Extension for Safari

not-planned

#1

Paperpile is great, but Chrome has steadily evolved into more and more of a bloated mess on Mac OS. It is slower, burns far more ram and idle CPU, and literally cuts laptop battery life in half relative to using Safari. I would really like first-class support for Paperpile in Safari.


#2

Desperate for this also. The standard answer from Paperpile is that Chrome is simply the only browser that can support their product. As a developer myself I can see that this is simply not true.

It just makes the user experience for portable users substantially worse when using Chrome because it’s such a resource hog.

Please guys. Invest in Safari! I would pay 10x more for Safari support.


#3

If you look through the forum, you’ll see that we try to answer all questions individually and completely open and honest. Of course I’m a bit disappointed to hear that despite these efforts we apparently come across as giving “standard answers” that are not true.

I’ll try again for this particular topic:

  1. Supporting Safari would also mean supporting Firefox because both have appr. the same market share (measured on potential users visiting our website — not existing users who are by definitiion 100% Chrome users).
  2. Their market share is extremely low at about 10% for Safari and Firefox. Edge is about 2% at the moment but will probably reach the same level soon and we would also need to support that.
  3. We never said Chrome is the only browser that would support all Paperpile features, we only said that it requires an extension since it uses many APIs that are not part of any web-standard. That means we would need to develop and maintain 3 different extensions. It’s safe to say, that if we had tried this initially there would be no Paperpile as you know it. If we tried this now, there would be no new features for the foreseeable future (no Android app, no iOS app, no Word plugin,…). Instead of winning a few Safari and Firefox users we would risk losing our very happy Chrome customers.

In short, supporting Safari would put the future of Paperpile at risk and that’s why we have not plans at the moment to go in this direction.

Paperpile is available for Chrome and very soon for iOS and Android (and next year most likely also for the Desktop). Our goal is to provide a great experience on all these platforms and that unfortunately means we can’t support all browsers.


#4

Hi Stefan,

Thanks for the detailed response.

I come from an environment where almost every scientist in the department is using a Mac. Hundreds of us. We are not writing papers using iPads, or iPhones (and never will be, very hard to write a paper on such a device). So we all, have to install Chrome in order to use Paperpile. When we have a new student start work in the department, we recommend Paperpile, and then have to tell them to install Chrome.

We all suffer 2-3 hours battery life loss because of using Chrome purely because we need Paperpile. Paperpile is probably the best referencing software we have ever used. It is beautiful, seamless, worth way more than currently charged by the mainstream competitors. But the critical problem for us is it is forcing us out of a heavily Apple ecosystem (no continuity, no handoff support with Chrome, no synced keychain, bookmarks with our iOS devices) purely to use this one website: Paperpile. It costs us in terms of battery life – which is important – especially to our students.

For as long as you keep forcing us to use Chrome – we will have to – because you have such a breakthrough product. But please consider the market share of Mac products in academia.

In my view, focusing on iOS over Safari integration is flawed. Myself and colleagues lament the iPads we purchased in the hope of reading and writing papers on them, only to find no good referencing software exists (sounds like you’re about to solve that problem). Even with referencing software, the multitasking support, poor keyboards, poor integration for graphing and stats software we would need to write papers make the iPad a substandard device.

I think if you look at academic papers written at the top 25 institutions in the UK, you would find that 60-70% of them are being written on Macs – Macs that do not have Chrome installed (unless, like us, we have to in order to use your great Paperpile software!).

Best wishes,
Oliver


#5

Stefan, I just wanted to say thanks for all of your efforts in developing Paperpile. For those of us who are fortunate enough to use desktop PC’s I can safely say that we don’t suffer any loss in battery life :horse:


#6

Interested in the plans for a desktop app. First I’ve heard about this. Wondering if it might not be possible to support Firefox and Safari by having a simple “open link in desktop app” extension?


#7

I am not sure I understand the resistance to use Chrome. While I have observed that Macs are more prevalent than usual in academic environments, Chrome is avaialble on Macs, and by almost any test benchmark - Javascript speed, HTML5 compatibility , extension compatibility etc, it is technically superior to Safari.

The battery life issues are solved in Chrome version 53 on Mac : https://9to5mac.com/2016/09/06/google-chrome-mac-battery-performance/

I am wondering why are you and your colleagues using Safari at all ?

That said, I am also perplexed about some of the decisions Paperpile makes - for example, the Desktop app is going to be a Mac only beta to begin with. Given that it is built on Electron.js , it should be cross-platform already. Why is then the beta Mac only?

Another decision Paperpile made is their extremely tight integration with Google Docs, which no serious academic can use to write papers. Word and Libreoffice integration would have much more “sell-able”.


#8

I’ve been a stuborn Safari user for many of the convenience-factors listed, but the battery-life issue was my trump card when I’d get to arguing with friends about it. But Nachiket is right - the battery problem is basically history and I now run Chrome all day with 30+ tabs with no noticeable difference in battery drain.

Password sync with my iphone still gets me down, but many people avoid this partially by doing all personal work in Safari and all professional work in Chrome. Not a bad idea to keep separate anyway. Another option would be to use a 3rd party password manager such as Lastpass or 1password.

Another tip: Use “cross share Air Drop” Chrome extension if another complaint you have is that Chrome doesn’t support Airdropping URLs.

And most importantly, Safari used to be a lot prettier than Safari, but it’s gotten some recent visual refinements which make it almost pretty. :smile:


#9

This is becoming a really great discussion. Thanks @Nachiket_Vartak for the link. It’s great timing with the latest release of Chrome 53 just a few days ago. It seems to have addressed the battery issue.

More generally, I have to say that we are not just building a product we are building a company. And to have a great product in the long term you need a robust company. So clearly business decisions go into the product decisions.

Some decisions are hard because there are always many options you can choose from. However, some are really easy because the data is so clear.

@Oliver_D: You are right many academics are using Macs but that does not change the fact that only 10% of all our potential users use Safari to begin with. I don’t know why that is, but this is what we see.

Second, mobile is not for writing papers is for reading papers. I’ve checked, in total we got 7 requests for Safari in the past 3 years. We get multiple requests for mobile every week. So the demand is 1-2 orders of magnitude higher for mobile apps than for Safari. That’s not us hating Safari that’s just the market we are operating in. If it was the other way round we would release a Safari app in October instead of an iOS app.

@Nachiket_Vartak Technical decision are based on so many factors that it’s hardly ever possible to judge them from outside. You are right, Electron is cross platform.

But that’s just the basic runtime there is lots of stuff going on which needs to be built specifically for macOS and Windows. Word for Mac and Windows are extremely different and we need native modules in Swift and .Net to communicate between Word and Electron. Packaging and updates require completely different server components. The UI paradigms are also different. We use the task bar on Mac as central UI element because it’s visible all the time. But on Windows it’s hidden on the bottom of the screen. Also we have been a Linux/Mac shop until now and we have to establish our development process on Windows. So with this information I don’t think anyone would be perplexed that we ship our Word plugin on Mac first.

Another decision Paperpile made is their extremely tight integration with Google Docs, which no serious academic can use to write papers. Word and Libreoffice integration would have much more “sell-able”.

We have many thousands of academics who have left Word for Google Docs and Paperpile. They would not be happy to be called “non serious academics” :wink:

And they are all paying customers and we fill a quickly growing niche. So I would strongly disagree that Word or Libre Office are more “sellable” in the current market. Actually our Word plugin will be free like our sidebar Docs add-on is.


#10

I don’t want to give an official confirmation yet, but we will release our Word plugin soon, which is basically a desktop app and eventually will run offline. Adding the actual app on top of that is not that hard. But we have to wait until the Word plugin is out to see how it behaves in real-life.

And yes, you are right, once we have a desktop app we could make a lightweight browser extension for Firefox and Safari.

Again, I can’t say if we actually go in this direction but from the experience with the Word plugin so far it’s very likely.


#11

Perhaps not the best choice of words - but let’s say “career academics”. All scientific staff in Germany has to sign the https://en.wikipedia.org/wiki/Bundesdatenschutzgesetz , the Data Protection Act, which explicity prohibits putting data or informtion that belongs to the organization on third-party servers (such as Google Docs). As such, using Google Docs for any serious work is out of the question.


#12

We have two PhD students writing up their theses, and have had a few masters students do the same, using Google Docs. Works quite well actually. I myself have written multiple papers using it (+ Paperpile of course!).


#13

That’s interesting. Can you elaborate on that? Who has to sign what exactly? Your link goes to the actual law. Every country has some sort of data/privacy protection legislature. But I would be interested what you sign and if that is a national thing (as opposed to federal state or university specific agreement).

I’m from Austria and have worked closely with scientists from Germany for years. It’s the first time I hear about it. It would mean no scientist could use Gmail for their work (which I’m sure many do).


#14

Thanks for engaging the discussion.

There’s clearly large potential for bias in your data: you run what has always been a Chrome-only service. You can’t use your own visitor numbers to estimate your potential user base in the alternate world where you did have Safari support without some big error bars. Many of us visit in Chrome because we’re required to, not at all because we want to or because that’s our primary browser. Also, the “battery issue” has been getting “fixed” according to the chrome issue tracker for years now. They make small incremental progress, but so does Safari, and the difference remains stark in my experience.

Also, the notion that, because you see similar usage of Safari and Firefox in your (biased) data, if you support one you’d “have to support [the other]” does not follow. You can put your resources towards whatever you want (including exclusively chrome).

That said, I’d be very happy with mobile + native desktop apps, and a much more light-weight browser integration for Safari (and MobileSafari, I hope!) Indeed, I’d personally prefer this to the current large browser-based app for my own usage.

Please do keep in mind those of us who don’t use Word, Docs, or any other word processor, though. 100% of computer science, math, and much of economics, at the very least, are done typesetting our own LaTeX, which we write in text editors. The good news is that integration with this world can be far less engineering-heavy than with Word or Docs.


#15

I’m happy to engage in this discussion. One thing you have to be aware of, however, that this is not an academic discussion.

I’ve done my fair share of data analysis in my previous life as scientist. So I’m not unaware of the potential of any sort of bias in any data. It’s even more severe now because if we misinterpret our data we put our business at risk (instead of repeating an analysis for a revised version of a paper).

The potential selection bias you mention does not affect the overall results.

Also, the notion that, because you see similar usage of Safari and Firefox in your (biased) data, if you support one you’d “have to support [the other]” does not follow. You can put your resources towards whatever you want (including exclusively chrome).

You’re right. What I meant was basically that if we go beyond Chrome we would first support Firefox before we support Safari. Again that’s a no brainer mainly because it can also be used on Windows and second because it’s open source and many universities use it on their networks as officially supported browser.

Also, the “battery issue” has been getting “fixed” according to the chrome issue tracker for years now. They make small incremental progress, but so does Safari, and the difference remains stark in my experience.

It has to be seen but it seems that version 53 makes significant improvements. At least it does for me on my Mac Book Pro.

Finally, we will never forget LaTeX/BibTeX users. I’ve used LaTeX for more than 10 years exclusively while doing science. We are busy with Word at the moment because it’s tricky to get right but we will come back to LaTeX again with the goal to make Paperpile the best reference manager for LaTeX users.


#16

A Paperpile desktop app would make all the difference for me — I’d use it immediately.


#17

Awesome discussion here. @stefan thank you for being so transparent with respect to the expected development directions. I still hope you will consider developing for Mac-apps too. Personally I prefer to work with one system only, in my case macOS-based apps. One of the advantages being the good integration with iCloud, and the other with iOS. So, I’m still a bit reluctant to work with PaperPile - it currently disrupts a work-process which is completely macOS/iOS based.

That said, great to hear about the Word-plugin efforts. And I’m looking forward to the iOS and macOS app.


#18

I have to say that I am surprised to see the reticence toward Safari support. I work at a large R1 and do a lot of interdisciplinary work, and the top two browsers are far and away Safari and Chrome. Anecdotal evidence, I know, but even at conferences and such seeing someone running Firefox is rare. It seems to me that supporting Safari and Chrome would capture most of the academic market. Perhaps a desktop application will solve all of these issues.

Two further comments:

  1. Running a full version of a reference manager on iOS is quickly becoming a must have–the first company to successfully do that will have a distinct advantage. What is out there currently is focused on reading articles, but more and more people are using their iPads for so much more. The ability to handle (i.e. read, import, manage, and cite) ebooks, websites, pdfs, and to a lesser extent audio/video material on iOS would be huge.

  2. Word integration is a serious sticking point. I really do think that Paperpile is the best when it comes to gathering research and organizing it. When it comes to citing, though, it works great in Google Docs but doesn’t go beyond that. For a variety of reasons, many will simply not use G-Docs. It is a testament to the strength of Paperpile that I am using G-Docs as much as I do (which isn’t much). Most of the time, I use Paperpile to gather/organize material and then do something else when it comes time to write and cite. That works okay for now, but I can’t see doing that indefinitely.