Free software in the scientific field intends to see what kind of
software used in the educational and scientific institutions that immediately
led to descend down orders and instructions "to remove unauthorized software
from any computer." The author proposes a solution where old computer no longer
exists, and the new money has not yet been given. And not give it. Computer
scientists at the moment are rare in the software desert oasis. Official policy
is formulated in a simple, commercial products to replace them open source/freeware-
analogues. So Linux Windows, MS Office Open Office, Origin QtiPlot, etc. It is
suggested that a previous heart paradigm of software products. Perhaps this
painful step, but on the other hand, it is an excellent opportunity to look
forward and try to build a new concept for use on a clean slate. More so for the
past ten years, many excellent open products, practising different approaches to
dealing with text and data processing, etc., and products that existed before,
but appeared very weak, have a strong development. Leaving aside the questions
of office automation software (postal customers, utilities recording disks,
graphics editor), to draw attention to the program, which enuzhny to solve
scientific problems. What can be included in this category? First, a system for
recruiting and layout text for a large number of formulas, graphics, and
references. Here you can specify a system for creating presentation materials,
again containing formulas, specific images (such as chemical or Feynman diagrams)
and a bibliography. Next on the block-drawing graphs. And as a daily tool-set to
enable the complex mathematical calculations, but still experimental and
language development / modification of software to manage the pilot installation,
data collection and data processing. Next, briefly consider that could
constitute a paradigm of the proposed Word / Origin / C + +. This proceed to a
standard through-put requirements on and try to find the most responsible of the
product. Text What requirements apply to systems usually work with the text? *
stability; * small file; * compatible file formats as down and up; * an
opportunity to work with multiple documents that contain hundreds or even
thousands of formulas; * stable and easy work of the references (number of
formulas, page numbers, bibliography, links to the figures and tables); * the
introduction of graphics links; * separation of content and external reporting.
My experience with a variety of text editors and publishing shows that better
LaTeX in this area is nothing. The most popular distribution LaTeX for Windows (OS
change and who hardly dare buy-still easier) MikTeX. No WYSIWYG, very different
approach to a wide range of documents and commands at first glance significantly
more complicated system. But this impression is misleading. Just do on a small
and a little effort to work with a bulky document physical-mathematical focus,
as the benefits of sharing content and approach are becoming evident. With clean
hands ASCII file of 100 pages, you can easily hold over horror in terms of a la
Word operation : to change the formula, delete paragraphs, insert /
udalyatkartinki, renumbering bibliography, etc. And integration with version
control system (that draw attention to Subversion, subversion.tigris.org, and
its Windows client TortoiseSVN) can save a document of history, work together on
a large article or book written back to the past and choose the best wording.
More important, and that the work with the bibliography built in LaTeX virtually
seamless manner. Since the number of links to bibliographic database rarely
exceed a few thousand, and speed access to records should not play, the storage
database to a text file (bibtex) falls entirely within the rules of the core
material, recruited in TeX. Moreover, there are several products that can work
with bibliographic databases using convenient interfaces. In particular,
deserves the attention of a remarkable package JabRef. Presentations will now
turn to a product presentation material. As head of the majority of users at
once a monument inscribed with great PowerPoint. Undoubtedly, the presentation
of PPT already become a de facto standard. However, a clear division of content
and presentation in PowerPoint does not. Difficulties redesigning the
presentation made to derogation from the basic style and layout (99% of cases),
the formula continues system models require enormous hand with minor design
changes presentations. Inconvenient animation tools do not do anything else
tells of documents containing mass formulas. What can be offered as an option?
It may seem paradoxical, this is still a LaTeX, but with three additional
packages : beamer, and pscyr pgf + TikZ (latest version pscyr, pgf and xcolor
must take a distribution MiKTeX or home pages on sourceforge connects with the
beamer are older versions). Source presentations in ASCII format, the content of
the rigid division of design, the possibility of multiple use their own crews to
remove blocks of text, graphics or formulas material support poslaydovoy
animation by raskadrovki text, pdf opportunities for integration with external
sources (exe, avi, etc.) to prepare a presentation with lots of formulas and
images in a matter of hours. It is balanced by mixing files document will look
professional. If necessary, change the design presentation takes place over one
or two minutes to select a new style file, and recompile the source document.
You can also read the contents of the presentation in print : screens screens
with commentaries for each frame. A language TikZ drawing illustrations, which
is, in fact, LaTeX package, allows graphics directly to the original document
presentation (in place or in the form of commands depending on frequency of use).
This language can create illustrations, using concepts of the subject area,
whether the nucleus and electrons, or boxes and ribs, or photon, and another
photo ... Moreover, the schedule TikZ, skooperirovannaya opportunities Output
animation beamer, can be beautiful, elegant things through low cost. Mathematics
And, the last part of the basic job-researcher packages for mathematics, graphic
display of results and management system for experimental complexes. In this
segment is referred to the popular programming language, belonging to the family
dynamic languages Python. Interesting language itself (for a discussion of
possible topics for a long interview), namely in the context of the task.
Strengths Python is that he supported OOP- paradigm can just easily be used as a
calculator, and a scripting language glue / batch processing, and as a language
for software and hardware management systems. The large number of packets to a
great many tasks significantly easier. For mathematical calculations lots of
specialized programs (such as wiki.python.org / moin / NumericAndScientific, www.enthought.com),
there assembly containing large quantities of specialized packages (such as code.enthought.com
/ enthon). There are separate packages for drawing graphs, such MatPlotLib. This
is a very convenient option, especially when viewed link "experimental setup +
document." Data from the pilot plant (or experiment) to be administered on a
more Python, are pre-processed (again Python), the information given in the form
of graphics files (2D, 3D, other formats), and then using ODBC scripts partially
formed and started to compile LATEX paper, using these graphics. As a result,
fully automatic mode can be obtained at the outlet professionally created PDF
file with the records of the experiment. Also, using Python's easy enough to
write GUI to various programs (such as packages TkInter, wxPython, TraitsUI).
And Rudukhina more. I think reasonable to pause and give readers the opportunity
to go on unfamiliar links, and more about those materials. Of course, all of
these products easy to learn, but the effort expended on the acquisition thereof,
that is. Taking advantage of the vacuum created, you can not try to recreate
earlier by the former environment erzatsev search, and try to change the vision.
This task alone challenging seriously, but on the other hand, there is an
enormous amount of technology and products to bring together various people in
the same direction. And they fit very well into the structure of the scientific
community, complementing and enhancing the existing ties between scientific
groups. Action team, which is the nucleus of informal community, will change the
paradigm of a painless way. For this, it is a system of positive feedback, and
once started, can support herself. As a result of such a group can demonstrate
the resource "TeX Institute of Mathematics and Mechanics UD RAS." He nearly
fourteen years (eight of them devoted to work with MiKTeX), and the history of
its development is a good example of how to successfully introduce new
technology in an academic environment. Why do so many readers probably say that
there are many wonderful packages and products that address these challenges.
That is true. Unfortunately, if begin to describe all possible combinations of
products, taking into account the releases, and especially integration, we get
multivolume work with maloupotrebimymi results, as a result of rapid
obsolescence. Therefore, the article suggested only approach to the challenges
and do the packages and products, which were chosen on the basis of personal
experience and a pilot were tested in a variety of areas, not just in science.
1. Why start with LaTeX? Because the word is the most frequent tool in the
parlance researcher. This can be used to receive wages and work measurement. And
now-continuous grants / reports / articles / presentations and other orgveschi
to other members of the group to carry out scientific research. LaTeX chosen as
a tool for writing articles, training presentations. Moreover, since the input
files are understandably ASCII format, automated system of filling documents are
written very easily. And done it with script languages. 2. Why this is a Python
rather than C + +. All simple. With the experience of industrial C + +, I can
imagine what the overheads associated with its use. What requirements should
submit to a programming language for scientific research? Experience shows that
the following list of close to the optimum : * unambiguous language constructs,
transparent syntax; * Ease of understanding, learning curve is acceptable; *
Krossplatformnost; * Flexibility; * Compact programs; * Support for the
scientific community *; * A wide range of libraries; * Concealing technological
complexity (COM, working with XML, lists, hash tables, Work strings, books); *
an opportunity to develop equal ease as CLI glue, and the GUI; * ease of
deployment; * Support OOP- concepts; * speed extradition finished code (preferably
avtotestami); * to introspection. Having experience with C + +, Java, Perl,
Python, I stopped at the last. And there is a lot of what was done. Amazingly
easy it is to generate Excel- Records (with paint and formatting), not knowing
glubinno of OM. Since XML is very comfortable ... and the mass of other things.
3. At Matplotlib light wedge not Tsiolkovsky. Packages much, but I say on the
specific decision, which I collected (appreciating the many ways). To list all
the packages in a popular article, perhaps nowhere. I would describe the general
approach to create ARM researcher. The idea was not to scuttle a package of the
package, and that can be collected by an instrument to further sharpen our
skills in specific areas. I do not insist on a specific package, I am talking
about the concept. People who decide to use an open source, in one way or
another should be involved in the community and look around carefully.
|