Merge (and other companies who take a unified API approach) are great for integrating many tools at the same time, but are not suited for going deep in a few critical integrations like Salesforce.
Many B2B SaaS products require deeper integrations and more customizability beyond what a common data model supports (e.g. custom objects). Supaglue is designed to solve for these use cases.
Honorable, but it could also be a slippery slope: how far back do you look? If an LP, or investor in general, uses their personally earned money, but they earned it through capital gains from tainted money - let’s say from stock in a company with human rights issues - do you ask about that and if so do you take their investment?
If you only look at one degree, ie are your direct investors’ money clean, you could also be targeted for laziness in the “supply chain” of your investors’ money. This is similar to problems with blood diamonds, organic foods, unethical coffee etc.
What's needed is a framework and policy in the community for what's acceptable due diligence.
You can't. What's needed is a framework and agreed-upon policy of what is acceptable due diligence in the community when taking investment money. Varying degrees of "organicness" for startups - like what you see in the meat section at Whole Foods.
In a lot of SEA and EMEA markets road signs, signals, traffic lights, etc. are often non-existent, wrong or broken. Intersections may have 10-20 signs all oriented in a non-standard position.
Throw in motorists and bicyclists that are constantly swerving between lanes and inches from cars. Seeing current videos of how self-driving cars react, it would constantly be jerking in different directions.
There are also lots of local habits that self-driving cars would need to be trained to do, eg in Nepal when in a two lane and you spot an oncoming car, you turn on your right blinker to signal you see each other and can pass safely.
I'm breaking a 4 yr HN commenting-hiatus to post this: it's a bit disappointing to see how the quality of HN comments has deteriorated over the years. Virtually all the comments I'm seeing here are sensationalist sentiments/anecdotes, backed with no facts, reverberated in an echo chamber. This is a plead to the HN moderators to continue to iterate on your product to improve the quality of the comments.
One of the worst parts of open source is when someone who has been happily getting value out of your free product for 4 years opens an issue to complain about how quality has gone downhill, and you need to fix a few things.
I don't think this is fair. The mods want the site to be successful and the parent is pointing out something that should be a cause for concern for them. No one wants HN to turn into the next Slashdot.
But a bug report that says "Your program seems to have gotten slower over the past four years and I feel like it's crashing more often. Please fix it" isn't very useful. A couple of concrete examples of specific problems with some suggestions on what should have been done in those cases would have made the post much more informative.
It's human tendency to glorify the past and undermine the present",
I vehemently disagree with the statement that HN is declining in quality.
concluding that HN is declining is like concluding that we live in a chaotic world, while globally violence is at an all time low [1] . Or undermining Dale Steyn in comparison to Malcolm Marshall or Dennis Lillee [2]. Cricket and Football fans would laugh at comparisons of Kohli with Tendulkar or Neymar with Ronaldinho while records speak otherwise [3] [4]. Aren't we falling prey to the same cognitive bias here.
Literally every forum I have ever been on, when it gets old enough, complains about declining quality and new users. There was a big discussion about this on a popular IRC I used to hang out on. And I spent the weekend setting up a bot to replay messages from 4 years ago in real time. And the general opinion was that the quality wasn't any higher back then.
Same thing with a default subreddit I moderated. We got endless complaints about declining quality. Using web archive to go a few years back, and the quality was just the same or worse!
All the time I read old discussion threads on HN. Either through the search feature or the "past" button. And I really don't see any difference between the old HN and the current one.
If anything it's slighty improved. I think many of the complaints discussed 4 years ago here aren't relevant any more: https://news.ycombinator.com/item?id=6747373 When was the last time you saw a "_____ in js" post? Honestly I miss those a bit.
Ideally I'd love to do statistics and see if stuff like average comment length or political words have increased over time. Maybe I'll try that sometime.
If you think the posted sensational anecdotes are not factual, then you're accusing the userbase of being a bunch of liars. Maybe the product isn't as good as you think?
On a more constructive note, why don't you raise the angle that you think is worth discussing? The allegations are pretty straightforward and Uber looks to be admitting to the mistake, hence the payout. Maybe the posted Bloomberg link should be changed to Quartz, which has a more substantial story with more to discuss than the settled payments: https://qz.com/989040/uber-is-paying-back-tens-of-millions-o...
Or you could also try to add to the discussion and enumerate those facts and sentiments that you feel are being overlooked in the discussion.
It's amusing that you have chosen to break your own self-imposed moratorium on participating in order to wag your finger and criticize everyone else while not contributing anything meaningful yourself.
It sounds like you might live in your own self-righteous echo chamber honestly.
I think the use of "product" here is a bit of a stretch, and what "moderators" are you pleading to? AFAIK it's only dang, and I'm not sure what we can expect him to do when the criticism levelled at him is "the comments aren't as good as I remember"
Also worth remembering is that when companies as polarising as Uber are in the news, there's gonna be heated discussion with lots of strong opinions and anecdata. No amount of moderation can avoid that
> Virtually all the comments I'm seeing here are sensationalist sentiments/anecdotes, backed with no facts, reverberated in an echo chamber.
Point them out here, if you disagree. How do I prove to you that I spoke with a Lyft driver who prefers other car service apps over Uber? I am not going to prove it. Take it as a fact, or don't.
Let people post what they want to say. We don't need everyone to write like a professional critic or columnist. This is not a competition for Best Comment. I like sharing my thoughts with people regardless of how much content I can fill my comment with. You like to keep yourself mute from commentary, that's okay. If you think you are hearing echoes, perhaps there is truth to this? Or perhaps we should conjecture a conspiracy theory that a competitor is running a destroy-uber campaign here on this thread, like they planned a bunch of accounts weeks or months ago. Yeah there are a lot of fake online identities out there (esp during election time). Shall we continue with this conspiracy theory?
Let's move on with our life, we have bigger things to worry about than an echo chamber on HN.
I think that's a reputation Uber has built for themselves. I've seen a couple of commenters here who consistently used to denounce every Lyft article and praise Uber without disclosing they worked there.
i've only recently joined the hn community, and i'm interested by this thought because i find the comments generally inquisitive, regardless of how much someone has posted. upon further research of when the golden age of hn may have been, i found an article talking about whether or not app developers should listen to hn (i'm not linking it because it wasn't particularly enlightening). it lead to the hn comments in reply to the advent of dropbox, where i notice the same structure you're describing here, ten years ago. i suppose i just wish you would provide links to what you're describing as the 'height of hn' if you're describing how it has 'deteriorated over the years'
it seems pretty amazing that a conversation like this takes place in the comments section to Uber ?
The FAQ continues to list the "HN becoming more like reddit" comments (which I'd classify this comment as, considering even in a vaccuum both exhibit regression-to-the-mean based quality deterioration, so it's familiar) as a newcomer misconception, so either the mods here are in denial (worse) or lazy enough to not care (better, but still not great).
I'd take an update to the FAQ as at least acknowledgement of issues here, but tbh I don't care much either. Once you have some feeds up for the voices you do like having weigh in the standard content curation set doesn't matter as much.
There's nothing preventing them from data-mining messages and stripping out identifiable information (like what Google does w/ Gmail) to better target ads in other services.
Photos allow you to discriminate which moment in time you want to convey to people. You can take and retake photos until you distill it to the right scene. Dead simple postprocess editing, e.g. filters, allowed you to enhance that even further.
Videos don't work the same way. Trying to distill the zeitgeist into a video is (24FPS * 15sec) times more work. Without a rich set of postprocess editing tools, i.e. what filters did for photos, that are equally dead-simple, the fantasy-aspiration of what instagram did for photos simply isn't there.
Sure, the trailer instagram showed for the introduction of videos had the same emotional and fantasy components in it that filters brought to photos, except they did a crap ton of postprocessing work on it to get it to that stage. Contrast this with when they showcase photos and filters - they must have done little or no work. Unless the app itself can create the level of videos that instagram showcased, instagram video has brought upon itself the death of fantasy.
The vast majority of that 1m+ employees are low-skilled wage laborers. Finding 3000 high-skill engineers is definitely a monstrous task. Twitter is 7+ years old and they're only at ~1600 employees about half of which are engineers.
Many B2B SaaS products require deeper integrations and more customizability beyond what a common data model supports (e.g. custom objects). Supaglue is designed to solve for these use cases.