[Taxacom] Zootaxa papers now in BioStor

[Mike Sadka]

Hi again Stephen

I guess I  must be incompetent at stating my case.  The only alternative
seems to be that you are wilfully trying to misunderstand me!  

>it is easier to generate wiki-type content from structured data than
the reverse

Yes - but my point is not that easiest is best, but that structured data
are capable of flexible use, and therefore represent a more efficient
use of resources.  A database can be used to support numerous different
applications of the data it contains - this is less true of a wiki, even
those such as Rod mentions.

I do have a different weighing machine, surely it's best to use as many
weighing machines as possible if you want the most reliable answer?  My
weighing machine is not weighing just what is visible from the outside
of an information system - it also evaluates the bits you aren't looking
at (which are actually what make it work in the first place).

Behind my point is a belief that achieve decent bioinformatics there
needs to be a synergy between those who understand the data domain
(biology) and those who understand how computers work!  Arguments based
on one aspect only are always going to miss the mark.  I'm heartily sick
of the ongoing antipathy some biologists seem to have for IT - it isn't
helpful and IMO it is holding back your discipline.

I'm not trying to promote a particular technology.  I'm suggesting that
in order to make decisions about information systems you need to
understand the technology as well as what you are trying to do with it.
And you need experts in both fields to get to the most appropriate
consensus.

Anyway, your responses indicate once again I've clearly failed to make
my point effectively, so I'll just admit failure end bow out now!

Cheerio, Mike

__________________________

Mike Sadka

Interactive Media

x5462

 


-----Original Message-----
From: Stephen Thorpe [mailto:s.thorpe at auckland.ac.nz] 
Sent: 11 February 2010 20:26
To: Mike Sadka; TAXACOM
Subject: RE: [Taxacom] Zootaxa papers now in BioStor

Hi Mike,

Maybe you should indeed have kept your mouth shut in the first place! At
least I am not obsessed with partisan arguments promoting technology!

>And is "easiest" the most important consideration anyway?  It's easier
for rogue taxonomists to erect spurious taxa than to do the job
properly, but that doesn't make it desirable

You should be a politician! You haven't quoted your own words that
prompted my response - let me remind you:

>it is easier to generate wiki-type content from structured data than
the reverse

It was you who was arguing that "easier is better" in the first place!

>*relatively* useless wikis are compared to structured databases (and
therefore that it is more *efficient* to spend effort on structured
datasets, even if that takes longer

This may be true in theory, but in practice useful information (from the
point of view of content) is appearing on Wikispecies now, whereas the
potential greater utility of more structured databases has yet to be
realised. How long should we wait? 1 year, 10 years, 100 years, ...

>So speed and quantity are more important to you that data quality and
how usefully and widely the data may be used?  Efficiency is more
important IMO. Surely, if you are going to spend your valuable time
entering data, you want it to be as widely and usefully available as
possible?  The best things ofte take a little longer

NO WAY did I indicate that speed and quantity are more important to me
than data quality, etc.! Data quality is my PRIMARY concern, and the
Wiki system facilitates it BETTER than other models IMHO! It is easier
to keep up-to-date, and crap data isn't locked in so can be fixed
easily, simply, and quickly (speed is important in relation to error
correction and updating). Maybe the best things take a little longer,
but how little is little?

>>I think only cladists find it useful!
>So are you completely anti-cladist?  I am no systematist, but that
seems a pretty extreme position for this century...  Does this not
support my comment about luddism?
Not completely - just enough to indulge in sarcasm about it now and
again! :)

>It's a bit like designing a car without any understanding of
engineering- it might look fantastic, but how well will it work and how
long will it last, and how easy will it be to fix, run, maintain or
operate?
On the other hand, if I have to go to work today - you still being in
the planning stage of designing my car because it might not suit me in
100 years time isn't much help ... Different factors need to be weighed
up and compromises made. The guiding principles IMHO are content,
quality and utility - but you seem to have a dirrefernt weighing machine
....

Cheers,

Stephen

________________________________________
From: Mike Sadka [M.Sadka at nhm.ac.uk]
Sent: Friday, 12 February 2010 3:30 a.m.
To: Stephen Thorpe; TAXACOM
Subject: RE: [Taxacom] Zootaxa papers now in BioStor

Stephen:

I don't really have time for this (so I guess I should have kept my
mouth shut in the first place!) but...



>Perhaps, but it is far easier (and far cheaper) to create wiki-type
content period (i.e., not from structured data, but from primary
taxonomic literature)

But is it efficient?  Arguably you are likely to get more use out of
your input effort putting the content of primary taxonomic literature
into a structured database.

And is "easiest" the most important consideration anyway?  It's easier
for rogue taxonomists to erect spurious taxa than to do the job
properly, but that doesn't make it desirable.

In any case, your response misses my point.  My point is not how easy it
is to create wikis, but how *relatively* useless wikis are compared to
structured databases (and therefore that it is more *efficient* to spend
effort on structured datasets, even if that takes longer).  Structured
datasets can be used to output wiki-type content AND to keep the data
much more flexibly so that you can do other things with it too.
Generally (Rod's previous points accepted) wiki content is just wiki
content - whereas a well designed data model should support any
logically appropriate use of the data it contains, not just generation
of static/flat and hard-to-search-or-manipulate content.

That said, I would probably support a plurality of approaches for
different circumstances.  Dogma is not a good thing in IT as in anything
else.




>If resources are limited (if???), I would spend them creating as much
useful information as possible, as quickly as possible

So speed and quantity are more important to you that data quality and
how usefully and widely the data may be used?  Efficiency is more
important IMO. Surely, if you are going to spend your valuable time
entering data, you want it to be as widely and usefully available as
possible?  The best things ofte take a little longer.

The point I am trying to make is that structured data are
probably/usually much more useful and more widely useful long term.  If
you'll forgive me saying so, I get the impression that many scientists
don't really understand information management, or how databases can and
should be used to improve efficiency in all sorts of ways - and why
should they, since IT isn't their speciality.  The sadness is that many
of them do not recognise this limitation of their own knowledge and
therefore do not seek appropriate informatics collaborators - or even
up-to-date and accurate knowledge of what is possible or appropriate.

Therefore experts often consider only half the equation - the part that
pertains to their data domain - but ignore the wider IT considerations,
and don't welcome or seek the synergy between domain specialists and
informatics technicians which might actually achieve an appropriate
result from all perspectives.



>I think only cladists find it useful!

So are you completely anti-cladist?  I am no systematist, but that seems
a pretty extreme position for this century...  Does this not support my
comment about luddism?



>At any rate, to cross threads just a little bit, EOL is no better off
than Wikispecies by these arguments ...

Yes - but it is not me who appears to be obsessed with partisan
arguments about wikispecies vs EoL!    ;-)

I wasn't supporting either against the other.  My point was that none of
your stated reasons for your position re use of wikispecies appears to
be based on any consideration of the *technological* implications.

It's a bit like designing a car without any understanding of engineering
- it might look fantastic, but how well will it work and how long will
it last, and how easy will it be to fix, run, maintain or operate?


Cheerio, Mike