QSCGZ
Posts:
175
Registered:
12/12/04
|
|
Re: why .ps ?
Posted:
Aug 5, 2000 3:47 PM
|
|
now I give all my answers in one post :
-----------------------------------
Ilya Kapovich wrote:
>Oh, come on, what's this bull about elitism etc?
can you _argue_ against it ?
>Are you a conspiracy theorist or something?
Maybe a bit. But no more than justified , I think.
>.ps is a standard product of TeX.
>TeX files are often also made available and they are in fact ASCII files and
>can be used to perform a text search.
yes, that's (partly) true for pure text. But then some text parts are
in different fonts , and formulas or even inclosed variable names
often aren't included.
>And TeX is used because it produces
>properly typeset mathematical texts and formulas (yes, in .ps format) which
>can be printed out and submitted to a journal or sent to another person.
it takes longer with modem and can't be read and answered online ,
like other emails.
>You don't really want to see $\int_0^5 \frac{x^2}{x+1} \, dx$
>printed in a book or an article, do you?
no, but I{0..5} x^2/(x+1)*dx would be OK.
And I wasn't talking about books but online articles.
>That's what TeX macros for math formulas look like
>before TeX files are compiled and converted into .dvi and then .ps
>
>I could agree that more mathematicians should convert .ps files into .pdf
>files, which are much smaller and printable on more printers.
--------------------------------------------
Steve L wrote:
>On 5 Aug 2000, QSCGZ wrote:
>>
>> why is most mathematical online-information only available in
>> .ps or .dvi encrypted format ???
>
>Because it's easy to form a .ps or a .dvi file from a LaTeX
>file. Besides, it's not really encrypted -- just download any of a number
>of free viewers (like Ghostview).
>
>> I assume , that .ps can handle pictures , greek letters , integral
>> symbol etc. , but there are usually ways to do this in ASCII as well ,
>> with only small decrease of readability.
>
>Including graphs of results, or complicated mathematical expressions, is
>easier with .ps files.
>
>> Most .ps files could also be written in text format , but unfortunately
>> text versions are usually not made available.
>
>Ask the person who holds the .ps file if they can send you the LaTeX file.
>
>> e.g. the sci.math posts here come in .txt format. I've never seen
>> .ps versions of sci.math posts attached , so it seems possible to do it
>> in .txt.
>>
>>
>> - .ps files cannot be searched for keywords.
>> - .ps files cannot be viewed online (AFAIK)
>
>Ghostview.
how long will it take me to learn this ?
what is the probability that it will work on my system ?
>> - it's difficult to download and install the (huge and complicated)
>> .ps-viewing software.
>
>Ghostview took about two minutes to install, and it wasn't very difficult
>at all.
not for me. And AFAIR it's in the MB range , so it took at least 5 min. for
download. And then I couldn't get a proper format setted for viewing/
printing when I tried , so I usually try to ignore .ps ,.dvi , .tex
files or read them undeciphered.
>> - .ps files are much larger , which is bad for downloading and archiving.
>
>Then ask for .tex files.
send emails , wait for answers (if any) , 20% of people would complain ,
as they don't like qscgz-names etc.
And .tex is still worse than .txt IMO.
>> - printing .ps files takes much longer on many printers (I still
>> don't know, how to print them in suitable format on my printer)
>> - good conversion software ps-->txt is not available AFAIK
>
>Ghostview, Edit->Text Extract.
OK , there are 81 executables in my \ghost directories , which is ~20MB.
But no ghostview.exe.
I tried gsview.bat - nothing.
Then tried gs.exe and got an error : interpreter doesn't match gs.init
or such. Then tried gs2.bat , gs3.bat invain. Then I lost patience.
Don't want to read the documentation now.
But I remember, I once used it with -partial- success , after lots
of learning time. Still couldn't get the correct format for the
printer AFAIR and it took long to print.
>> it seems to me , that mathematicians use .ps as an attribute of elitist
>> behaviour ; to make it difficult for outsiders to enter their domain.
>
>No, it's more that ways to process LaTeX files into dvi (which are
>supposed to be, and AFAIK are, DeVice Independent) and from dvi into ps.
.ps and .dvi are usually converted LaTeX files ? I didn't know that.
But then , why don't they let the downloader convert them , if he want's to ?
>> What we need is a good,free .ps-->.txt conversion software , and more
>> mathematicians who are willing to provide their papers in .txt format.
>
>You can't include figures and diagrams in .txt format.
the rest would be just fine. To my experience diagrams and figures
usually contribute only some few % to the understanding of math-articles.
But I admit , I'm mainly interested in combinatorics,programming which isn't
very picturous.
And pics are often included for better appearance.
sci.math e.g. can live pretty well without pictures.
>> They can still provide them in .ps format too , if they really want.
>> But .txt should be the standard.
>
>No, LaTeX .tex files should be standard ...
If .txt can be converted to .tex and .tex can't be converted to .txt ,
then how can .tex be better ? (except for graphics)
>and that's what journals request IIRC.
nonelectronic journals will die in the long run. Seaching them
is too difficult. Storing them takes too much room. Mailing them is
expensive and slow. Copying them is difficult.
electronic journals with nonsearchable files will die in the long run too ,
or people will only read the abstracts , provided they are searchable.
> .ps files are graphical representations of the .tex files.
----------------------------------------------
Patrick wrote in parts:
> However, the Acrobat Reader is free from Adobe.
I once tried that too. Took me hours (download errors , ) ,
failed at last , don't remember why , finally removed it.
>If the author of the pdf file did not set protection against copying the
>text, the text can be copied via the clipboard.
Steve L wrote in parts:
>... but now try to write a matrix,
>let's say a 50-by-50 ... try to write _that_ in plain text without getting
>some confusion.
_I_ should have given that example. These things are garbled with TeX :
the columns aren't parallel.
>> pdf files can be converted to .txt ?
>Probably.
>You can search in a ps file with GSview.
but do yahoo,altavista etc. know allow that ?
-------------------------------------------------
Hong Ooi wrote:
>>I{0 to oo} exp(-x^2/2)*dx = sqrt(pi/2)
>>or such works fine for me.
>
>Now:
>- does that I{} thingy mean "integrate" or the Identity function?
>- do the curly brackets denote a set, or do they have some sort of
>special meaning?
maybe [] are better !?
>- do the letters "pi" mean the mathematical constant 3.14159+, or the
>multiplication of two variables, p and i?
use p*i for that
>- similarly does "dx" stand for the differential element, or two
>variables d and x?
>- similarly, couldn't you replace the "oo" with "o^2"?
hmm, maybe ascii 236 : ÃÂì
>You may argue that for this example such things are clear from context,
yes
>but there'll be lots of situations where they _aren't_ clear.
>Typesetting mathematical formulae so that the meaning is clear (or at
>least isn't needlessly obfuscated) is not a trivial job.
but that makes communicating with computer difficult.
How handle Mathematica,Maple,.. these ?
Some conventions are always needed. Most papers begin with some.
I often saw oo,I,pi in sci.math and noone complained.
You still can use infinity,integral,from,to in situations where you're
concerned about clearness.
----------------------------------------
Steve L wrote:
>What's I? What's oo? What if I need to draw the graph of some
>complicated function and want a picture of that in the paper ... I'd use
>Matlab or the like to graph it and include it in the file.
you still can use TeX or whatever you want for that. I don't want you
to give them up. But even the text-part in .ps etc. often is garbled.
I tried pstotxt and often get spaces like "num b er" etc.
BTW , how would you do it in sci.math ?
>Or let's say I have a set of data points (x_i,y_i) and I want to show that
>x and y are correlated?
compute the correlation coefficiant
>> hmm, let me see : They _do_ want to publify but _don't_ want to
>> be copied,cited,changed, (partly) used within other's work ,...
>
>Copied, changed -- yes, they want to prevent this. Cited -- no, you can
>cite a ps paper just as easily as you can cite a normal paper.
>
>> If s.o. wants to change it , he can still do , just a little
>> bit more work. But then fewer people will suspect the changes.
>>
>> >So that is why amongst mathematicians dvi and postscript
>> >became standard - really an accident of history.
>>
>> so,let's change it !
>
>You haven't given a good reason why we need to do so.
I gave 7 : keyword search,viewed online,no SW necessary,smaller files,
faster printing,convertable,computer readable.
>It's no harder to create .tex files than it is to create .txt, and .tex
>files can do more than plain text files [If you want your paper to look
>_exactly_ like this, or be readable on just about any computer, TeX can do
>it.]
...with suitable additional software installed
>> btw., how long does it take to find , download , install these programs
>> and to get familar with them for proper usage ?
>
>20 minutes, give or take.
>
>> There are whole newsgroups full of problems with them.
>
>If you're looking for something that _doesn't_ have problems, you're not
>going to find it. There is _always_ a problem with _anything_ -- but such
>is life.
but sometimes more and sometimes less.
>>>This program will also allow you to view postscript online.
>>
>> within AOL browser ?
>
>Possibly. I haven't had any experience with AOL's browser.
>
>> proof it. Why aren't they distributing .txt files as well ?
>
>Because you can't _do_ some things you can do with .tex by using .txt
>files ...
but you can _do_ some(most) of these things. I wouldn't care , if some
graphics were missing.
>or it would be a whole lot more confusing to do so. [large
>matrices, including figures, etc.]
-----------------------------------------
Steve L wrote:
>> then . how long would it take for someone
>> not familiar with them to get them ready
>> for use ? Including searching for the right files, downloading,installing,
>> reading the manual , etc.
>
>Searching for the right files -- a minute or two with Altavista.
>Downloading -- 10 minutes, maybe 15. That's not terrible -- and you can
>do it in the background while you're doing other things.
>Installing -- I don't know, but not _that_ long. If you just want to view
>files you don't really need to read the manual, the installation should
>associate the appropriate files automatically (or IIRC you check boxes in
>a form in the installation.)
thanks for your answers.
that's still considerable for nonmathematicians , who only search
for one special paper or one special subject.
------------------------------------------------
Steve L wrote:
>Another poster already give one reason why .tex files are not released --
>it's too easy to change them and claim that the file is yours.
wouldn't that be illegal or at least immoral ?
What , if someone converts to .txt or .tex with proper acknowledging ?
>Don't worry about (future) ai-programs ... by the time we have an
>ai-program that can work with general mathematical concepts we should be
>able to have it read in .ps files (or we will have found something
>better.)
maybe the ai-program only handles special concepts.
maybe we'll have dozends of .ps - like formats then.
maybe they want to read in and verify (parts of) "old"
papers from 2000.
>> >Just as 50 years
>> >ago every mathematician was simply assumed to be able to read
>> >French, German, and Italian,
>>
>> I doubt this. Certainly not _every_ mathematician.
>
>Fine. 'almost every' mathematician then.
>
>> And do you think, that this has changed now ? and why ?
>
>No, it hasn't.
>
>> Is TeX assumed to replace these languages ?
>
>No. If you want to read a paper in a foreign language, knowing LaTeX will
>allow you to get the structure but none of the content.
>
>> >so today every mathematician is assumed
>> >to use TeX for promulgating his work.
>>
>> or .ps or .dvi or pdf or MathML or ...
>
>Once you have a LaTeX document, there are MORE than enough programs to
"MORE than enough" is exactly the problem , I was adressing..
>convert it to dvi, ps, pdf, etc. Theoretically, I could read a document
>in LaTeX and get the information from there ... but it would be MUCH
>easier to read the Postscript version obtained from that .tex file.
once you've installed it , got familiar with it , are offline , don't
have to quote from it (e.g. to sci.math) , don't have to print it ,
search it , ... - agreed (except "MUCH")
---------------------------------------------------
seems that everyone disagrees with me on this ?
--qscgz
|
|
