Computer Science Encyclopedia Can Fill a Gap | Hacker News Reader

Computer Science Encyclopedia Can Fill a Gap

cacm.acm.org

80 pointscarlehewitt29 comments

It's true wikipedia can be hit-or-miss sometimes, but in the STEM fields it's mostly solid, so not sure why the need to start a new one from scratch. Maybe just make some initiative to fix existing pages and make them citable?

http://www.scholarpedia.org/ has nice articles written by experts in the field. Usually pretty good and thorough reviews.

I think I remember Peter Thiel talking about that the main value proposition of Facebook was the real identities of people. This wants to produce an encyclopedia using the real identities of people. So it's looking for some combination of Wikipedia, Stack Overflow and Quora [0].

[0]: https://www.quora.com/Do-I-have-to-use-my-real-name-on-Quora...

The main issue with Wikipedia is that its main focus is providing references rather than being spot-on accurate[1], which is more useful for subjective and cultural subjects where the objective accuracy can not be trivially determined. For a STEM-focused platform, it would be beneficial to have a process where articles can only be edited if sources are provided (so there would never be a "citation needed" tag).

[1]: https://en.wikipedia.org/wiki/Wikipedia:Researching_with_Wik...

In computing and computer science it is regularly appalling, even on basic topics.

The article on wait() et al., for example, is significantly wrong, some of which is called out on its talk page.

* https://en.wikipedia.org/wiki/Wait_(system_call)

There are similar errors in its article on MAC times.

* https://en.wikipedia.org/wiki/Talk:MAC_times

Its article on pax, a standard Unix utility for a fairly long while, is very misleading too.

* https://en.wikipedia.org/wiki/Talk:Pax_(Unix)

Its article on CubeHash in 2013 did not include information from 2009.

* https://en.wikipedia.org/wiki/Talk:CubeHash#This_article_is_....

I've pointed out errors in its article on systemd, one of which was called out by one of the systemd authors.

* https://en.wikipedia.org/w/index.php?diff=723240462&oldid=72...

* https://en.wikipedia.org/w/index.php?diff=631112520&oldid=63...

The irony of this edit to the NFS article is that in fact the problems and limitation of NFS with respect to full Unix filesystem semantics were called out by its authors when they first wrote about it, and are the widespread view.

* https://en.wikipedia.org/w/index.php?title=Network_File_Syst...

Wikipedia people actually tried to delete articles on "Is" functions, C++ placement syntax, and the Process Environment Block.

* https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletio...

* https://en.wikipedia.org/w/index.php?title=Is_functions&diff...

* https://en.wikipedia.org/w/index.php?title=Process_Environme...

They in contrast kept an article about a purported computer science concept that was invented within Wikipedia based upon a vague phrase in a book that did not in fact describe the Wikipedia concept.

* http://jdebp.eu./FGA/legacy-encoding-has-no-definition.html

* https://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletio...

So did you help make it better?

I haven't touched these topics. But in my experience, nearly none of my edits remained, despite providing credible and neutral sources for whatever I added. Usually, there are long-term Wikipedia editors who have put plenty of pages on their watchlist, and treat the pages as their own personal property. Every tiny change is treated as if it were controversial, and must be fought over in the talk pages. Beginners to Wikilawyering and WikiBullying lose these fights. "Anyone can edit" has become a joke.

I was going to say that computing and computer science topics are free of such silliness, and that clearly you were editing a far more fraught subject area such as American current affairs; but:

* https://en.wikipedia.org/w/index.php?title=Make_Compatible&d...

hehe yeah I've noticed that. They are the most boring arguments in the world, with those people, who with a life has time for that. Well, most pages aren't like that I think.

Wikipedia is really good, and if there is a mis surely more WikiBooks can be written, a separate Encyclopedia isnt required.

> ...supported by appropriate professionally-relevant advertising. ... The nonprofit professional Encyclopedia will be self-supporting through appropriate professionally-relevant advertising carefully curated for high standards using existing advertising programs.

It blows my mind that something which purports to be so important couldn't find a way to sustain itself without commercial advertising. This isn't about improving the status quo of CS education; it's about lining the pockets of a handful while reaping the rewards of volunteer experts.

<quote> There is an important gap in Computer Science education and professional collaboration that can be filled by an nonprofit online reputable, referenceable Encyclopedia... </quote>

The author spends a lot of words envisioning the encyclopedia, but says nothing about what gap needs to be filled.

Uhm... Why? I don't see any gaps here. And the potential ammount of work is immense.

> Over time, the Encyclopedia should be organized using ontological services supporting programmatic interfaces for a knowledge graph.

And this phrase shows how quickly this encyclopedia will become a memorial to itself rather than a source of up-to-date knowledge.

The Stanford Encyclopedia of Philosophy may be a decent model for this. Especially for the more conceptual / core areas of CS.

Link for convenience [1]. I have yet to find anything on par with the Stanford Encyclopedia of Philosophy (SEP) for other fields. It truly is an amazing, expert-curated project.

If anyone else has an interest in the semantic web or ontologies [2] in general, you might have a look at the Indiana Philosophy Ontology project [3]. It works to use data from SEP and other sources to build a ontology of the field.

[1] https://plato.stanford.edu/

[2] https://en.wikipedia.org/wiki/Ontology_(information_science)

[3] https://www.inphoproject.org/

I had a nightmare once about something like this. In the dream, I woke up, went to work, wanted to check something online, found out stackoverflow no longer existed.

Is it possible we might put too much trust into this one source? I know it is amazing. I know we love it. But something like this could help.

An open-source and non-profit version of something like StackOverflow would be nice. Dependence on one commercial company is risky. The now-defunct C2 wiki had interesting discussions about software engineering, but couldn't really deal with specific products or languages. Such a tool(s) would probably need to be split or marked to distinguish between formal research, philosophical discussions (such as whether to optimize for programmer heads or machines), and specific languages and tools.

«The Encyclopedia should be managed by a prestigious Editorial Board which appoints a hierarchy of editors to moderate articles. [...] Serving as a member of the Editorial Board could become a prestigious office for senior professionals to provide their experience and judgment»

Sounds a lot like the current peer-reviewing system and journal editors mafia. A bunch of old men with their hand on the system and doing the maximum so it does not change, for their own interest of keeping their position and dominance of the system.

Mr Hewitt has apparently disrupted Wikipedia [1] with excessive self-interested and self-promoting edits and so is not someone who should be involved in running a new Encyclopedia

[1] https://www.theguardian.com/technology/2007/dec/09/wikipedia...

Please don’t forget to launch a PDF/ EPUB version as well. Similar to Usenix’s full proceedings.

I've always wondered why something like this doesn't yet exist.

Something like a Wikipedia of man pages, programming language documentation (and official tutorials), answered/archived Stack Overflow questions, etc. A central (but maybe decentralised) collection of manuals and reference sheets and all the rest.

So often you go to find something and the site has expired because they gave up on the project (but it's still being used), or the forum posts have been deleted, or it's impossible to find because Google has decided you actually wanted to find <insert something unrelated> and they know better than you do.

An Archive.org for tech info, with machine readable formatting so we can have a comprehensive search function.

Then throw in a Wikipedia-style packaged archive you can download for Internet-free local searching and working while traveling or whatever.

According to [1] Carl Hewitt, the guy advocating for this encyclopedia has

disrupted Wikipedia for more than two years by using it for self-promotion, tampering with his own biography and manipulating computer science articles to inflate the importance of his own research.

How would be assured that the members of the "editorial board" of the encyclopedia wouldn't use it for their own benefit inflating the importance of their research, like allegedly Carl did?

[1] https://www.theguardian.com/technology/2007/dec/09/wikipedia...

Similar to Princeton Companion to Mathematics perhaps?

Yes, that's exactly what popped into my head. It is just at the right level for explaining subjects for someone who isn't a specialist.

This is what we're talking about for those not familiar: https://press.princeton.edu/titles/8350.html

All that stuff about registering with real names and having procedures for fairness and inclusivity already exists. M. Hewitt might want to learn about Citizendium and what has happened to it.

* http://en.citizendium.org/wiki/CZ:Policies

They should follow the model of The Stanford Encyclopedia of Philosphy, which is accessible online for free.

Too much talk of prestige and reputation, perhaps unsurprising since it's from a professional organization.

> The Encyclopedia must establish procedures to be fair and inclusive on the basis of race, sex, religion, age, disability, and national origin.

No, first and foremost the Encyclopedia must establish procedures to ensure high-quality, accurate, and concise content. Being "fair and inclusive" is a secondary concern. If it has no users, no one cares whether it is "fair and inclusive".

You've misread the article. It never said such procedures must come before everything else. What it says is entirely compatible with what you want.

Ideas don’t matter much, execution is far more important and much harder.

Reply on news.ycombinator.com