4 ms·
I'm actually quite upset about this. I love PyCharm and I've always upgraded to most recent major version because even though I didn't necessarily need all the
by shrugisor 11y ago
I'm actually quite upset about this. I love PyCharm and I've always upgraded to most recent major version because even though I didn't necessarily need all the new features, I wanted to support JetBrains for making such great software.
So now I'm stuck using PyCharm 4.x forever unless I pay a monthly fee to upgrade. And if I stop paying I lose access to my IDE? Truly an awful way to treat loyal customers.
I can only hope that JetBrains abandons this new SaaS model or has some plan to allow users to purchase either a perpetual license or a subscription a-la Microsoft Office.
If not, then I hope Atom or Sublime Text get some robust debugging plugins for Python like PyCharm has.
- thawkins 11y agoI agree, im a cLion and a PhpStorm licensee, and this is a disaster. Im going to have to start moving back to netbeans. We where about to kit out our entire department with phpstorm but now its going to be really really hard to justify this. This is nothing about being more convient for customers, its about screwing more money out of people. I am very very disapointed.
- gorohoroh 11y agoFor commercial PhpStorm licenses a rough comparison would as follows. Before November 2: First year: $199 a seat Each subsequent year: $129 a seat After November 2: First year and each subsequent year: $119-149 a seat (with/without a promotion available)
- lol768 11y agoExcept that before November 2nd a company could pay $199 once and never again until they felt the need to upgrade. Following November 2nd, a company has no choice but to cough up the yearly fee or the software stops working and the developers can't do any work without changing the software they use.
- timv 11y agocough up the yearly fee or the software stops working and the developers can't do any work Yup, and all it takes is for an invoice to get misplaced or delayed in your accounts payable team and development grinds to a halt.
- thawkins 11y agoAnd no tool available if we decide not to pay the subscription one year. This is all about forcing companies and licensees into continious payments, its not extending thier choice its severly restricting it. I cannot justify this to my board, as it takes all control over our costs out of our hands and puts us at the mercy of jetbrains pricing policies.
- thawkins 11y agoAnd no tool available if we decide not to pay the subscription one year. This is all about forcing companies and licensees into continious payments, its not extending thier choice its severly restricting it. I cannot justify this to my board, as it takes all control over our costs out of our hands and puts us at the mercy of jetbrains pricing policies.
- konradb 11y agoIs it hard to justify because the department was not going to get regular updates (so you'd half the cost by say, only updating every two years)? If they were going to get yearly updates I can't see the downside other than psychological.
- timv 11y agoThe main issue in places where I work would be that it totally eliminates options. Under the old model: - If the company moves away from a particular technology they can just keep an unsupported, but legal, license for the IDE in case they need to fix bugs, etc. - If the company needs to improve cashflow it can defer upgrades - If JetBrains goes out of business, you can keep using the IDE - If JetBrains discontinues the product, you can keep using the IDE - If JetBrains stops delivering worthwhile upgrades, you stop paying to upgrade. - If JetBrains decides to unreasonably increase the upgrade price you stop paying. (edit: Added) The benefits of the new model are: - It's easier to account for the licensing cost as OpEx rather than CapEx, which matters to some businesses. - It's easier to adjust your license costs in-line with your headcount. Hire some contractors? Add a couple of licenses. Downsizing that team? Drop some licenses. Some organisations will consider that to be a beneficial trade-off, but many won't.