Hacker News new | past | comments | ask | show | jobs | submit login

Or, people could just put PDF files on the web, and search engines could index them.



Not everybody has personal hosting; sites like YouTube or Scribd are nice because they'll worry about the costs of DNS, bandwidth, storage, backups, etc without any charge.


Megaupload and most similar services are easier to use for file downloading than Scribd. Which does look like a benchmark of some kind...


I disagree; Megaupload, Rapidshare, Mediafire, etc are basically black holes in the Internet. Files go in, but what comes back out is not usable.

First, most of them have wait timers. After opening their page, the user must wait 60-120 seconds before downloading.

Second, they're plastered with ads. And not Google text ads, but awful '90s-style popups / popunders / flashing seizure GIFs. Ad blockers are useless, because the sites are written to prevent the download link from working when ads aren't visible.

Third, most (all?) of them don't handle Unicode correctly. Upload a file in anything except US ASCII, and you'll be lucky if even the file extension comes across intact. For example, Преве́д.zip could come out "Ð�Ñ�евеÌ�д.zip" or "Преве́" or nearly anything else.

Lastly, sites like Rapidshare are known for imposing download limits, such as "2 files per hour" or "max 50 Kb/s". This is hugely annoying when trying to download a half-dozen 30MB files.


Depends on a site, but taking MU and going point by point:

- first - wait timers: Megaupload has a timer which takes ~1min. to count down - that's quicker for me than dealing with scribd's registration.

- second - ads: I haven't noticed, I use adblock and never seen an ad on Megaupload. Not sure what's the current status of Scribd, but disabling scripts brought out a lot of spam words years ago (maybe they don't use it anymore), so annoyance is pretty much the same (or isn't there, depending on your view)

- third - unicode: We're talking about pdf-s. They have their own tags, for example for the title which makes the name pretty much irrelevant. Unless you're downloading a book with "Author - Title.pdf" name, you're most likely going to run into "some_serial_number", "ModelNumberDescription_code" for manuals, "thesis.pdf" or some other internal convention. Actually zipping the file allows you to see the original one.

- fourth - limits: Megaupload has none that I know of and I get >200kB/s most of the times.

So it might be very subjective, but even considering all the crap download pages give us, I'd take a Megaupload-ed .pdf over Scribd any time.

+ After writing this, I got back to scribd and tried to download something - it's MUCH better than it was before. I only had to scroll through 2 pages and find a small link at the bottom and it opened a new window. Right now, for me it's only as bad as MU.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: