Another voice in the fray, I also did well enough to bypass the technical screen. Interviewed with four companies they recommended. Picked one. Still with them two(-ish) years later and an outspoken advocate.
It's a tricky situation because you have to balance being accepting of people with non-traditional backgrounds while also making sure they can do the basics. I liked that TripleByte could condense DAYS of interviews into a few hours.
In general, long tech screens are hard on people that can't afford to take off a full day, so it's tricky for people already working in crushing jobs that need a change. The TripleByte quiz worked nicely for this.
On the flip side, if their tech screen is filtering people for reasons other than ability, it's removing possibly good candidates who deserve a shot at something.
The new approach is sounding like it's trying to satisfy the latter, which is good for improving inclusivity, but won't be to everyone's liking.
For what kind of tagging are you looking? Per-image labels? Single-tag-per-image? Image rectangles?
Most of the tagging programs I've encountered really aren't that awesome. For such a relatively common task (for ML, at least), there are few, if any, really good pieces of workflow software out there.
How many people do you have on your ml team so far? I'm doing ML research (NLP for determining writing quality and similarity, amusingly) for my company and it's getting a bit lonely.
We have
- 1 ML/NLP engineer (me)
- 1 CEO w/ a linguistics background
- 1 founder is an English professor at Berkeley
- 2 with previous experience teaching English
- 1 Berkeley PhD in Deep Learning / NLP advising us
- 1 Berkeley PhD in English helping us categorize writing issues
We're working on assessing writing quality too. Get in touch!
Hi jo_
I'd love to talk with you about Alexa's NLP and ML research team in Cambridge, MA ebbounty@amazon.com send me a note! We have a robust team of senior and principal engineers and scientists to learn from
Perhaps they were fixed as part of unrelated bug fixes? That would explain why they didn't credit anyone as having reported them. Something like, "Fixed bad return value leak," or "refactoring old method" could break the exploits.
Forgive my ignorance, but it seems like this is just attempting to take advantage of the optimization done by LLVM, yes?
What I would love is a simple way of writing standalone functions that compile into a cross-platform LLVM file that I can call from a variety of other languages on a variety of other systems. In particular, if I train a recurrent network on text data for a chat bot, I want to be able to use that LLVM file + model in a game I release for the PC and for Android without worrying about the NDK/gcc/clang/Windows/OSX build nightmare. The ability to easily and quickly define a model in TensorFlow, write a Python function that takes an array of data, and spits out an array of data would be incredible and would mean that all the work I'm doing for a native Rust library is unneeded.
Admittedly, with Bazel I could create a C++ wrapper for the function which loads the library. It's just... that produces a 150mb shared library with all the dependencies and it's also a pain in the ass.
This is actually easy to do, you just need to generate the IR and then merge them into a Module, after that you apply passes over the entire module to optimize, to do function inlining, etc.
I'm a current WebPass subscriber and I'm dreading the move to another ISP when I change apartments. May I inquire what it takes to set up and maintain the transmitter?
I believe the transmitters are pretty expensive. In the $10,000 range, depending on the frequency and power required.
So there's cost if you want to own one. Plus you need to convince someone (Webpass I guess?) to point one of theirs at you to complete the link.
It's probably easiest to convince the owner of your new apartment to pay Webpass's setup fee, where they own the equipment and wire the building. Although having tried that a couple of times, it's a pain in the ass.
The receivers themselves for a consumer you can get for < ~$1,000 for one that will be able to handle a couple of hundred Mbps easy. Getting someone else to point directly at you would cost a lot. Or doing both ends yourself. The business model for ie. WebPass is not to point to every customer, but to point to one point in say a neighborhood and then run cables from that to every house in the neighborhood. Which is a fraction of the cost of running cables all over cities.
At Least, that will be the business model under Google Fiber.
People buckle because they are victims of NEEDS. If I didn't have marketable skills, or if I existed in a market which was oversaturated with people as skilled as myself, sure, I could argue with my boss or exercise my powers, but that means risking a job loss. I'd be replaced by someone as capable, but with less inclination to argue. My 'wants' might include such things as, "being able to eat," and, "not sleeping on the street." Sure, it's possible to get overzealous with one's aspirations for physical goods, but I think that overlooks the authentic fear that many people have over being in a precarious financial situation.
This looks exceptionally helpful. I have around 50 tabs open, nominally, and my CPU fan is almost perpetually on high for it. Just before installing it, however, I found myself wondering, "Why am I using an app which allows me to keep more tabs open? Why am I not using bookmarks?"
Why do we keep open 500 tabs instead of using bookmarks these days?
That's an excellent question, and one I started asking myself about a year ago. My conclusion: it's a lousy band-aid over poor browser document and state management.
If I had a _better_ way of keeping track of what it is I'm currently accessing (or interested in accessing) than tabs, I'd use them.
A key problem is that browsers are keeping _every last tab fully open and rendered in the fear that you might need to immediately access is_. The tendency of websites toward full interactive designs exacerbates this problem.
Really: most text is pretty static. Generally I'm viewing one page at a time. Perhaps 2-5. The rest ... they can just go off and die.
Having them _saved locally in a state that's suitable for quick re-render_ would be kind of cool. Again, yes, this means that a lot of interactive page shit dies. But then, it should.
IMO webpages should earn their privileges. Including whether or not to animate, or play video, or audio, or display nonrelevant sidebars, headers, footers, and the rest.
The present browser / Web paradigm is grossly overextended and deserves to die.
I've never been to your subreddit before, but now I think I've found a blog I'll be keeping/catching up with for the next few weeks.
I agree with you in general. A while ago I tried thinking of some other paradigm to keep track of webpages I might want to get back to in the near future without the permanence of bookmarks. Some of my thoughts:
Web browsers are used for multiple activities that have their own "stack" in my mind. Often I find myself opening a whole set of tabs that are all related to a general activity I am doing. For example, I'm be doing some coursework and open a million tabs (man pages, tutorials, etc.), then I close my laptop and go to bed. The next day it's class time and I just want a clean tab pane to focus only on what is being discussed in class. Then maybe the next day it's the weekend and I'm relaxing, reading some blogs, going on social media, etc. The point is that every time I use my browser, I might not be performing the same overall activity, and I don't want to be distracted by what I did hours ago.
I don't want to lose information on what I was doing earlier, though. I want to be able to come back to the particular state of my browser as it was when I was last doing a certain activity. Let's say I have Tuesday/Thursday classes, and on Thursday I want to return to where I left off on Tuesday. I need a way to get back to that state without what I may have done on the browser between then and now to get in the way.
Bookmarks don't cut it for me in this regard. Bookmarks only work for a single webpage. But maybe I have a whole group of webpages that are all inter-related! Using the bookmark model we have now, I'll have to bookmark each page individually, then open them up individually when I want to return. There are add-ons that bookmark tab sets, but I think that UI needs to be a core part of browser UI, not wedged into a browser extension.
Right now, I think the most useful way to manage tabs (for me, YMMV) is some combination of tab pinning, tab groups and tree style tabs, and tab suspension/delayed loading. Each of those has a particular advantage:
Tab pinning (as seen in Firefox and Chrome) is useful to me. It lets me identify single tab activities (usually web apps) that I always want to have available to me because they are commonly accessed. Facebook, email, Youtube, Calendar, etc. It is useful to have a muscle-memory location on my tab pane that I can quickly switch back to those common activities.
Tab groups are one of my top favorite features in Firefox. It so naturally fits my particular workflow where I might be switching browsing contexts on a daily basis (class, work, relaxing, Wikipedia bingeing, blog reading, ...). The main brilliance behind tab groups is that they act as a less permanent form of a bookmark group. I can easily switch between contexts in a context overview mode (Cmd/Ctrl+Shift+E), and all of my contexts are clearly delineated. All I have to do to organize the contexts is to think "I am starting a new activity" and open a new tab group. However, I only need to close a tab for it to be gone from my mind.
Tree Style Tabs does this pretty well too, but for some reason I just found the UI unintuitive. Plenty of other people like it so I won't knock it. It does do a good job of organizing tabs by activity.
Another style that I thought was interesting is tab stacking as seen in Opera[1]. I think it is an idea not mutually exclusive to tab groups.
I also really like Firefox's delayed loading of tabs on startup. It fits perfectly with the concept of tab groups because it automatically means that other tab groups won't load up, and little memory and CPU cycles are wasted on them. Right now I have 81 tabs "open" in various tab groups, but Firefox has only loaded the ones I clicked or opened since starting.
Where things can improve:
1. Memory management/unloading of old tabs
I really think if Firefox wants to embrace the idea of tab groups, it needs to clear out tabs that haven't been interacted with for a long time. That would bring the best of all worlds: only the tabs I care about being open are open, and the tabs (and tab groups) I care about (with associated states, tab histories, and page scroll placement) are still preserved. Some add-ons like Bartab[2] are a good step in this direction, but it would be nice to see this put into mainline Firefox (with a preference to adjust the time limit to, say, infinity).
2. Tab group classification or tagging
Tab pins also could be more useful if they had some way to classify them based on importance. Currently, if you pin a tab in Firefox, the pin is available within all tab groups. However, I think the idea can be expanded so that you can:
* Pin for all tab groups
* Pin for a certain subset of tab groups (e.g. YouTube pin only on groups categorized/tagged with "relax")
* Pin for a single tab group
3. Tab stacking
As I said earlier, tab stacking is available in Opera, but I'd like to see it in other browsers too. It's another way to save on tab bar space while also grouping tabs by concept. One could use tab stacking and tab groups and pinning at the same time.
I think this idea would particularly match the use case of having a web application like a Google Doc open while also doing some research: create a new tab group, pin the document tab to just a single tab group (so that it is permanent and also takes up less horizontal space), and then go about browsing as normal with the rest of your tab space. Similar tabs (e.g. tabs on the same domain) can be stacked, and switching tab stacks automatically maximizes one stack while minimizing the others.
4. Open the browser to the tab overview page
This one is pretty self-explanatory. I think opening into a tab group overview is a better way to deal with the concept that a browser is used for multiple activities, and it may not be the same (or last) activity on startup every time.
These are just wishful ideas that I have. Firefox does the best job for me right now, I can't even use Chrome for my workflow because it just doesn't handle >50 tabs well enough.
...I've found a blog I'll be keeping/catching up with...
Thanks :)
Web browsers are used for multiple activities...
Something I've concluded Web browser developers put insufficient thought into. They're also far too often optimizing for content providers and advertisers rather than readers.
...a whole set of tabs ... related to a general activity...
Quite. And I'll often have multiple projects going on, each with their own set of related references.
The failure of the Desktop (and GUI) paradigms in general to reflect this is a growing and tremendous frustration of mine. Our activities are organized and grouped by application and increasingly by Web silo rather than by task. It's where the UNIX shell philosophy is increadibly powerful, and the abilty to both create different "workspaces" (generally: directories) for projects, and to string data through commands and pipelines, is phenomenally useful.
Just to highlight a frustration: increasingly desktop tools are utilizing Web tools for documentation. I arrange my work by Workspace. So when I fire off the Help command in a particular utility ... a browser tab opens somewhere three workspaces away, that I'm not aware of. Often in the midst of a bunch of other unrelated tabs....
For example...
Excellent use cases, and while my own work differs in details, the general outline is the same: different tasks, often persisting across days, weeks, months, or even years, but alternating with time, which I wish to group. Often quite complex and involving many related sites and/or pages. And retaining that user state is at an extreme premium.
Bookmarks don't cut it for me in this regard....
I'm leaning toward a middle ground between Bookmarks, History, and some level of annotation. Readability is an interesting experiment but ultimately frustrating and limiting.
My own curated lists of stuff are somewhat useful but extremely tedious to create. My subreddit is as much a bookmarking tool (of both Websites / references, and my own thoughts) as it is a publication and discussion.
The ability to create and readily access sets of bookmarks, tag and tab them, filter them usefully (by tag, author, site, dates, etc.) would be tremendously useful.
I'd really love a browser feature which let me search for text within pages I'd recently visited (hour, day, week, month, ...). Often what I'm looking for is something I've already recently visited.
It is* possible to organize Bookmarks hierarchically, etc. But I find that the UI/UX for doing this in numerous browsers is simply tedious beyond description.
...some combination of tab pinning, tab groups and tree style tabs, and tab suspension/delayed loading...
Yes to all of this.
For Pinning, I'd prefer to simply break out Web Apps as their own local app instance. Not associated with a browser at all. While Web-as-App-delivery is interesting, it's ultimately highly frustrating in my experience. A neither-man-nor-beast experience.
Tab groups...
I'll need to play with that.
Tree Style Tabs...
I use and love the plugin (for Firefox), and use it much as you do Tab Groups from what I can tell.
...tab stacking...
Not familiar with that.
...delayed loading of tabs...
Genius.
1. Memory management/unloading of old tabs
2. Tab group classification or tagging
4. Open the browser to the tab overview page
A-fucking-men to all of this. Your point that there is no overview page for tabs, or similarly that there is not way to perform group operations on tabs, is another massive failing.
I'm still fighting to keep my number of tabs down. It causes significant mental overhead and task switching costs. I've noticed a bump in my productivity by taking the time to close/bookmark/onetab tabs.
Agreed. OneTab is a great way to organize your "sessions" of browsing. If you're like me, you keep separate windows for separate topics and it's great to OneTab them all when something needs to be put on hold.
I do realize it's glorified bookmarking, but the concept and ease-of-use make it nice.
Because maintaining a large catalogue of bookmarks is a nightmare; good luck finding anything unless you are incredibly meticulous about organizing them. In addition, I think it's a spatial locality thing. I find that I'm a lot more likely to come back to an article/page/etc. if it's still open than I am to remember that I bookmarked it and wanted to come back later.
From you description, you do sound kinda' like a bot. Disabled cookies. Disabled Javascript. Irregular searches. I understand the frustration with saying, "You have to have these features supported to use the product," but let's face it: providing an experience to people who deliberately disable huge chunks of browser functionality is a tremendous pain in the ass. I think I can understand both sides of the argument using different strawmen:
"Can I read this paper, please?"
"Yes, of course, just put on these reading glasses."
"Why do I have to put on the reading glasses?"
"Well the font is quite small. If you don't wear the glasses, you probably won't be able to make out anything on the page. Your experience will be seriously degraded."
"I don't want to wear the glasses. Why can't I just read the page?"
"Well, we can fit a lot more data and make the page more robust by printing the text smaller. Why don't you just wear the glasses?"
"I have concerns about the glasses. I'd rather strain my eyes."
"We're not going to make a special page for you when 99% of the people are totally okay with wearing the glasses or wear the glasses anyways."
It's a tricky situation because you have to balance being accepting of people with non-traditional backgrounds while also making sure they can do the basics. I liked that TripleByte could condense DAYS of interviews into a few hours.
In general, long tech screens are hard on people that can't afford to take off a full day, so it's tricky for people already working in crushing jobs that need a change. The TripleByte quiz worked nicely for this.
On the flip side, if their tech screen is filtering people for reasons other than ability, it's removing possibly good candidates who deserve a shot at something.
The new approach is sounding like it's trying to satisfy the latter, which is good for improving inclusivity, but won't be to everyone's liking.