How much online press coverage has the J# programming language received, as 
compared to the amount of press C# has received?  A ratio is fine.  If 
individual media sources generally tend to cover one but not the other, please 
note that, if time allows.

---

Clarification of Question by dotnet-ga on 22 Apr 2002 18:42 PDT

Note that Google finds the terms A# through G#, but doesn't seem to find J#, 
even though the term exists on many web pages.

Hello dotnet,

As you mentioned, searching for “J#” reveals nothing (on Google at
least), but as J# in not an actual language, per se, but more of a
tool for helping migrate Java developers to C#, searching for the name
of the tool (Visual J, minus the #) leads to many results concerning
J#.  Admittedly, you also have quite a few “Visual J++”, but if you
pare your search down to “Visual J –J++” you receive a list of many
results that do not contain J++.

After going through the first few dozen pages or so, I found two
things to be true: 1) There are a particularly large number of Chinese
and Japanese sites with articles on J#, far more than would normally
show up for a search on a popular Microsoft product, and 2) The ratio
of “press” articles to “other” (how to use, how to install, special
features of J# compared to Java, not J# at all, etc.) articles seems
to be roughly 40% press, 60% other.  If that ratio were to hold true
to the rest of the 629,000 articles, it looks like there are roughly
250,000 “press” articles about J#.  Actually, this ratio does not hold
true as you progress past 50 results.  It quickly drops
logarithmically to the point where results past 500 links give a ratio
of only 0.0001% “press” and the majority having nothing to do with J#.

Considering that we have lost at least a few articles by using “-J++”,
and that there are probably at least one or two articles past the 500
mark that aren’t being counted, it’s reasonable to just say that the
likely number of J# articles hovers somewhere close to 30-40.

Probably many of these are not unique, however, as news agencies tend
to share their stories.  That said, I found quite a few unique news or
news-like organizations in just the first few pages of results
including:

Microsoft
ZDNET
Infoweek
CNET
Chinabyte
CNN
@Market
Reachlive
Slashdot

…and many more.  But as long as we treat C# references with the same
rules as above, we should be able to produce a reasonable “layman”
style ratio (for a more scientifically accurate ratio, you would have
to individually download each of the 600k pages and compare them
against each other for duplicates and content related to J#).

As for C#, doing a quick google search on “C#” give us roughly a
million links, however you can quickly see that the majority of them
are not press links.  The majority are “how to” or similar.  That
said, however, the relevance of sites to the C# languages stays
constant all the way out to result 1000 (the highest that google will
return).  If even 1% of these articles were press related, that would
mean there are almost 10,000 articles on C#.  I went through the first
100 or so links, and I could only find 5-10 articles that could be
fully considered “press” (some were half press/half tutorial).  Most
likely that lost number (5-10%) is a result of having most of the
articles about C# being pushed lower in the list as time goes on. 
Still, even such a low number gives us more than enough articles
(50,000) to trounce J#.  I cannot believe that 50,000 unique press
reports would have been written about C#, as most of those would be
duplicates, and there must be a fall-off point somewhere that we
cannot see die to Google’s 1000-result max, but if we just assume that
the number of duplications is somewhat constant between the two
topics, and we assume that the fall-off is trivial, we still get a
ratio close to 1000 C# article to every 1 J# article.

As a control, I did the same calculation with Java:

23,500,000 pages, 98% “other” (Java the country, java coffee, Java
tutorials, Java applet downloads, etc.), 2% “press”, falloff of
relevancy not until far past 1000 results.  This comes out to roughly
500,000 press articles about Java.

Considering that Java has been around for about 10 times as long as C#
the ratio looks like it could be close to correct.  Even if there
aren’t actually 500,000 press articles about Java or actually 50,000
press articles about C#, I still feel pretty comfortable saying that a
1000:1 ratio for C# and J# sounds very close to accurate.

I hope you can use that!

I'm astounded at the quality of this answer!  If Google Answers
allowed tipping, I would definitely tip for this one.  Some serious
thought went into this answer, and some very useful techniques came
out of it.  Thank you!

Not much, AFAIK.  I've never even heard of J#, what is it?  Now I'm curious!  
Can you provide a link to some info it?

I ran this Google search:
://www.google.com/search?num=100&hl=en&lr=lang_en&safe=off&as_qdr=all&q=%
22J%23%22+%22.NET%22+%22java%22+-javascript

and it pulled up quite a number of links, most of which seem to deal with 
J#.NET.

Hope that helps,
researchmonkey-ga

"October 2001 : Microsoft releases a beta version of Visual J#.NET, a Java-esque
language embedded within Visual Studio.NET that lets developers use a Java-like
syntax in order to build applications that compile to the .NET CLR. The reaction
among Java developers thus far has been predictably underwhelming."
http://www.devx.com/free/hotlinks/2001/pointcounter102401/pointcounter_CPvsRJ.asp

Thanks researchmonkey... your search for

"J#" ".NET" "java" -javascript

is identical to a search for

j net java -javascript

which is why I'm posting this to Google Answers.  Try it.  Google removes the 
# and the . which makes searches on this subject area unnecessarily 
difficult.  Also, removing results with "javascript" actually removes good J# 
content, since J# is often mentioned along with javascript.  Sometimes you can 
get lucky when an author has used a "euphemism" somewhere on the page such as 
jsharp, j-sharp, dotnet, or dot-net, but the vast majority of pages don't see 
the need to use such search-friendly euphemisms.

try j sharp to find results in google!
afaik microsoft sees j# as "migration" to c# --> and as a result of this j# 
should not be seen as complete programming language, it should be used to use 
your j components in the .net framework!

Regarding gnovos-ga's observations regarding the prevalence of Java in
China and Japan, I wanted to add this: I lived and worked as a
developer in both Japan and China. In China, the government is the
biggest buyer (as something like 90% of business in China is
state-owned) and the govt. steers clear of Microsoft technologies
because of the license fees. I worked on an eLearning system for
education (university level) and it had to be ported to Unix, Java in
order to bid on govt contracts. The govt was look at our solution for
the Ministry of Railway but didn't want to pay the MS license fees to
roll it out at training centers across the country.

In Japan, where the computer is a relatively new thing in business
(most Japanese don't have personal computers; most web access in Japan
is by mobile (keitai) telephone) DoCoMo the largest mobile phone
service provided adopted Java for the development language for their
mobile phone system platform. This creates a huge interest in Java for
upcoming developers and Microsoft never had a chance to establish a
monopoly there.

Also with regard to CHina, Microsoft has a huge research park in
Beijing (probably cheaper than trying to get so many VISAs for CHinese
engineers) but I don't think Microsoft even considers CHina as a
market. In fact I read that MS pulled out of CHina because they can't
complete with pirate versions of software in a country with no respect
for intellectual property.

Small update (after nearly a year): Google does now support searches
in the style of "j#":

Google Search - "j#"
://www.google.com/search?q=%22j%23%22