Lol For me I’m just too lazy to click out of the repl editor when I’m done instead of returning to whatever reason why I needed a temporary sandbox in the first place, finding my repl, clicking the three dots, deleting it, and confirming the deletion, and making sure that I didn’t delete the wrong one. It just takes 1 occurrence to fall into the infinite-repls-to-delete fallacy, which is really easy to happen.
10 GiBs on the pro plan is such a small amount. I have a raspberry pi and it has 64 GBs.
there are 10 GiBs on the free plan
oh I must not be updated on this
it’s been like that since it rolled out to me
This change makes sense and it was like this before but it was a soft limit but wasn’t really enforced if you did not have bad intent. And having a few hundred users with over 100GBs of storage isent cost-effective by any means and it makes sense.
What bothers is me 2 things
- Replit staff said (IIRC the community slack) there would be tools for us to easly manage our repls/storage so we can get rid of a lot when this was launched.
- We cant buy more with cycles. Weather its a smaller amount per month or a set amount for life for a “dedicated drive”. Instead its hacker and pro. And if you are still using a lot too bad you cant use Replit. This doesn’t affect me using .5GiB but for people who build huge projects it becomes an issue where the single repl takes up that much space on Pro and they cant use Replit at all.
I would like to add some points about the free plan,
Replit attracts beginner users because of its simplicity. Everything is already set up to your liking, and all you need to do is write a single line of code and press a button. At least in my opinion, this is one of the main reasons Replit has reached the popularity it has today.
I can bet thousands of users have watched tutorials from Python programming, to full-stack web development, and used Replit as their IDE. Some of these users probably have no experience with such topics and are just watching to copy off whatever these tutorials suggest. These users don’t want to think about storage capacity, or their RAM and CPU. This is the reason Replit keeps its free users onboard, but this could easily change. How do you think users would react after creating a Node application, and instantly running out of storage?
Replit would most definitely lose some of its free users if something isn’t done. If these beginner users knew how, they would probably switch to programming on their local machine, or use something such as GitHub Codespaces which is much more flexible and has the same if not more capabilities than Replit currently does. It just requires some more technical knowledge. All these users want is to learn how to code, not to pay upwards of twenty dollars a month to be able to create a Repl and some React app.
IIRC node modules are not included in your storage
step to get unlimited storage:
- Move your programs to the node_modules folder
a. make main.py or whatever the main program is refer to your program in node_modules or
b. change .replit to use the file in node_modules as its entrypoint/run target
It may be packages for any languge and not only node modules. But I do not know for sure about node modules. Either way Replit probably checks if its a real package or not
Honestly I was under the impression packages were always counted in storage usage and would continue to be.
I would think programs could be split into an extremely large amount of packages, and use very little source code themselves. If Replit does check
node_modules for actual packages, would you be able to modify legitimate packages to your liking?
Entrypoint and Run are very different. Entrypoint is the file that a user goes in when they enter the repl. Run is the bash command that is run when the run button is pressed
well in python program the entrypoint is the run command, and the entrypoint generally dictates the default file to open regardless.
That would take up so much computational resources and bandwidth to enforce a seemingly ridiculous task, as there are probably countless methods to bypass this. For example, it wasn’t even that hard to release
replit2 on pypi, and I’d assume that the same goes for npm and other open-source project repos. and even if the replit package proxy blocks new packages, users could change the packaging url to the pypi one (.config/pip), and I’d assume there is an equivalent for npm.
TBH you should treat all files equally, and not allow the user to install infinite packages. Instead, just put them under the same shared quota, just like how code intelligence is mixed in with the RAM limits. if the user wants more storage space, then don’t install every package in the packaging index.
I completely agree. Countless users are posting topics on their RAM/CPU usage is almost maxed, even when just browsing code. Constantly checking any edit to
node_modules as a whole would take an extremely large amount of computing power, especially if the amount of packages is large, or a simple
npm update is initiated.
This is pretty off-topic, so someone may want to split this into another topic.
Back on topic though, people are beginning to leave Replit. In its current state, not one of its three plans can be used for large-scale projects, with expansive databases and locally stored assets. It’s just simply not enough storage, and I feel like Replit may be focusing on the wrong things right now.
Let me evaluate that. Replit seems to be focusing most of its attention on trying to make users buy Cycles and upgrade by extremely limiting the Free plan. I don’t agree with this strategy, because some developers do not have the money to upgrade to paid plans and are just using Replit for fun, and the paid plans are, not worth the money at all.
Let’s compare some Replit with some alternatives. GitHub is overall the best option currently. Unlimited repositories, and much higher storage limits, are in their Free plan. I believe GitHub’s Free plan is comparable to Replit’s paid Hacker plan. And if all you want to do is store some old code, you can create a new Google account and instantly get 20 GiB of Free storage. Upgrading is extremely cheap and very much worth it too.
These days, Replit is targeted toward a single type of user. A beginner programmer with projects smaller than a megabyte. It attracts beginners because of its ease. And this is the reason you don’t see many small or large companies developing on Replit. For small companies, the pricing is too unreasonable to consider, and larger ones are probably already using five times more storage than the Pro plan offers.
While Replit may not have the resources as larger companies like GitHub and Google do, if you are charging 20 dollars for 10 GiB, don’t expect to get a large paid user base. Build up funding to provide larger storage, and then consider raising prices to what they are now. A much more reasonable deal would be 100 GiB for 20 dollars, but that is probably still more than what it is worth.
Limiting the Free plan as much as it is currently limited is discouraging users from upgrading as well. Users will want to like the service if they even want to consider upgrading, and limiting them as much as possible does not encourage users to upgrade, or even continue using Replit at all.
Apologies if this sounded rude, I think a lot of improvements to the plans are in order. Here are some of my ideas:
- Private Repl’s need to become 100% free, its very odd this costs any money at all
- Hacker should get at least 4 cores, and preferably more if possible
- Free plan should get larger storage (at least 5 GiB) and outbound data transfer. This may require Replit to moderate alt creation, but 0.5 GiB is almost impossible to work with when doing anything with a backend
- Hacker and Pro should get at least 10x their current storage, they are not worth the money and it is probably the main reason why people choose not to upgrade
- Pro should have increased amounts of always-on and boosted Repls compared to hacker
I’m guessing nothing is going to change, but please consider at least boosting the paid plans’ current storage, as there are many other alternatives with much higher capacities, and Replit will only continue to lose more paying and non-paying users.
Wait a minute, if it was infinite before, why is it so limited now?
So why is Pro limited to 10 GiB? That is an absolutely terrible deal in 2023, and Replit could change it, but they just aren’t?
Sorry for chain of replies, I needed to add this though
Why doesn’t somebody just make an online IDE?
GitHub and Replit are the only IDEs I know that have high language support.
But they both have their problems.
So the best option here is for somebody to make an Online IDE
In fact, I was thinking of making an online IDE myself.
There are many reasons why GitHub and Replit are some of the more used ones. GitHub Codespaces has integration with VSCode, allowing developers to install extensions and work online with https://vscode.dev. Replit is extremely simple to use and has real-time multiuser collaboration. I feel like Replit may be mistreating their users though, and more advanced users are probably moving to Codespaces and locally on their own machines.
Making an online IDE with modern features is a challenge, and is probably one of the reasons there are not many options beyond basic websites with an output screen.