Hacker News Reader

Comment by Dangeranger | Hacker News Reader

If you are interested in reading the original paper without the annotations, as I was, you can find it on Alan's Viewpoints Research Institute site here [0].

[0] http://www.vpri.org/pdf/tr1984001_comp_soft.pdf

View on HN

Thanks, much appreciated! Prefer this to all that JS non-sense.

Er, can't you just ignore the annotations? They are in a different part of the screen and don't modify the original paper. Not affiliated with this website but wondering why you felt the need to seek the original paper from a different place when it's clearly right there...

Not the parent but when you click it slides out a sidebar to show the discussions, covering the content. The annotations also take up a lot of space making the text smaller.

On the plus side they (Fermat's Library) converted it to text, where the PDF versions are all image scans.

I'm new to Fermat's Library, but I'm perplexed by what they've done. They appear to have OCR'ed the paper, and it appears that you can select a region of text, but if you copy that region and paste it, it consists entirely of garbage characters. Similarly, they've somehow broken CMD+f.

Looking at the HTML source, it appears they've used pdf2htmlEX: https://github.com/coolwanglu/pdf2htmlEX / http://coolwanglu.github.io/pdf2htmlEX/ -- the first demo there (which is also from a scan) also seems to have the same property you mentioned.

Whenever I read material from a syndicated site I want to know where the canonical source is. This is essentially an electronic transcription, and it may not be 100% correct.

In addition I wanted to download the material, and Fermat's Library made that difficult by design.

Also I disable JavaScript for pretty much every website, and don't really like when I have to selectively re-enable scripts in order to achieve basic functionality, as was required for Fermat's Library.

Thanks for explaining. From previous looks at this website, I did not consider it an electronic transcription but an embedding of the original PDF; a closer look reveals some differences in this case (the image at the side of the first page has disappeared, and the subheading / pullquote at the top of the first page has been retypeset somehow). Apart from that, all the line breaks and character positions seem to match, so it does appear to be identical to the canonical source. The other reasons make sense (it didn't occur to me that it may not display properly; my understanding of their technology was that they simply took the original PDF and added dots as markers; I hadn't noticed that it wasn't downloadable).

Not displaying properly for me.

Reply on news.ycombinator.com