Community sizes

parnell wrote:
>> On the contrary, the evidence you just cited shows that hundreds of
>> Haskell packages are being created and the evidence I cited shows that
>> all but two (Darcs and HPodder) remain unused.
>>
>> Surely the only logical explanation is that these Haskell packages are
>> failures, presumably because they are useless.
>
> The evidence you cited is based on a flawed premise: that all installs
> of the libraries in question will be recorded by the Debian and Ubuntu
> package popularity contest.

That is wrong. In reality, I assumed there were no significant systematic
errors that would bias these results, i.e. OCaml users are not far more
likely than Haskell users to subscribe themselves to popcon. This is a
reasonable assumption. Moreover, the results correlate with completely
different metrics such as Google Trends and job demand.

> Because that is not true then your conclusion is flawed.

By that logic, all conclusions based upon finite experiments are flawed.

>> They are certainly imperfect but they are among the best indicators I
>> have found. Google Trends is another good one. The Windows platform is
>> the obvious omission.
>
> Google fight is more fun:
> http://www.googlefight.com/index.php...word2=Haskell+

The problem is, of course, that "Haskell" is a homonym. Indeed, you cannot
even do a job search to reliably ascertain the number of OCaml vs Haskell
jobs because there are always several jobs related to the word "Haskell" in
another context (typically someone's name).

Try a Google Fight between uniquely-named compilers instead:

http://www.googlefight.com/index.php...mlc&word2=ghc6

--
Dr Jon D Harrop, Flying Frog Consultancy
http://www.ffconsultancy.com/products/?u

Ben Franksen wrote:
> Jon Harrop wrote:
>> Haskell programmers will almost all have either hugs or ghc6 or both
>> installed, so there are between 6,606 and 9,794 of them.
>
> I am using ghc-6.8.3 which is not yet in my distro. Also, one should not
> forget that there are people who do not take part in such surveys. That
> means the real numbers are higher (it is unclear how much).

Well, then you should also consider that many people will have installed
these language implementations and almost never used them. For example, I
have several Haskell implementations installed on several machines but I
very rarely use any of them.

--
Dr Jon D Harrop, Flying Frog Consultancy
http://www.ffconsultancy.com/products/?u

On Aug 22, 6:09 am, Jon Harrop <j...@ffconsultancy.com> wrote:
> parnell wrote:
> >> On the contrary, the evidence you just cited shows that hundreds of
> >> Haskell packages are being created and the evidence I cited shows that
> >> all but two (Darcs and HPodder) remain unused.

This is flaw in your logic, I and many others use many of those
libraries listed on OhlOh. Because the debian and Ubuntu packages tend
to be behind the latest (as was mentioned by an earlier poster) and
because there are several other avenues (possibly more popular) that
one can use to install and use these packages your conclusion that
"all but two (Darcs and HPodder) remain unused" is a flawed
conclusion.


> That is wrong. In reality, I assumed there were no significant systematic
> errors that would bias these results, i.e. OCaml users are not far more
> likely than Haskell users to subscribe themselves to popcon. This is a
> reasonable assumption. Moreover, the results correlate with completely
> different metrics such as Google Trends and job demand.
>
> > Because that is not true then your conclusion is flawed.
>
> By that logic, all conclusions based upon finite experiments are flawed.

Your conclusions remind me of the conclusions of the proponents of the
miasma theory of the spread of cholera in 19th century London. They
looked at elevation and distance from the Thames and drew conclusions
that the feculent air rising from the Thames was killing people. It
was not until John Snow's much more thorough study involving location,
elevation, class, and which company supplied the water that bad water
was found to be the real culprit. In short I believe that your
measurement needs a few more inputs.


> The problem is, of course, that "Haskell" is a homonym. Indeed, you cannot
> even do a job search to reliably ascertain the number of OCaml vs Haskell
> jobs because there are always several jobs related to the word "Haskell" in
> another context (typically someone's name).
>
> Try a Google Fight between uniquely-named compilers instead:
>
> http://www.googlefight.com/index.php...mlc&word2=ghc6

You are very correct about the "Haskell homonym" problem. Your google
fight is a much better metric than my google fight. But this seems to
bear out what we all really "know" that both languages have rather
small followings of very dedicated (not really the right word)
individuals. And that many of those individuals (myself included)
should be included in both camps.

Parnell

parnell wrote:
> On Aug 22, 6:09 am, Jon Harrop <j...@ffconsultancy.com> wrote:
>> parnell wrote:
>> >> On the contrary, the evidence you just cited shows that hundreds of
>> >> Haskell packages are being created and the evidence I cited shows that
>> >> all but two (Darcs and HPodder) remain unused.
>
> This is flaw in your logic, I and many others use many of those
> libraries listed on OhlOh. Because the debian and Ubuntu packages tend
> to be behind the latest (as was mentioned by an earlier poster) and
> because there are several other avenues (possibly more popular) that
> one can use to install and use these packages your conclusion that
> "all but two (Darcs and HPodder) remain unused" is a flawed
> conclusion.

In other words, I cannot disprove the existence of a huge number of
invisible people silently using Haskell software. While true, that is not
useful.

>> > Because that is not true then your conclusion is flawed.
>>
>> By that logic, all conclusions based upon finite experiments are flawed.
>
> Your conclusions remind me of the conclusions of the proponents of the
> miasma theory of the spread of cholera in 19th century London. They
> looked at elevation and distance from the Thames and drew conclusions
> that the feculent air rising from the Thames was killing people. It
> was not until John Snow's much more thorough study involving location,
> elevation, class, and which company supplied the water that bad water
> was found to be the real culprit. In short I believe that your
> measurement needs a few more inputs.

You can say that of all measurements, so it conveys no information.

--
Dr Jon D Harrop, Flying Frog Consultancy
http://www.ffconsultancy.com/products/?u

> In other words, I cannot disprove the existence of a huge number of
> invisible people silently using Haskell software. While true, that is not
> useful.

They are not invisible, you have just chosen not to count them.

>You can say that of all measurements, so it conveys no information.

Not at all, your measurements are far too narrow to conclude that the
libraries in question are unused. Drawing the conclusion that the
OCaml libraries are more popular from your data is a valid conclusion,
since theoretically all libraries in question have other avenues
through which they can be installed and so if a more comprehensive
analysis were to be done one would think that both the Ocaml and
haskell libraries in question would get a boost. But saying that the
Haskell libraries are unused and broken based on this data is like
saying the emperor penguin is extinct because I could not find any
living in the Falkland Islands.


Parnell

Ben Franksen wrote:

> I am using ghc-6.8.3 which is not yet in my distro. Also, one should not
> forget that there are people who do not take part in such surveys. That
> means the real numbers are higher (it is unclear how much).

Same here. Ubuntu, at least, is chronically behind the times with haskell
(ghc) packages. Anyone who cares about using an up-to-date
compiler/libraries is pretty much precluded from using the Ubuntu packages.
I can't speak to the debian situation.

parnell wrote:
>> In other words, I cannot disprove the existence of a huge number of
>> invisible people silently using Haskell software. While true, that is not
>> useful.
>
> They are not invisible, you have just chosen not to count them.

Then where are they?

>>You can say that of all measurements, so it conveys no information.
>
> Not at all, your measurements are far too narrow to conclude that the
> libraries in question are unused. Drawing the conclusion that the
> OCaml libraries are more popular from your data is a valid conclusion,
> since theoretically all libraries in question have other avenues
> through which they can be installed and so if a more comprehensive
> analysis were to be done one would think that both the Ocaml and
> haskell libraries in question would get a boost. But saying that the
> Haskell libraries are unused and broken based on this data is like
> saying the emperor penguin is extinct because I could not find any
> living in the Falkland Islands.

I can point you to emperor penguins but you still have not pointed me to
hundreds of thousands of users of open source Haskell software.

--
Dr Jon D Harrop, Flying Frog Consultancy
http://www.ffconsultancy.com/products/?u

On 22 ago, 15:12, Jon Harrop <j...@ffconsultancy.com> wrote:
> I can point you to emperor penguins but you still have not pointed me to
> hundreds of thousands of users of open source Haskell software.

I think all is about to change:
http://www.realworldhaskell.org/blog...-now-complete/

A 700 page online book about Haskell in real-world usage?
I guess your OCaml books are in for some serious competition.

see more:
http://groups.google.com/group/fa.ha...dbf0311d22e99#

Expect many people becoming enlightned on how to effectively cope with
real world problems using Haskell...

namekuseijin wrote:
> On 22 ago, 15:12, Jon Harrop <j...@ffconsultancy.com> wrote:
>> I can point you to emperor penguins but you still have not pointed me to
>> hundreds of thousands of users of open source Haskell software.
>
> I think all is about to change:
>
http://www.realworldhaskell.org/blog...-now-complete/
>
> A 700 page online book about Haskell in real-world usage?

I also have high hopes for that book.

> I guess your OCaml books are in for some serious competition.

OCaml for Scientists maybe but I doubt Real World Haskell will touch F# for
Scientists. :-)

> see more:
>
http://groups.google.com/group/fa.ha...dbf0311d22e99#
>
> Expect many people becoming enlightned on how to effectively cope with
> real world problems using Haskell...

Only time will tell...

--
Dr Jon D Harrop, Flying Frog Consultancy
http://www.ffconsultancy.com/products/?u