Delphi in the right direction
Fred Derf
![Paul Nichols [TeamB]'s profile photo](https://lh3.googleusercontent.com/a/default-user=s40-c)
Paul Nichols [TeamB]
No, Delphi does not have the visibility of Java,C, C++, or even VB!!
Is this something new?
Now what you could take away from this is that all (with the exception
of VB) in the top 6, are all XPlatform.
Wonder where Delphi would be if it were XPlatform?
Fred Derf
I say that it would be at position #20 or so, because it would be twice as
buggy as it is now.
Paul Nichols [TeamB]
Nick Hodges (CodeGear)
> Sounds like you have an axe to grind..
Ya think? ;-)
--
Nick Hodges
Delphi Product Manager - CodeGear
http://blogs.codegear.com/nickhodges
Robin
> http://www.tiobe.com/tpci.htm
Suckers :-)
I am glad the competition are not using Delphi.
--
Robin.
Australian Bridal Accessories := http://www.bridalbuzz.com.au
Turbo for Noobs (a work in progress) := http://turbofornoobs.blogspot.com/
Paul Nichols [TeamB]
> Paul Nichols [TeamB] wrote:
>
>> Sounds like you have an axe to grind..
>
> Ya think? ;-)
>
Fred Derf
Having fun? Good for you, Delphi Product Manager. ;-)
What is your opinion on the decline of Delphi?
Dave Nottage [TeamB]
> Sounds like you have an axe to grind..
There seems to be an axe that hacks those figures together. Beats me
why anyone would want to use that index to plan their future. I
wouldn't, even if Delphi or C# were at the top of the list.
--
Dave Nottage [TeamB]
Dave Nottage [TeamB]
> Agree there are flaws to this list. Popularity cannot be measures by
> hits, but by projects.
Something that puzzles is their assertion: "The ratings are based on
the world-wide availability of skilled engineers, courses and third
party vendors.."
then: "The popular search engines Google, MSN, and Yahoo! are used to
calculate the ratings"
Which begs the question: what methods can you use from those search
engines to come up with data that resembles what they describe in the
first statement?
Note also that they added MSN and Yahoo search because Google changed
their engine in 2004 which resulted in skewed figures, which they kept
anyway. Sheesh.
--
Dave Nottage [TeamB]
Paul Nichols [TeamB]
John
> Wonder where Delphi would be if it were XPlatform?
one more reason to have a Cross Platform Delphi :-))
Bruce McGee
> Which begs the question:
These "statistics" were discredited when they were in Delphi's favour.
I don't see why they would be any more accurate now.
--
Regards,
Bruce McGee
Glooscap Software
!["Henrick Hellström [StreamSec]"'s profile photo](http://lh3.googleusercontent.com/a-/ALV-UjV3KHenxfvxLOAqv2Ffez_9o81Et7uiVcF5l_3bHlpoQ-o7=s40-c)
"Henrick Hellström [StreamSec]"
> Something that puzzles is their assertion: "The ratings are based on
> the world-wide availability of skilled engineers, courses and third
> party vendors.."
> Which begs the question: what methods can you use from those search
> engines to come up with data that resembles what they describe in the
> first statement?
If there are skilled engineers, courses and third party vendors, they
surely have some form of web presence. Hence, you can locate them using
search engines. The question is rather how the entities mirrored by the
measured figures are defined. For instance, how do you define "course"
in a manner that makes it clear how to tell one course from either two
separate courses or half of a course? Is a course with 1 participant
just as much of a course as a course with 1000 participants? Is a course
about general web design that happens to use PHP or C# as the language
for the examples as much of a course for those languages, as a course
that focuses specifically on Delphi development is a course for Delphi?
Once you got those definitions clear (which obviously is far from
trivial), it makes sense to start discussing the error margin of the
figures actually measured.
Kim Madsen
meaningless.
The problem is that the shown results can be/are used as ammunition by
people favoriting the top languages.
The majority of people reading the results, dont actually analyze them to
figure out how they came about. They take the list as a truth or at least as
an indicator.
That makes this site PR wise very bad for CodeGear and us as such ammunition
would not favor Delphi.
What could be needed are measurements positively favoriting Delphi. Without
having done any investigation, I could think of areas like productivity,
maintainability and lifetime of Delphi applications vs C#, Java, PHP, Ruby
and Basic based ones on the various platforms/IDEs.
That would give the pro Delphi managers some much needed ammunition.
I would even think that although we all guess that the number of Delphi
developers has declined over the past few years, it would actually bolster
Delphis reputation with public announcements from CodeGear about numbers.
Currently many managers and C#/Java/Basic evangelists spread information
about their beliefe that only one nerd a million (or so) use Delphi.
Even if the number of Delphi users are 'only' in the hundred of thousands it
would be positive ammunition to know the number to be able to counter the
arguments that 'Nobody use Delphi today'.
In fact... what the Delphi community and CG should focus on right now is to
cast ammunition that we as a community can use in the always ongoing
discussions about what next project should be programmed in.
Just my few cents.
best regards
Kim Madsen
k...@components4developers.com
www.components4developers.com
The best components for the best developers
High performance, high availability and highly scalable n-tier development
tools
"Dave Nottage [TeamB]" <qni...@enqfbsg.pbz.nh> wrote in message
news:46b80bec$1...@newsgroups.borland.com...
"Henrick Hellström [StreamSec]"
> That makes this site PR wise very bad for CodeGear and us as such ammunition
> would not favor Delphi.
> What could be needed are measurements positively favoriting Delphi. Without
> having done any investigation, I could think of areas like productivity,
> maintainability and lifetime of Delphi applications vs C#, Java, PHP, Ruby
> and Basic based ones on the various platforms/IDEs.
That's actually a good point.
<satire>
Developers HATE Delphi for the following reasons:
* Delphi developers can finish complex projects themselves. Often there
isn't a fellow Delphi developer around for you to grab a beer with after
work. Who likes that?
* Delphi developers build projects that last. They are done in a month
and their employer never asks them to return to fix bugs and other
issues. If you code in other languages at least you can be certain you
have a steady job for a very long time.
* Delphi is too darn easy to learn. You never get to go to fancy courses
and conferences to learn how to use it, and it is hard to convince
anyone but beginners to hire you to teach them to use it.
* You are STUPID if you write a modular Delphi application, because that
just means you will never have to get back to work on the application.
You have to convince your employer it is so easy to extend existing
Delphi code that it doesn't have to be modular.
</satire>
Dave Nottage [TeamB]
> Hence, you can locate them using search engines.
Except there are no details as to how they are doing so. IME using
search engines to do such a thing is hit and miss at the best of times.
--
Dave Nottage [TeamB]
Paul Nichols [TeamB]
> Dave Nottage [TeamB] wrote:
>
>> Which begs the question:
>
>
> These "statistics" were discredited when they were in Delphi's favour.
> I don't see why they would be any more accurate now.
>
I do not doubt there is some validity to the numbers. I am surprised
however, that one could conclude that VB numbers are so high (even
though the Visual is in parenthesis, so is this counting all forms of
basic, including scripting and VBA on both Windows and Open Office ?)
but no one should be surprised that some of the others (Java, C/C++) are
much higher than Delphi.
With Delphi you are basically speaking of more traditional Client/Server
or Client side applications for MS Windows only, which of course, is
where Delphi excels.
The other languages (with the exception of VB, which makes no sense
occupying this position), are more XPlatform or primarily used as Web
based and/or OS based technologies, meaning they have higher target
areas and audiences by default (more targets=more changes for development).
Notice however, that even in this survey, Microsoft's new standard, C#,
has no where near the user base that Java or C/C++ has either. So if we
assume (subjective of course) that out of all of the languages for
Windows only type development, Delphi actually,is not that far behind
C#. Looking at things in what I would consider the correct perspective,
this report actually makes some sense.
In other words, you should not be attempting to compare Delphi
popularity in the large as compared to Java, C/C++, or Java Script. Why?
Because the former languages are not targeting a single platform and are
generally not used for a specific purpose or target. That would be like
comparing the Honda automobile division against the Honda motorcycle
division and concluding that Honda Motorcycles are not popular, since
they are dwarfed by the sale of Honda cars. The automobile division is a
general use device, with potential use by everyone for general
transportation, whereas the motorcycle division is for a focused group
and a focused purpose. In reality, both Honda automobile division and
Honda motorcycle division, are doing quite well, though the numbers of
comparative sales, would convince you that Honda motorcycles are failing.
VLmartins
"Fred Derf" <nos...@nospam.com> escreveu na mensagem
news:46b7...@newsgroups.borland.com...
Visual basic is in the TOP three, you can choose VB to your projects, and
you could have a nice futur...
Ray Porter
"Paul Nichols [TeamB]" <pa...@none.com> wrote in message
news:46b874e2$1...@newsgroups.borland.com...
>
> In other words, you should not be attempting to compare Delphi popularity
> in the large as compared to Java, C/C++, or Java Script. Why? Because the
> former languages are not targeting a single platform and are generally not
> used for a specific purpose or target. That would be like comparing the
> Honda automobile division against the Honda motorcycle division and
> concluding that Honda Motorcycles are not popular, since they are dwarfed
> by the sale of Honda cars. The automobile division is a general use
> device, with potential use by everyone for general transportation, whereas
> the motorcycle division is for a focused group and a focused purpose. In
> reality, both Honda automobile division and Honda motorcycle division, are
> doing quite well, though the numbers of comparative sales, would convince
> you that Honda motorcycles are failing.
Very nice analogy, Paul. Viewed in this light, the figures do make some
sense even though I have serious doubts about any figures based on search
engines. However, I don't think the typical PHB will give the matter the
kind of thought you did. In fact, I'm not sure they'd even understand the
differences between the tools.
Ray
I have serious doubts about the validity of any statistics developed based
on popular search engines, at least as long as we don't know how they
structured their queries.
Captain Jake
news:46b80bec$1...@newsgroups.borland.com...
> engines to come up with data that resembles what they describe in the
> first statement?
Perhaps they are using the search engines to estimate the availability of
skilled engineers, courses and third-party vendors.
Here in the United States, it is my perception that Delphi is a niche
product at this time, such that if you want to continue doing Delphi and get
a decent wage for doing so, you must look at the national market and be
willing to relocate or work away from home for long periods of time.
Senior developers usually don't want to move anymore, so by and large they
have left Delphi for other languages.
Paul Nichols [TeamB]
> Ray
>
> I have serious doubts about the validity of any statistics developed based
> on popular search engines, at least as long as we don't know how they
> structured their queries.
>
>
Understand. Although statistic, if arrived through proven scientific and
mathetical equations, do not lie, the way they are presented can make
all of the difference in the world.
Look at it another way. Since China and India have over 3 times the
population of the US, you would think that logic would dictate that
these are the markets you would want target:
US: 300 million potential customers/consumers
China: 1.3 Billion potential customers/consumers
India: 1.0 Billion potential customers/consumers
Conclusion: since US has 1/3 potential customers of China or India, US
marketing focus should be 1/3 of these others countries.
While the statistics are accurate, the application of market potential
where they are applied in this case, is not. Why? While raw numbers of
market potential are important, the ability of the customer/consumer to
purchase goods and service is even more important. So this marketing
analysis is flawed, though statistically, the information is correct.
If you take market potential by ability to buy or purchase, the US is a
much better market than both India and China, even though population
would dictate differently. (2005 stats)
Average Medium income
China: $932.00 per year
India: $710.00 per year
US: $46,326.00 per year
So which market would you target first?
Of course the one who has the ability to make purchases, which is
neither China nor India.
Markus.Humm
> Dave Nottage [TeamB] wrote:
>> Something that puzzles is their assertion: "The ratings are based on
>> the world-wide availability of skilled engineers, courses and third
>> party vendors.."
> [...]
>> Which begs the question: what methods can you use from those search
>> engines to come up with data that resembles what they describe in the
>> first statement?
>
> If there are skilled engineers, courses and third party vendors, they
Hm, I'd say I'm skilled, but I have nowhere any web presence (except if
you consider my NG postings) which states that I'm skilled in Delphi.
I assume many people simply working in companies and programming there
fall into that category. => the search engines can only be inaccurate here.
Greetings
Markus
"Henrick Hellström [StreamSec]"
> Hm, I'd say I'm skilled, but I have nowhere any web presence (except if
> you consider my NG postings) which states that I'm skilled in Delphi.
> I assume many people simply working in companies and programming there
> fall into that category. => the search engines can only be inaccurate here.
In practice, probably yes, but in theory it *might* work, provided that
you find a formula that describes the relation between newsgroup
postings and availability of skilled engineers. An educated guess is
that the formula is kxlnx, where the logarithmic factor describes that
some developers will search Google for answers first, and the linear
factor describes that some developers will always post first and search
Google later.
GrandmasterB
news:46b80bec$1...@newsgroups.borland.com...
> the world-wide availability of skilled engineers, courses and third
> party vendors.."
Wierd, isnt it? See, to me, if they're basing it on the availability of
skilled engineers, that may instead just be reporting what type of engineer
isnt very employable. :-)
> then: "The popular search engines Google, MSN, and Yahoo! are used to
> calculate the ratings"
Or in this case, it could measure the difficulty of the language - more
difficult = need to do more searches.
Judging anything based on 'internet popularity' as this outfit has done is
just bad statistics.
Michael C.
> Dave Nottage [TeamB] wrote:
>
>> Which begs the question:
>
>
> These "statistics" were discredited when they were in Delphi's favour.
> I don't see why they would be any more accurate now.
>
You're talking about something being "discredited" without
backing it up ...
Michael C.
I like the TIOBE index.
I think it's a fairly reasonable indication what's going on in the
programming world.
I'm not sure I would use it to plan a future project;
I would just use it for an indication for what is popular.
Now if anyone wants to present some evidence that indicates
why the TIOBE index is flawed - please present a reasonable argument.
"Henrick Hellström [StreamSec]"
> Now if anyone wants to present some evidence that indicates
> why the TIOBE index is flawed - please present a reasonable argument.
Dave Nottage gave such an argument.
> Something that puzzles is their assertion: "The ratings are based on
> the world-wide availability of skilled engineers, courses and third
> party vendors.."
>
> then: "The popular search engines Google, MSN, and Yahoo! are used to
> calculate the ratings"
>
> Which begs the question: what methods can you use from those search
> engines to come up with data that resembles what they describe in the
> first statement?
This is a reasonable argument for the same reason you might claim that
astrology is flawed - not because you necessarily have hard evidence to
prove astrology isn't a sound scientific method, but because it isn't
self evident astrology is a sound scientific method and no astrologer
has presented hard evidence to prove astrology is a sound scientific method.
Bruce McGee
> You're talking about something being "discredited" without
> backing it up ...
Google for discussions on TIOBE and Delphi.
Mike Vance
is just at 10.11% -- so obviously we should all be programming in
straight "C" now. C# is at 3.99% I hope the C# and C++ users are ready
to "increase" their productivity by jumping ship to straight "C". There
is no perfect language/framework out their, but Delphi is just plain the
best out their.
Dave Nottage [TeamB]
> Now if anyone wants to present some evidence that indicates
> why the TIOBE index is flawed - please present a reasonable argument.
I'm saying that the index is questionable. There don't seem to be any
answers to my questions, eg what search method(s) are being used. My
experience of searching for similar data using those engines is very
hit-and-miss.
Any reasonable arguments as to why/how using those search engines could
produce reliable data?
--
Dave Nottage [TeamB]
Michael C.
> Michael C. wrote:
>
>> You're talking about something being "discredited" without
>> backing it up ...
>
> Google for discussions on TIOBE and Delphi.
>
That produces many links.
Did you have any specific links in mind that give evidence that
the TIOBE is poor?
Michael C.
> Michael C. wrote:
>> Now if anyone wants to present some evidence that indicates
>> why the TIOBE index is flawed - please present a reasonable argument.
>
> Dave Nottage gave such an argument.
>
He did?
I thought all his did was ask a few questions.
Asking a few questions doesn't indicated something is flawed.
You must produce evidence that INDICATES something is flawed.
> This is a reasonable argument for the same reason you might claim that
> astrology is flawed
<snip>
I don't agree.
Michael C.
You can question the index all you want.
That doesn't indicate it's not accurate or inaccurate.
However,
the index seems a reasonable indication about what is popular to me.
That's why I'm giving you and anyone else here an
opportunity to present evidence why it's a bad or "flawed" index
from a statistical point of view.
Dave Nottage [TeamB]
> You can question the index all you want.
> That doesn't indicate it's not accurate or inaccurate
I am asking the questions *because* there's no indication as to how/why
the index is accurate or not.
> However,
> the index seems a reasonable indication about what is popular to me.
> That's why I'm giving you and anyone else here an
> opportunity to present evidence why it's a bad or "flawed" index
> from a statistical point of view.
Conversely, I'd like to see evidence why it's a "reasonable indication
about what is popular". Until I see any, I consider it risky to rely on
it for anything.
--
Dave Nottage [TeamB]
Anders Isaksson
> You must produce evidence that INDICATES something is flawed.
We don't *must* anything! The party saying "X is Y" is the one that
should supply the evidence. Others can get by with reasonable doubt.
--
Anders Isaksson, Sweden
BlockCAD: http://web.telia.com/~u16122508/proglego.htm
Gallery: http://web.telia.com/~u16122508/gallery/index.htm
Anders Isaksson
> the index seems a reasonable indication about what is popular to me.
How could it? Have you seen the query used to get the index? If not,
how can you say *anything* about its relation to 'popularity'? Maybe C
is high on the list because it is mentioned very often in connection
with music (what's the position for A,B,D,E,F,G?), and children's sites
(easy as A B C)...
As long as we don't know *anything* about the querys and filters, we
don't know *anything* about the relevance of the result. Period.
Did you know that there is a very good correlation between the length
of women's skirts and the index of the stock exchange (at least up till
about 1970)?
"Henrick Hellström [StreamSec]"
> Conversely, I'd like to see evidence why it's a "reasonable indication
> about what is popular". Until I see any, I consider it risky to rely on
> it for anything.
What it probably *does* indicate is how many hits you get when you use a
search engine for trying to solve your programming problems.
Bruce McGee
> That produces many links.
> Did you have any specific links in mind that give evidence that
> the TIOBE is poor?
I don't have a specific link for you, but if you're interested, you'll
find it in there.
Do you have any that give evidence that the TIOBE trends are indicative
of reality?
John Jacobson
> Any reasonable arguments as to why/how using those search engines could
> produce reliable data?
They don't need to be reliable, just unbiased. I don't see any reason to
think that they are biased for or against any of the languages in the
comparison.
Craig Stuntz [TeamB]
> Did you have any specific links in mind that give evidence that
> the TIOBE is poor?
How hard is it to come up with reasons why Ghits might not reflect
reality? Or that the TIOBE results may be even worse than raw Ghits?
Fact: When Google changes their algorithms, the TIOBE results can
change dramatically, in spite of no evidence for actual programming
language usage changes at that time. TIOBE openly acknoledges this on
their site.
Fact: TIOBE results are weighted by search engine based on Alexa
rankings, and there is strong evidence that Alexa is (1) not
representative of the web as a whole, and (2) especially wrong when it
comes to technical readership. (See
http://slashdot.org/articles/07/07/23/152243.shtml)
Fact: TIOBE presents an absurd level of precision (down to .001%!)
without providing any evidence that its metric is actually that
accurate. See http://en.wikipedia.org/wiki/False_precision
Then there's the whole issue of the "search engine test" in the first
place, a notion which is hardly unique to TIOBE. This page has a good
discussion of the issues:
http://en.wikipedia.org/wiki/Wikipedia:Search_engine_test
In short, Ghits have some relationship to reality, but it's not at all
clear that it's a very close one. They're useful as a rough guess, but
presenting figures with false precision borders on lying to me.
--
Craig Stuntz [TeamB] · Vertex Systems Corp. · Columbus, OH
Delphi/InterBase Weblog : http://blogs.teamb.com/craigstuntz
IB 6 versions prior to 6.0.1.6 are pre-release and may corrupt
your DBs! Open Edition users, get 6.0.1.6 from http://mers.com
Craig Stuntz [TeamB]
> "Dave Nottage [TeamB]" <qni...@enqfbsg.pbz.nh> wrote in message
> news:46ba6bfa$1...@newsgroups.borland.com...
> > Any reasonable arguments as to why/how using those search engines
> > could produce reliable data?
>
> They don't need to be reliable, just unbiased.
Why? If I throw darts at a board while blindfolded, I'll come up with
unbiased results. But what would it tell me?
--
Craig Stuntz [TeamB] · Vertex Systems Corp. · Columbus, OH
Delphi/InterBase Weblog : http://blogs.teamb.com/craigstuntz
Want to help make Delphi and InterBase better? Use QC!
http://qc.borland.com -- Vote for important issues
David Erbas-White
>
> Why? If I throw darts at a board while blindfolded, I'll come up with
> unbiased results. But what would it tell me?
>
It will tell you how consistent you are (really -- this can be
important!!! <G>).
I'm an occasional archer. Sometimes when I want to 'test' myself I'll
close my eyes while shooting in order to check my consistency of
positioning my body (and yes, I DO hit the target).
David Erbas-White
Nick Hodges (CodeGear)
> However,
> the index seems a reasonable indication about what is popular to me.
> That's why I'm giving you and anyone else here an
> opportunity to present evidence why it's a bad or "flawed" index
> from a statistical point of view.
I'm not too concerned about whether the TIOBE index is flawed or not,
however, it is ridiculous for you to assert the above -- the burden of
proof is always on the affirmative. You are assertive an affirmative
statement, and you have to prove it.
Or, put more colloquially: Everything is BS until it's proved not to
be. ;-)
--
Nick Hodges
Delphi Product Manager - CodeGear
http://blogs.codegear.com/nickhodges
Michael C.
> Michael C. wrote:
>
>> That produces many links.
>> Did you have any specific links in mind that give evidence that
>> the TIOBE is poor?
>
> I don't have a specific link for you, but if you're interested, you'll
> find it in there.
>
> Do you have any that give evidence that the TIOBE trends are indicative
> of reality?
>
Generally, the index seems to be accurate according to my own
empirical observations.
Now, if you have any evidence that indicates it's not,
please do share.
Michael C.
> Michael C. wrote:
>
>> Did you have any specific links in mind that give evidence that
>> the TIOBE is poor?
>
> How hard is it to come up with reasons why Ghits might not reflect
> reality? Or that the TIOBE results may be even worse than raw Ghits?
>
> Fact: When Google changes their algorithms, the TIOBE results can
> change dramatically, in spite of no evidence for actual programming
> language usage changes at that time. TIOBE openly acknoledges this on
> their site.
They also acknowledge that there index is only an "indication of the
popularity of programming languages."
In that spirit,
they change their algorithm so that they are not only checking google
for statistics but also MSN and Yahoo.
>
> Fact: TIOBE results are weighted by search engine based on Alexa
> rankings, and there is strong evidence that Alexa is (1) not
> representative of the web as a whole, and (2) especially wrong when it
> comes to technical readership. (See
> http://slashdot.org/articles/07/07/23/152243.shtml)
>
A person doesn't need "the best" ranking system in order to collect
statistics. A person should, however, pick a ranking system that they
think is good.
If the people who create TIOBE choose Alexa,
I believe they did it because they thought it was "good enough" for
their index.
> Fact: TIOBE presents an absurd level of precision (down to .001%!)
> without providing any evidence that its metric is actually that
> accurate. See http://en.wikipedia.org/wiki/False_precision
However, they do explain at the beginning of their web page that their
index is only an "indication" of popularity of programming languages.
I don't think a person should be expecting accuracy after that statement.
> Then there's the whole issue of the "search engine test" in the first
> place, a notion which is hardly unique to TIOBE. This page has a good
> discussion of the issues:
>
> http://en.wikipedia.org/wiki/Wikipedia:Search_engine_test
>
> In short, Ghits have some relationship to reality, but it's not at all
> clear that it's a very close one. They're useful as a rough guess, but
> presenting figures with false precision borders on lying to me.
I think we should all realize that statistics sometimes can only give
us an "indication" about a certain topic;
Statistics don't have to be 100 percent accurate in order to be
considered good.
I believe the TIOBE index is more accurate than a "rough guess."
I think it's even better than "educated guess".
I find the TIOBE index a reasonable index.
So far, no one has convinced me otherwise.
I haven't found a better "popular programming language" index yet.
If you know of a better index that measures
"popularity of programming languages",
please do share.
Michael C.
> Michael C. wrote:
>
>> You must produce evidence that INDICATES something is flawed.
>
> We don't *must* anything!
True.
Although, no one here has convinced me that the TIOBE index is not good
or "good enough".
Paul Nichols [TeamB]
I would expect C to be more popular than C++, due to the huge amount of
Open Source software that is written in C. Most database systems are
also written in C, not C++, although there may be some algorythms
written in C++.
C usually produces the fastest compiled code, which is one reason why it
is preferred for low level operations. Not always the case, but usually
this is proven to be true.
Most OS systems are written in C, not C++. C++ is generally better for
Device Drivers and application development.
This does not mean there is no reason to use C++, nor more than it means
you should not use X. The best way to determine what you used is by
utilizing a Use Case analysis, not only on the requirements, but also on
the technology used.
Michael C.
> Michael C. wrote:
>
>> You can question the index all you want.
>> That doesn't indicate it's not accurate or inaccurate
>
> I am asking the questions *because* there's no indication as to how/why
> the index is accurate or not.
>
The truth is statistics don't have to be 100 percent accurate.
Sometimes, statistics can only be an indication about a certain trend.
Personally I find the TIOBE index reasonable based on my own observation
around me.
I do find the following reasonable:
Java is the most popular language.
The most popular languages are in this order:
JAVA, C, Visual Basic, and C++.
Python ( a cross platform language ) is
more popular than Delphi.
Javascript is twice as popular as Delphi.
Delphi isn't a very popular language.
<snip>
Michael C.
> Michael C. wrote:
>
>> the index seems a reasonable indication about what is popular to me.
>
> How could it?
See my reply to Dave Nottage above.
<snip>
Nick Hodges (CodeGear)
> Although, no one here has convinced me that the TIOBE index is not
> good or "good enough".
Indeed -- and no one has convinced me that I shouldn't be a movie star.
Michael C.
> Michael C. wrote:
>
>> However,
>> the index seems a reasonable indication about what is popular to me.
>> That's why I'm giving you and anyone else here an
>> opportunity to present evidence why it's a bad or "flawed" index
>> from a statistical point of view.
>
> I'm not too concerned about whether the TIOBE index is flawed or not,
> however, it is ridiculous for you to assert the above -- the burden of
> proof is always on the affirmative. You are assertive an affirmative
> statement, and you have to prove it.
>
> Or, put more colloquially: Everything is BS until it's proved not to
> be. ;-)
I find the index a reasonable indication popular programming languages.
For example, I do find that my local bookstores tend to stock a lot of Java books,
more so than books related to Visual Basic, C++, and C#.
I also find the various C# and Visual Basic newsgroups are much more
popular than the Delphi newsgroups.
I also think it's reasonable that Python is more popular than Delphi
based on my experiences with other programmers.
In short, the TIOBE index seems to generally reflects the
"popular programming languages" I see around me.
I'm giving others the chance to convince me otherwise.
David Erbas-White
> Michael C. wrote:
>
>> Although, no one here has convinced me that the TIOBE index is not
>> good or "good enough".
>
> Indeed -- and no one has convinced me that I shouldn't be a movie star.
> ;-)
>
We all thought that was so obvious it didn't need to be stated (ducking
<G>).
David Erbas-White
Brian Evans
Any statistic that uses proxy data over time will also show any
other trends between the proxy and what your trying to measure. It
can be hard to separate out and it's common to just ignore it as too
much work or assume it's not significant enough to matter.
From reading how TIOBE is calculated (http://www.tiobe.com/tpci.htm)
it's very much prone to showing these other trends. An example would be
link farms using popular search terms or copying content of
popular sites which could amplify the rating for popular languages.
Brian
Nick Hodges (CodeGear)
> In short, the TIOBE index seems to generally reflects the
> "popular programming languages"
I'm not sure that anyone is disputing that.
Nick Hodges (CodeGear)
> We all thought that was so obvious it didn't need to be stated
> (ducking <G>).
I care not what you all think. /I/ am not convinced. ;-)
Paul Nichols [TeamB]
> Personally I find the TIOBE index reasonable based on my own observation
> around me.
> I do find the following reasonable:
>
> Java is the most popular language.
>
> The most popular languages are in this order:
> JAVA, C, Visual Basic, and C++.
>
> Python ( a cross platform language ) is
> more popular than Delphi.
>
> Javascript is twice as popular as Delphi.
>
> Delphi isn't a very popular language.
>
>
> <snip>
>
OK using your rationale, C# is not a popular language either, since it
is way down on the list and percentages are not very high.
Although I would agree with the first two (Java and C), there is no way
I could see VB being number 3. Of course, the (Visual) is in
parenthesis, so I do not know if this means all Basic and Visual Basic
or just VB. I could see all Basics at number 3. (Visual, Gambas, Real
Basic,and VB Script or VBA).
Go up a few and read what I stated about interpretation starting with
8/7/2007. Think this puts these readings in a more realistic light.
Rudy Velthuis [TeamB]
> David Erbas-White wrote:
>
> > We all thought that was so obvious it didn't need to be stated
> > (ducking <G>).
>
> I care not what you all think. I am not convinced. ;-)
Think of this: if you should have been a movie star, you'd probably
already be one, by now. <eg>
--
Rudy Velthuis [TeamB] http://rvelthuis.de
"Sterling's Corollary to Clarke's Law: Any sufficiently advanced
garbage is indistinguishable from magic."
Rudy Velthuis [TeamB]
> "Rudy Velthuis [TeamB]" wrote:
> > Nick Hodges (CodeGear) wrote:
> >
> > > I care not what you all think. I am not convinced. ;-)
> >
> > Think of this: if you should have been a movie star, you'd probably
> > already be one, by now. <eg>
>
> I hear that he got a star role in the new Simpson Movie and John
> Kaster does his voice-over. :)
I saw it. I only saw one character that looked a bit like Nick, but
AFAIK, Nick doesn't smoke. <g>
--
Rudy Velthuis [TeamB] http://rvelthuis.de
"I can write better than anybody who can write faster, and I can
write faster than anybody who can write better."
-- A. J. Liebling (1904-1963)
I.P. Nichols
> Nick Hodges (CodeGear) wrote:
>
>
> Think of this: if you should have been a movie star, you'd probably
> already be one, by now. <eg>
I hear that he got a star role in the new Simpson Movie and John Kaster does
his voice-over. :)
I.P. Nichols
> I.P. Nichols wrote:
>> "Rudy Velthuis [TeamB]" wrote:
>> > Nick Hodges (CodeGear) wrote:
>> >
>> > > I care not what you all think. I am not convinced. ;-)
>> >
>> > Think of this: if you should have been a movie star, you'd probably
>> > already be one, by now. <eg>
>>
>> I hear that he got a star role in the new Simpson Movie and John
>> Kaster does his voice-over. :)
>
> I saw it. I only saw one character that looked a bit like Nick, but
> AFAIK, Nick doesn't smoke. <g>
Maybe he only tried it this one time....
http://youtube.com/watch?v=wSE5hCPNpCg&mode=related&search=
Graham Stratford
Craig Stuntz [TeamB]
> They also acknowledge that there index is only an "indication of the
> popularity of programming languages."
Then why the false precision? TIOBE, when pressed, has acknowledged
the limitations of their method. And yet they take every opportunity to
pretend that it's something more than a search engine count.
False precision is lying.
> A person doesn't need "the best" ranking system in order to collect
> statistics. A person should, however, pick a ranking system that
> they think is good. If the people who create TIOBE choose Alexa,
> I believe they did it because they thought it was "good enough" for
> their index.
It doesn't matter whether thy thought it was good; there are very good
reasons to believe that it isn't, especially for technical readership.
Simply asserting that you think that they think it's OK doesn't change
any of this.
> I believe the TIOBE index is more accurate than a "rough guess."
> I think it's even better than "educated guess".
You've given no substantive reason for this, and appear to be just
ignoring the reasons why it's probably not true.
> So far, no one has convinced me otherwise.
Again, what you think doesn't change the facts of the matter.
--
Craig Stuntz [TeamB] · Vertex Systems Corp. · Columbus, OH
Delphi/InterBase Weblog : http://blogs.teamb.com/craigstuntz
Please read and follow Borland's rules for the user of their
server: http://support.borland.com/entry.jspa?externalID=293
FL
Dave Nottage [TeamB]
> I am sure you could make a good movie star, when I see your photo on
> the Delphi Wiki it has a certain 1980'ies Jack Nicholson mad man look
> to it...
>
> <http://delphi.wikia.com/wiki/Nick_Hodges>
>
> I can the movie poster now...Starring Nick Hodges as Jack Torrance in
> the Shining - redux ;-).
Even "Heeeeeeere's Johnny!!.." would be accurate <g>
--
Dave Nottage [TeamB]
Richard Foersom
> Indeed -- and no one has convinced me that I shouldn't be a movie
> star. ;-)
I am sure you could make a good movie star, when I see your photo on
the Delphi Wiki it has a certain 1980'ies Jack Nicholson mad man look
to it...
<http://delphi.wikia.com/wiki/Nick_Hodges>
I can the movie poster now...Starring Nick Hodges as Jack Torrance in
the Shining - redux ;-).
<http://www.imdb.com/title/tt0081505/>
Doei RIF
Q Correll
| I can the movie poster now...Starring Nick Hodges as Jack Torrance in
| the Shining - redux ;-).
|
| <http://www.imdb.com/title/tt0081505/>
<chuckle> Works for me. ;-)
--
Q
08/12/2007 17:49:51
XanaNews Version 1.17.5.7 [Q's salutation mod]
Nick Hodges (CodeGear)
> I am sure you could make a good movie star, when I see your photo on
> the Delphi Wiki it has a certain 1980'ies Jack Nicholson mad man look
> to it...
That picture is far too flattering. ;-)
Bob S
"Nick Hodges (CodeGear)" <nick....@codegear.com> wrote:
>Richard Foersom wrote:
>
>> I am sure you could make a good movie star, when I see your photo on
>> the Delphi Wiki it has a certain 1980'ies Jack Nicholson mad man look
>> to it...
>
>That picture is far too flattering. ;-)
This help any?:
http://img142.imageshack.us/my.php?image=jacktanicksp8.swf
8^)
James Gibbons
> Michael C. wrote:
>
>> They also acknowledge that there index is only an "indication of the
>> popularity of programming languages."
>
> Then why the false precision? TIOBE, when pressed, has acknowledged
> the limitations of their method. And yet they take every opportunity to
> pretend that it's something more than a search engine count.
>
> False precision is lying.
>
Language number 50 is 0.074%. If they limited it to the precision of the
results it would possibly be 0% and we wouldn't know how to compare it.
I believe they show 3 places because the smaller languages only give
reasonable percentages with 3 places.
Accuracy and precision are two entirely different things. It is quite
possible to present values to as many places as you want as long as you
provide a +/- accuracy range. As this is a non-scientific study,
accuracy is not really something they can calculate. I'm not sure how
you would even go about trying to even determine accuracy for something
like this. They do it for political polls but these methods are suspect
too. Even with hard scientific measurements there can be measurement
biases and improper assumptions that get past the analysis.
The only way to understand the accuracy of these results would be to
know more about the methods and this seems to be lacking from the results.
James
Nick Hodges (CodeGear)
> This help any?:
>
> http://img142.imageshack.us/my.php?image=jacktanicksp8.swf
LOL --
My real name is John Nicholson Hodges, so who knows?
John Jacobson
news:46bfdc31$1...@newsgroups.borland.com...
> My real name is John Nicholson Hodges, so who knows?
So we could call you Jack Nicholson?
John Jacobson
> Richard Foersom wrote:
>
>> I am sure you could make a good movie star, when I see your photo on
>> the Delphi Wiki it has a certain 1980'ies Jack Nicholson mad man look
>> to it...
>
> That picture is far too flattering. ;-)
Nick in movies? There's an interesting idea. He could play a left-wing pinko
hippie in Easy Rider! That would be a good test of his acting abilities.
Craig Stuntz [TeamB]
> Language number 50 is 0.074%. If they limited it to the precision of
> the results it would possibly be 0% and we wouldn't know how to
> compare it.
That is precisely the point. You *don't* know how to compare it (based
on TIOBE anyway), and they are lying to you to make it seem like you do.
--
Craig Stuntz [TeamB] · Vertex Systems Corp. · Columbus, OH
Delphi/InterBase Weblog : http://blogs.teamb.com/craigstuntz
Everything You Need to Know About InterBase Character Sets:
http://blogs.teamb.com/craigstuntz/articles/403.aspx
Q Correll
| This help any?:
LOL!
Good job! ;-)
--
Q
08/13/2007 09:59:54