Scribd Had A Blowout Year, And So Did the Web Document • Hacker News Board

Re: Scribd Had A Blowout Year, And So Did the Web Document

#21

Post by zain » Thu, Jan 01, 2009, 12:49 AM UTC

axod wrote: post #12

Mazy wrote: post #10

Earlier quoted context omitted.

I highly doubt comscore is claiming to be tracking embed/widget traffic. Those numbers are supposed to be for the site itself. They are quantified, and quantcast shows the vast majority of their traffic is to their destination: http://www.quantcast.com/scribd.com see scribd.com vs The Scribd Network

True, so people are linking instead of embedding. But the point is the same... If one person makes the decision to link to scribd instead of the pdf, and 100 people follow that link, it doesn't really mean 100 people value the service.

The traffic isn't just from people following links. What about the people uploading 50,000 documents every day?

Source: Scribd says so, according to washingtonpost http://www.washingtonpost.com/wp-dyn/content/article/2008/12...

Re: Scribd Had A Blowout Year, And So Did the Web Document

#22

Post by mattmaroon » Thu, Jan 01, 2009, 12:50 AM UTC

axod wrote: post #18

marketer wrote: post #15

Earlier quoted context omitted.

Have you ever thought about using flash for your web chat service? I'd actually prefer it over standard javascript, especially because it would have a real socket connection. Also, flash has pretty good support for text rendering. You can embed any font into any swf file, and it has anti-aliasing and effects. There was a presentation about the new flash 10 text features recently - http://is.gd/ejKj

From a user point of view, why do you care if it has a real socket connection, or a emulated connection that does exactly the same thing? I'm not sure any user anywhere would notice the difference. There is no noticable difference in terms of latency, negligible difference in terms of bandwidth. I have thought about the possibility of using it where available, but I don't think it'd really be worth the effort and pot…

The problem with HTML is there's a decade lag between when the new font rendering is introduced and when you can rely on it to be accessible to all of your customers.

Re: Scribd Had A Blowout Year, And So Did the Web Document

#23

Post by mattmaroon » Thu, Jan 01, 2009, 12:52 AM UTC

axod wrote: post #9

zain wrote: post #6

Man, you guys are harsh. Even if it isn't a service you yourself would use, why bash a service useful for others? The first goal of many Y Combinator startups is to make something people want, and Scribd clearly fits the bill if millions of people are using it.

I disagree. ONE person decides to use scribd to show some text on techcrunch, and suddenly the whole readership of techcrunch are "users" of scribd? Measuring success in terms of pageviews in a passive widget isn't a good measure.

How about the 50k documents uploaded every day? Can you really question that Scribd is a product that a lot of people want to use?

Re: Scribd Had A Blowout Year, And So Did the Web Document

#24

Post by axod » Thu, Jan 01, 2009, 12:53 AM UTC

marketer wrote: post #17

hapless wrote: post #4

I realize that someone, somewhere, is running webtv and would love a service like Scribd. That person is not able to use it because it's a flash and javascript nightmare. The rest of us, who have regular PCs with the modern software that scribd requires, have document readers for all of the various file formats. The very worst of these readers is faster and more pleasant to use than the best scribd rendering. Lastly,…

About the speed issue - how is a 20 MB reader (Adobe Reader 9) plus the cost of downloading the entire pdf faster than a 200 kb swf file and having the document streamed? That makes no sense.

Half of the "documents" seem to be either plain text, or a plain image (porn).

The most viewed "document" on scribd of all time: http://www.scribd.com/doc/26896/SUING-FOR-TOO-MUCH-SEX

Do you think it's sane to "stream" that document? What the hell are these people thinking??? They're wrapping plain text inside flash. One can only imagine their warped mentality. Why not convert plain text to a video, upload it to youtube, and embed the video. It'd make about as much sense.

Re: Scribd Had A Blowout Year, And So Did the Web Document

#25

Post by abstractbill » Thu, Jan 01, 2009, 12:56 AM UTC

axod wrote: post #24

marketer wrote: post #17

Earlier quoted context omitted.

About the speed issue - how is a 20 MB reader (Adobe Reader 9) plus the cost of downloading the entire pdf faster than a 200 kb swf file and having the document streamed? That makes no sense.

Half of the "documents" seem to be either plain text, or a plain image (porn). The most viewed "document" on scribd of all time: http://www.scribd.com/doc/26896/SUING-FOR-TOO-MUCH-SEX Do you think it's sane to "stream" that document? What the hell are these people thinking??? They're wrapping plain text inside flash. One can only imagine their warped mentality. Why not convert plain text to a video, upload it to yout…

This is the modern version of people emailing each other Microsoft Word documents with photos in them. I never understood that either.

Re: Scribd Had A Blowout Year, And So Did the Web Document

#26

Post by abstractbill » Thu, Jan 01, 2009, 1:01 AM UTC

axod wrote: post #18

marketer wrote: post #15

Earlier quoted context omitted.

Have you ever thought about using flash for your web chat service? I'd actually prefer it over standard javascript, especially because it would have a real socket connection. Also, flash has pretty good support for text rendering. You can embed any font into any swf file, and it has anti-aliasing and effects. There was a presentation about the new flash 10 text features recently - http://is.gd/ejKj

From a user point of view, why do you care if it has a real socket connection, or a emulated connection that does exactly the same thing? I'm not sure any user anywhere would notice the difference. There is no noticable difference in terms of latency, negligible difference in terms of bandwidth. I have thought about the possibility of using it where available, but I don't think it'd really be worth the effort and pot…

The real place I could see Flash being great for mibbit is it would allow you to build an embed that could be used on sites like Myspace which allow Flash but not Javascript content.

Let me know if you ever want me to show you how to get a sane Flash dev environment set up - I have the war-wounds from writing most of Justin.TV's Flash code ;-)

Re: Scribd Had A Blowout Year, And So Did the Web Document

#27

Post by axod » Thu, Jan 01, 2009, 1:05 AM UTC

mattmaroon wrote: post #22

axod wrote: post #18

Earlier quoted context omitted.

From a user point of view, why do you care if it has a real socket connection, or a emulated connection that does exactly the same thing? I'm not sure any user anywhere would notice the difference. There is no noticable difference in terms of latency, negligible difference in terms of bandwidth. I have thought about the possibility of using it where available, but I don't think it'd really be worth the effort and pot…

The problem with HTML is there's a decade lag between when the new font rendering is introduced and when you can rely on it to be accessible to all of your customers.

True, although I think things are speeding up. Chrome, firefox, safari, emergence of the iPhone as a browser...

HTML5 is definitely heading in the right direction though. WebSocket will be awesome when it arrives, making comet a thing of the past. and tags will pretty much put flash out of business, etc etc

Re: Scribd Had A Blowout Year, And So Did the Web Document

#28

Post by axod » Thu, Jan 01, 2009, 1:08 AM UTC

zain wrote: post #21

axod wrote: post #12

Earlier quoted context omitted.

True, so people are linking instead of embedding. But the point is the same... If one person makes the decision to link to scribd instead of the pdf, and 100 people follow that link, it doesn't really mean 100 people value the service.

The traffic isn't just from people following links. What about the people uploading 50,000 documents every day ? Source: Scribd says so, according to washingtonpost http://www.washingtonpost.com/wp-dyn/content/article/2008/12...

True, although I'd guess a fair amount is webmasters trying to cross promote their content. Seems like if you own a porn site, you may as well post some "tasters" to scribd with your web address on, etc etc

Re: Scribd Had A Blowout Year, And So Did the Web Document

#29

Post by axod » Thu, Jan 01, 2009, 1:15 AM UTC

mattmaroon wrote: post #23

axod wrote: post #9

Earlier quoted context omitted.

I disagree. ONE person decides to use scribd to show some text on techcrunch, and suddenly the whole readership of techcrunch are "users" of scribd? Measuring success in terms of pageviews in a passive widget isn't a good measure.

How about the 50k documents uploaded every day? Can you really question that Scribd is a product that a lot of people want to use?

From a quick look at the "most recent" page:

  20 docs per page, 50 pages covers last 18 hours.
  =1,000 docs in last 18 hours
  =approx 1,333/day

Maybe that page isn't showing every document, but I don't see why it wouldn't. And sure, it's new years, but even so...

Also you can't deny its used by webmasters promoting their websites....Posting endless tips etc with their link at the bottom. Take a look at some of the recently posted docs.

Re: Scribd Had A Blowout Year, And So Did the Web Document

#30

Post by justin » Thu, Jan 01, 2009, 1:20 AM UTC

Fuck the haters. Scribd is awesome because it's a great utility for easily sharing your shitty Word docs and PDFs, and there is obviously demand for it as Scribd and all the Scribd-like knockoff sites each have more users than your worthless websites. Complaining about Scribd putting text in Flash is like complaining about how ugly Myspace pages are: while you were out bitching, a few talented people were out giving the masses what they wanted, no matter how many standards got violated, and will subsequently get paid for it.