Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
lsh123
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
62 ms
·
751.
▲
by
lsh123
13y ago
Assume your DB query takes 10 secs. It's irrelevant if C++ code takes 0.01 sec when your PHP code takes 0.1 sec.
752.
▲
by
lsh123
13y ago
Well, all applications are different so can't comment on your experience. For me APC does at least 2-3x in the max number of non-DB requests per second. Personally I think it is a lot. As with any technology, ORM is just a tool. Used w
753.
▲
by
lsh123
13y ago
In my statement I kind of assumed that all the "cheap" optimizations in the framework are already done (e.g. PHP APC cache is enabled, lazy classes/configs loading is implemented, etc.) so we compare "apples-to-apples&qu
754.
▲
by
lsh123
13y ago
Disclaimer: I used to write Web apps in C++ in late 90s/early 00s for the largest web company of the time. 1) DB is often a bottleneck for web apps. Not always. But often. Very often. Optimizing pages rendering will improve the perform
755.
▲
by
lsh123
13y ago
It's not about "assuming" things. It's about severe performance degradation beyond any expected levels. Yes, there are solutions to workaround these things. But it will cost way more than a dedicated hardware setup if you account all the co
756.
▲
by
lsh123
13y ago
I've run the numbers for our PCI-compliant setup about 1.5 years ago and Amazon was about 3-4x more expensive than the average bid for my RFP for several smaller hosting providers for a mix of dedicated and virtual servers. Another thing I'
757.
▲
by
lsh123
14y ago
I personally had a frozen pitot tube due to failed tube heating in the solid IMC (no thunderstorm though). You just disable autopilot and manually fly by the known manifold pressure numbers, vertical speed and attitude. Plus double check wi
758.
▲
by
lsh123
14y ago
The old vacuum attitude indicators have a tendency to fail in a really slow and subtle way as the gyro slows down. Yet, there are other instruments in the cabin to figure this out. There are rules of how to detect failed or unreliable ind
759.
▲
by
lsh123
14y ago
Brief technical explanation. The "attack" consists of the following steps: 1) Modify the desktop simulator FMS code to support commands in the data protocols (e.g. ADS-B). 2) Send commands through data link to utilize the newly created cont
760.
▲
by
lsh123
14y ago
For, say, adding a new column or changing indexes, one can probably just run update w/o any code changes. For more complex cases, you might need to modify the code to work with both old/new schemas. It's all about details :)
761.
▲
by
lsh123
14y ago
"Zero Downtime" migration but if the migration will take less than 15 seconds. Honestly, this is not very interesting. The same effect could be achieved by increasing connection timeout on the clients and the server and then just letting th
762.
▲
by
lsh123
14y ago
Not every problem needs to have a technology solution. In this case, the non-technology solution is to pass the fraud cost to the merchants (through fee) who in turn pass it to you (consumer). There is nothing wrong with it as long as eve
763.
▲
by
lsh123
14y ago
The biggest thing in flying is not controlling the airplane but understanding that you are in charge and there is no anyone else who can help you. In the story mentioned above ( http://forumserver.twoplustwo.com/34/other-other-topics/prop..
764.
▲
by
lsh123
14y ago
Compass is one of the most complicated instruments to master in the aircraft because it behaves differently depending on many factors: where you are, where you turn, and are you going up or down: https://en.wikipedia.org/wiki/Aircraft_comp
765.
▲
by
lsh123
14y ago
IMHO, this article is a complete BS from someone who doesn't understand the difference between Bourbaki's math and real modern math (P.S. I used to be a mathematician in my previous life)
766.
▲
by
lsh123
14y ago
Complexity of operations. Simple planes don't really need checklists. There are not that many knobs to turn and not many things can go wrong. As the planes got bigger, faster, and got more equipment, it became impossible to just "follow the
767.
▲
by
lsh123
14y ago
If you read my comment, then you might see that my objections are about citing "scientific studies" incorrectly.
768.
▲
by
lsh123
14y ago
"...and planes will generally continue to fly straight ahead even if you let go of the control yoke." Not really. P-factor, difference in the wings surface, etc. will create a force that will turn the plane. As an IFR certified pilot I can
769.
▲
by
lsh123
14y ago
To say the least, the post is a worst possible example of "science": 1) Bad generalization: "Multitaskers actually become bad at multitasking, by multitasking". Reading the referenced article, we can find out that it talks not about generic
770.
▲
by
lsh123
14y ago
Yes. However, you have same issues with DB tables receiving tons of inserts/deletes/selects. This is the worst possible DB load :)
771.
▲
by
lsh123
14y ago
1) The article claims that queues is a replacement for a database thus it solves all the problems with DB solution. In reality, if you have to have reliability and data integrity guarantees, then a messaging queue system will be using a DB
772.
▲
by
lsh123
14y ago
The performance improvements have probably NOTHING to do with node.js but with the re-architecture goals set by the team: "For our inevitable rearchitecting and rewrite, we want to cache content aggressively, store templates client-side (wi
773.
▲
by
lsh123
14y ago
With WePay: 1) Create "credit card" object from JS 2) Store the credit card ID in your database to charge customer later 3) Charge the customer With Stripe 1) Create a single use token from JS 2) Create Customer object 3) Store the Customer
774.
▲
by
lsh123
14y ago
I've been quoted as WePay employee on HN quite a few times including yesterday :) Regarding PCI compliance, we have a very different opinion about security and compliance. You can compare these two articles and judge by yourself: https://a
775.
▲
by
lsh123
14y ago
* Full support for platform use case including the easiest possible merchant onboarding and ability to charge "app fee" for services. * Support for all experience levels from hosted pages to payment buttons to iFrame solution (full PCI comp
776.
▲
by
lsh123
14y ago
Thanks. Great to be popular
777.
▲
WePay Drops Prices, Rolls Out White Label Payments API
(techcrunch.com)
51 points
by
lsh123
14y ago
|
24 comments
778.
▲
Application Security Checklist
(wepay.com)
1 points
by
lsh123
14y ago
|
0 comments
779.
▲
by
lsh123
14y ago
http://www.wepay.com/api
780.
▲
by
lsh123
14y ago
there is no magic in this world (even if stripe's marketing team wants you to believe that there is)
More ›