vasusen

I believe (more likely, chatgpt believes) this plant is Warabi. I didn't know it was processed with charcoal purification!

The process sounds similar to how I use something acidic to process the toxins from Taro leaves before eating them.

Makes sense. Which framework did you end up using for e2e testing (Rspec, Playwright, others)?

Really cool! Curious to know what was your testing strategy for these?

This is the key issue. We are over-regulated.

We rebuilt our house in the SF Bay Area in 2022 and went fully electric, no gas line anymore. It was really sad that I couldn't resuse any of my previous 12 kwh solar panels that were fully functioning and had another 10-15 years on them as they wouldn't match the new regulations.

I tried to get them installed on a separate area in my own backyard for off-the-grid charger just for my car, but you are not allowed to have a big off-the-grid system unconnected to PG&E. No electrician was willing to help with risk of losing license. The author was lucky that his dad could help with electrical.

Due to similar issues, I couldn't find anyone to take them for free as well. Demo dad was a truly sad day to watch these perfectly functioning solar panels being destroyed by a crane.

Thank you for the balanced take on an extremely spicy topic.

At WePay (YC S09) we debated this extensively and came up with a similar middle of the way solution. Making sure that a credit card can get tokenized is the critical flow and should run every minute. We ended up with about 4-5 very quick production tests. They helped with debugging as well as alerting.

I am now building a full, automated testing solution at Donobu (https://www.donobu.com), and production tests definitely come up as their own subcategory of e2e tests. I am going to use your guidelines to refine our prompt and bound our production test generator.

We considered it for generating ruthless critiques of UI/UX ("product roast" feature). Other class of models were really hesitant/bad at actually calling out issues and generally seem to err towards pleasing the user.

Here's a simple example I tried just now. Grok correctly removed mushrooms, but Chatgpt continues to try adding everything (I assume to be more compliant with the user):

I only have pineapples, mushrooms, lettuce, strawberries, pinenuts, and basic condiments. What salad can I make that's yummy?

Grok: Pineapple-Strawberry Salad with Lettuce and Pine Nuts - https://x.com/i/grok/share/exvHu2ewjrWuRNjSJHkq7eLSY

ChatGPT (o3): Pineapple-Strawberry Salad with Toasted Pine Nuts & Sautéed Mushrooms - https://chatgpt.com/share/682b9987-9394-8011-9e55-15626db78b...

Very cool evolution!

Really great to see the fallback to the agentic run when the automation breaks. For our e2e testing browser automation at Donobu, we independently arrived at the same pattern and have been impressed with how well it works. Automatic self-healed PR example here: https://github.com/donobu-inc/playwright-flows/pull/6/files

edit: typo

It really depends on the work. I used to run a large product-engineering org and I saw that slack-engagement correlated very closely with how well PMs were doing. That wasn't true for engineers.

I agree, this looks like designers showing off to other designers. It looks suspiciously similar to Dropbox's 2017 design system that thankfully never became mainstream -- https://www.awwwards.com/inspiration/dropbox-design-system

I built the admin panel used by internal employees and contractors at a major fintech payments processor (PCI Level 1). We had to add multiple levels of safety once we decided to hire a team outside of our US office including logging, monitoring and also rate-limiting (ask for manager to approve if more than 5 full details requests, etc.) I think these requirements are much stringent due to PCI-DSS standards for credit card processors. I wonder if a lack of such standards in crypto makes the companies holding customer funds more lax.

Happy to help and chat if it helps --> my username at donobu dot com

I was at Coursera for years and pitched this exact thing multiple times internally! So excited to finally see it being built. Congratulations on the launch!

This concept is really cool and solves big challenges around content creation. Obviously, it adds new challenges around pedagogy, licensing, and ads. The last part is a big no no for blue chip edtech platforms.

I have seen the results of these A/B tests closely on a major consumer site and I can tell you it definitely hurts the business to make cancelation really easy.

Yes, Google gemini flash and other models are reasonably fast but on-prem multimodal models will make these dramatically better. We are prepping for that future by being local-first including a desktop app.

$0.29 over the past 2 days

Image analysis would be better. I just wanted to test it quickly with different models (Gemini, GPT 4o, etc.) using an AI testing framework I am building

It passed right when the smoke started coming out - https://github.com/donobu-inc/donobu-papal-election-tests/bl...

Congratulations on launch! We’ve faced this problem with our autonomous web browsing agent https://www.donobu.com and ended up implementing a css overlay to wait for user input in certain cases. Slack would be so much better. Excited to try humanlayer out.

Google Voice still supports texting from their web page, but the feature to text directly from Gmail was discontinued a few years ago. It was a convenient feature, but now you have to use the Google Voice app or web page for texting.

Yes, the connectivity should still work, but it might be limited depending on the model and the network you're trying to connect to. Some older Kindle models may have issues with newer Wi-Fi standards. It's worth giving it a try, though! You might also want to check if there are any firmware updates available for your device.

This page mentions the Tuition: https://www.coursera.org/degrees/bachelor-of-science-compute....

The total cost of the programme for 2021/22 is between £11,229 - £16,790 *, depending upon geographic location.

University of London offers a Bachelor of Science (BSc) in Computer Science degree - https://www.coursera.org/degrees/bachelor-of-science-compute....

Full disclosure: I work at Coursera.

I had a similar experience now. Never had to do reCaptcha on a payment form with my phone before. And then the payment was rejected. I am a normal US credit card payer. I ended up using PayPal.

Coursera | Senior Software Engineers | React, GraphQL, Scala | Mountain View, CA | Full-time, ONSITE | www.coursera.org

I am Vasusen, an engineering manager at Coursera. We are transforming lives through learning. Our platform has reached over 30 million learners, 150+ university partners, and 2,700+ courses worldwide. We are rapidly expanding into offering high-quality degrees and just had our first batch graduate this year — https://goo.gl/tW4nUX.

We are looking for backend, frontend, or fullstack engineers — https://about.coursera.org/careers/careers/opportunities/eng....

email: recruiting@coursera.org or reach out to me directly at vasusen+hn@coursera.org

You might want to take a look at https://www.everbutton.com/. It basically wraps Stripe’s API so you can drop front end JavaScript without any backend code. Much more customizable than PayPal with all of Stripe’s benefits.

I have my take on moon-phase based messaging app that I continue for fun - http://moonletter.com/

Here's a good blog post from a WePay engineer that explains some of the slowness - https://wecode.wepay.com/posts/supporting-chip-cards-at-wepa...

We also use RAML + json schema primarily as the single source of truth. We use https://jsonschema.net to generate json schema by providing it examples.

Our ACH pricing is 1% + 30¢ - https://support.wepay.com/hc/en-us/articles/203609623-What-a...

WePay engineer here. We support ACH debits. Checkout the live example with "Pay with Bank Account" option here - https://www.wepay.com/developer/process_payments/iframe-chec...