In article <l63bem$6me$1@speranza.aioe.org>, This.is@not.Me says...
>
> Hi Tim,
>
> On 11/14/2013 10:51 AM, Tim Wescott wrote:
> > On Thu, 14 Nov 2013 00:19:17 -0700, Don Y wrote:
> >
> >> <http://www.infoworld.com/t/application-development/c-pulls-away-java-
> > among-top-programming-languages-230603>
> >>
> >> That'ss gotta say *something* about (perceived) "usability"
> >> despite all the other contenders that have come (and gone)
> >> over that timespan... :-/
> >
> > "The popular search engines Google, Bing, Yahoo!, Wikipedia, Amazon,
> > YouTube and Baidu are used to calculate the ratings. Observe that the
> > TIOBE index is not about the best programming language or the language in
> > which most lines of code have been written."
> >
> > So, it's not about lines of code written, or $$ made on the code so
> > produced, it's about what shows up most on search engines. That doesn't
> > see to be a very useful measure, unless you're selling advertising space
> > on programming-language websites.
>
> I don't think it can be trivialized away that easily.
>
> You can't deduce the per-unit profit of a Betamax vs. VHS VCR
> based on the number of *titles* available in each format.
> Nor can you deduce the technical merits of the two technologies.
> Or, the reliability of the transports and electronics.
Meaningless see later
> But, you can see what MOVIE VENDORS perceive the markets for
> each of these technologies to be! The "relative interest".
They tend to do that through figures of sales around the world of the
playing media (from machines to website visits) to figures of new sales
trends (actual existing media and stats for downloads).
> It is a measure of one aspect of the "market" -- the number
> of LIVE web pages concerned with the particular language.
> Assuming (!) web page creators are reflecting their idea of what
> "the market" (defined as people wanting to find/create web pages)
> is interested in, this is a great way of "measuring".
It is a limited form of market of search engines that does not count
in the interlinked nature of the web. Where unless they have also
downloaded spyware onto everyones computer the cases where different
search methods are used they miss the following -
1/ those who may start with a school/university website its content
and its direct links to known sites
2/ those who start from books that contain still active links and maybe
direct links elsewhwere as well
3/ those who spend time in newsgroups/forums and get direct links from
them
4/ those who get links from colleagues
5/ it only shows the start of a search on the search engine find a page
that then takes them to a myriad of other resources NOT using the
search engine
6/ those who have bookmarked sites they frequent for such information
A well ranked site generally has lots of links to and/or from other
RELEVANT and RELATED sites.
For the figures to have any accuracy beyond what happens on search
engines, this assumes all searches are done from the search engine and
NOBODY follows any links on that site etc. . In other words they ALL
ONLY look at the single page presented and repeat the search or go back
to look at the next site.
> You note "BASIC" isn;t on the list. And, it would be a lead pipe
> cinch to create a web page with scads of BASIC code.
>
> BUT, NO ONE WOULD BE INTERESTED IN IT!
Many of the existing ones may already be visited many times from
other mthods. see later
> Note that the second metric cited on the page (PyPL) goes hunting
> for *tutorials*. These are web page creators casting their votes
> as to what languages they "see" as "up and coming". Presumably,
> mature languages (and simple languages) have fewer needed tutorials;
> "everyone ALREADY knows how to do this"!
Depends on definition on tutorial.
...
> What the index is *missing* is the consumer-side view of all
> this: how many searches were *issued* with these terms? How
> many pages were *visited*, etc.
How many pages did they visit on each site for that search link
Was that tutorial 10 pages and video for one topic
Did that page contain 10 topics and did they find the others relevant
Did they follow links out of that site
How many of those searches are REPEATS of the same term from same person
or others
More importantly did they skip the search engine and go directly to the
site. Typical examples would be for PHP, Python and those using
w3schools.
Many other metrics missing
Some languages may be lower in the ranks for many reasons
--
Paul Carpenter | paul@pcserviceselectronics.co.uk
<http://www.pcserviceselectronics.co.uk/> PC Services
<http://www.pcserviceselectronics.co.uk/pi/> Raspberry Pi Add-ons
<http://www.pcserviceselectronics.co.uk/fonts/> Timing Diagram Font
<http://www.gnuh8.org.uk/> GNU H8 - compiler & Renesas H8/H8S/H8 Tiny
<http://www.badweb.org.uk/> For those web sites you hate
Reply by Don Y●November 14, 20132013-11-14
Hi Tim,
On 11/14/2013 10:51 AM, Tim Wescott wrote:
> On Thu, 14 Nov 2013 00:19:17 -0700, Don Y wrote:
>
>> <http://www.infoworld.com/t/application-development/c-pulls-away-java-
> among-top-programming-languages-230603>
>>
>> That'ss gotta say *something* about (perceived) "usability"
>> despite all the other contenders that have come (and gone)
>> over that timespan... :-/
>
> "The popular search engines Google, Bing, Yahoo!, Wikipedia, Amazon,
> YouTube and Baidu are used to calculate the ratings. Observe that the
> TIOBE index is not about the best programming language or the language in
> which most lines of code have been written."
>
> So, it's not about lines of code written, or $$ made on the code so
> produced, it's about what shows up most on search engines. That doesn't
> see to be a very useful measure, unless you're selling advertising space
> on programming-language websites.
I don't think it can be trivialized away that easily.
You can't deduce the per-unit profit of a Betamax vs. VHS VCR
based on the number of *titles* available in each format.
Nor can you deduce the technical merits of the two technologies.
Or, the reliability of the transports and electronics.
But, you can see what MOVIE VENDORS perceive the markets for
each of these technologies to be! The "relative interest".
It is a measure of one aspect of the "market" -- the number
of LIVE web pages concerned with the particular language.
Assuming (!) web page creators are reflecting their idea of what
"the market" (defined as people wanting to find/create web pages)
is interested in, this is a great way of "measuring".
You note "BASIC" isn;t on the list. And, it would be a lead pipe
cinch to create a web page with scads of BASIC code.
BUT, NO ONE WOULD BE INTERESTED IN IT!
Note that the second metric cited on the page (PyPL) goes hunting
for *tutorials*. These are web page creators casting their votes
as to what languages they "see" as "up and coming". Presumably,
mature languages (and simple languages) have fewer needed tutorials;
"everyone ALREADY knows how to do this"!
You could argue that simple/mature languages might have MORE such pages
(because any web-page-creator-wannabe could throw together even
the simplest of inane tutorials!).
You could argue that simple/mature languages might have *less* (as above).
You could argue that complex languages would have more as there are
more *issues* that need clarification.
You could argue that complex languages would have LESS as there are
fewer folks knowledgeable in those issues willing to spend time
creating web pages to address them. :>
The fact that C has remained at the top of this list for SO LONG
has to mean "something".
E.g., I don't see APL, Limbo, etc.
What the index is *missing* is the consumer-side view of all
this: how many searches were *issued* with these terms? How
many pages were *visited*, etc.
>
> That'ss gotta say *something* about (perceived) "usability"
> despite all the other contenders that have come (and gone)
> over that timespan... :-/
"The popular search engines Google, Bing, Yahoo!, Wikipedia, Amazon,
YouTube and Baidu are used to calculate the ratings. Observe that the
TIOBE index is not about the best programming language or the language in
which most lines of code have been written."
So, it's not about lines of code written, or $$ made on the code so
produced, it's about what shows up most on search engines. That doesn't
see to be a very useful measure, unless you're selling advertising space
on programming-language websites.
--
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com
Reply by David Brown●November 14, 20132013-11-14
On 14/11/13 10:37, John Devereux wrote:
> David Brown <david@westcontrol.removethisbit.com> writes:
>
>> On 14/11/13 08:19, Don Y wrote:
>>> <http://www.infoworld.com/t/application-development/c-pulls-away-java-among-top-programming-languages-230603>
>>>
>>>
>>> That'ss gotta say *something* about (perceived) "usability"
>>> despite all the other contenders that have come (and gone)
>>> over that timespan... :-/
>>
>> To paraphrase a rather famous politician, C is the worst of all
>> programming languages, but it's the best we've got!
>>
>> What surprises me about that list is that Objective C is more popular
>> than C++ according to this survey. I know Objective C is common for
>> Macs and iThingies, but it is rare outside that - and it is certainly
>> not the only language used for programming such systems. I find it hard
>> to believe that 9.4% of worldwide software development work is purely
>> for Apple-only work.
>
> I don't.
>
> <http://ipod.about.com/od/iphonesoftwareterms/qt/apps-in-app-store.htm>
>
> Someone has to write all that crap. (Well actually I fully expect
> someone has automated this by now, spamming the iphone store with 10s of
> thousands of look-alike junk!)
>
> My customer keeps asking for an "iphone app", for an embedded device we
> make for them. No idea what they actually want it to do, and never mind
> that iphones are only 15% of the smartphone market, vs 80% for Android.
>
> But I suppose I am going to have to get into this stuff :(
>
I know there are a lot of apps written for iPod/iPhone - often well out
of proportion given the market share compared to Android. But how many
of these are actually written in Objective C? I haven't written
anything for an iThingy myself (I never even owned one), but a large
number of apps are cross-platform between Android and iOS. Surely these
are written in something else - Python, HTML5, C++, Java, etc.?
>> The numbers in this survey come from how often people search for
>> tutorials for languages. Maybe it is just that people find Objective C
>> harder than C++, and need more tutorials. Or maybe good tutorials for
>> Objective C are harder to find, and people have to search more for them!
>
>
Reply by John Devereux●November 14, 20132013-11-14
David Brown <david@westcontrol.removethisbit.com> writes:
> On 14/11/13 08:19, Don Y wrote:
>> <http://www.infoworld.com/t/application-development/c-pulls-away-java-among-top-programming-languages-230603>
>>
>>
>> That'ss gotta say *something* about (perceived) "usability"
>> despite all the other contenders that have come (and gone)
>> over that timespan... :-/
>
> To paraphrase a rather famous politician, C is the worst of all
> programming languages, but it's the best we've got!
>
> What surprises me about that list is that Objective C is more popular
> than C++ according to this survey. I know Objective C is common for
> Macs and iThingies, but it is rare outside that - and it is certainly
> not the only language used for programming such systems. I find it hard
> to believe that 9.4% of worldwide software development work is purely
> for Apple-only work.
I don't.
<http://ipod.about.com/od/iphonesoftwareterms/qt/apps-in-app-store.htm>
Someone has to write all that crap. (Well actually I fully expect
someone has automated this by now, spamming the iphone store with 10s of
thousands of look-alike junk!)
My customer keeps asking for an "iphone app", for an embedded device we
make for them. No idea what they actually want it to do, and never mind
that iphones are only 15% of the smartphone market, vs 80% for Android.
But I suppose I am going to have to get into this stuff :(
> The numbers in this survey come from how often people search for
> tutorials for languages. Maybe it is just that people find Objective C
> harder than C++, and need more tutorials. Or maybe good tutorials for
> Objective C are harder to find, and people have to search more for them!
To paraphrase a rather famous politician, C is the worst of all
programming languages, but it's the best we've got!
What surprises me about that list is that Objective C is more popular
than C++ according to this survey. I know Objective C is common for
Macs and iThingies, but it is rare outside that - and it is certainly
not the only language used for programming such systems. I find it hard
to believe that 9.4% of worldwide software development work is purely
for Apple-only work.
The numbers in this survey come from how often people search for
tutorials for languages. Maybe it is just that people find Objective C
harder than C++, and need more tutorials. Or maybe good tutorials for
Objective C are harder to find, and people have to search more for them!