Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
_tom_
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
27 ms
·
271.
▲
by
_tom_
4y ago
How do you get germs in something that is only running municipal water in it? There no (or very little)biological matter to feed on, no light to provide energy.
272.
▲
by
_tom_
4y ago
Im seeing ads now for "create content faster with GPT-3. So expect quality to go down even farther, to meet the demands of capturing eyeballs as cheaply as possible. I've also started seeing review websites that are clearly ai gen
273.
▲
by
_tom_
4y ago
No, but NLP toolkits like opennlp already do this quite well.
274.
▲
by
_tom_
4y ago
You could write normally and then split on sentence boundaries with any NLP tool. Then pipe the output into a diff. Don't format your writing to suit your tools. Hacker news today says that OpenNLP 2.0 just came out. That should work
275.
▲
by
_tom_
4y ago
If it starts making money, publisher will put in the effort to block it. It's not a viable business.
276.
▲
by
_tom_
4y ago
Why do I doubt that they are doing this with their C-level compensation?
277.
▲
by
_tom_
4y ago
The problem with non-cloud cameras, I think, is that burglars can, and likely will steal them. This kind of defeats the purpose of a security system.
278.
▲
by
_tom_
4y ago
The source is 10 years old. Is there a more recent approach? (e.g. Amazon FPGA instances.)
279.
▲
by
_tom_
4y ago
1% is, of course, a ridiculous number. You'd need scientific notation to express how wonderful you would be after 10 years, if you could do that.
280.
▲
by
_tom_
4y ago
Assets don't have value, just because the are on the books. If the assets are bonds of questionable value, the face value of the bonds may not be meaningful.
281.
▲
by
_tom_
4y ago
Oh, man. I remember that. It seemed like no matter how careful you were, your machines would not be the same, and you'd hit issues. We went to VMs to get things consistent.
282.
▲
by
_tom_
4y ago
The Lucene search library has something like this. Data is stored /retrieved via codecs. Most write binary data, but one codec actually writes out all data as text. It is, of course, too slow for production use, but it is great for fig
283.
▲
by
_tom_
4y ago
The average Apartment in California is over $3000. If there is anything that is 500, it will be worse than a tent.
284.
▲
by
_tom_
4y ago
It's so much worse than that. Some payroll tax laws vary by SCHOOL DISTRICT of the employee.
285.
▲
by
_tom_
4y ago
I worked on a tax app. I was amazed to find that some tax tables necessary to compute taxes were not released until months after taxes were due. WTF?
286.
▲
by
_tom_
4y ago
If they get the self driving cars working, they become a tech company. Of course, then they won't be able to shift expenses to drivers.
287.
▲
by
_tom_
4y ago
I think many people in the comments here, and most users, are missing that you index a SMALL subset of the web. This leads to people running a default test search, finding no results, and concluding your search engine is bad, and leaving. W
288.
▲
by
_tom_
4y ago
Google is demonstrating this nicely now. It's become almost useless, replacing the query I actually typed with something more popular. And when that doesn't happen, the results are likely seo'd junk. (The latter is not purel
289.
▲
by
_tom_
4y ago
It doesn't index sites like Reddit, so, not too surprising Reddit wasn't in the result.
290.
▲
by
_tom_
4y ago
I received a "we have filled all positions" email from them and I have not been interviewing, or even responding to their emails.
291.
▲
by
_tom_
4y ago
Would you care to talk about how you build it? I assume static, but do you use something like Hugo?
292.
▲
by
_tom_
4y ago
I ran a quick test, averaged 268 milliseconds. Yep, that's "essentially immediately. We forget how fast the web can be, because of all the bloated sites out there.
293.
▲
by
_tom_
4y ago
If I were rich, sure. $10k for Me is like $0.10 for Bezos. But with your last $10k? That's just stupid (or fiction)
294.
▲
by
_tom_
4y ago
If you are familiar, could you post a short explanation? I'm also curious as to why buying in may is too late. Edit: https://www.treasurydirect.gov/indiv/research/indepth/ibonds... Doesn't explain w
295.
▲
by
_tom_
4y ago
He might have a look at NAND2tetris and see if that grabs his interest. Arduinos are also fun.
296.
▲
by
_tom_
4y ago
The purchase is completed before he gets it home and finds new conditions being imposed. That the part that I question about being legal.
297.
▲
by
_tom_
4y ago
> completely objective. The really sad thing is that people have discovered that "completely objective" doesn't sell very well. Taking an extreme view sells much better, but provides much less useful information.
298.
▲
by
_tom_
4y ago
Anyone that thinks renting houses is passive income has never done it. Repairs, collections, advertising, evictions, etc.
299.
▲
by
_tom_
4y ago
Many (Most?) ebooks on amazon require steady advertising to not drop into oblivion. This makes it decidedly less passive. Source: my authors group
300.
▲
by
_tom_
4y ago
May depend on your market, it amazon sells the vast majority of ebooks in the world. So, neither of those.
More ›