 3 years at the old price is a move that inspries a lot of loyalty (as shown by my immediate purchase/extension).
I understand if you don't program enough to make it worth paying for a tool but if you develop professionally then the JetBrains products are a steal. I really only use IDEA (PHP, TS, Java) and DataGrip (databases) but I pay for the full product pack in case I ever need another tool. I recently used AppCode and was very pleasantly surprised by it as I quite dislike Xcode for actual coding.