www.digitalmars.com         C & C++   DMDScript  

D - linux

reply "Walter" <walter digitalmars.com> writes:
You're probably all wondering why I haven't done an update in a while. There
are two reasons:

1) I was sick for a couple weeks. Ugh.

2) I'm working on a linux version. dmd is working under linux now, but the
runtime library needs much work. Work is frustratingly slow on it, mainly
because of my inexperience with linux and the dearth of good information on
how the internals of file formats, etc., actually work.
Apr 23 2003
next sibling parent Russ Lewis <spamhole-2001-07-16 deming-os.org> writes:
ALLRIGHT!!!

I've been using dli for forever, and have been looking forward to re-integrating
with dmd.  I just tried to use dmd with wine; it's possible, but not easy.

Thanks for the work!
    Russ

Walter wrote:


 You're probably all wondering why I haven't done an update in a while. There
 are two reasons:

 1) I was sick for a couple weeks. Ugh.

 2) I'm working on a linux version. dmd is working under linux now, but the
 runtime library needs much work. Work is frustratingly slow on it, mainly
 because of my inexperience with linux and the dearth of good information on
 how the internals of file formats, etc., actually work.


--
The Villagers are Online! villagersonline.com

.[ (the fox.(quick,brown)) jumped.over(the dog.lazy) ]
.[ (a version.of(English).(precise.more)) is(possible) ]
?[ you want.to(help(develop(it))) ]
Apr 23 2003
prev sibling next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
 1) I was sick for a couple weeks. Ugh.


Get well soon!
It was not SARS I hope?!

Jan
Apr 23 2003
parent "Walter" <walter digitalmars.com> writes:
"Jan Knepper" <jan smartsoft.us> wrote in message
news:3EA70C20.12AF234A smartsoft.us...

 1) I was sick for a couple weeks. Ugh.


 Get well soon!
 It was not SARS I hope?!


I wasn't bedridden or anything, just down enough that I was demotivated from
doing anything but watching Seinfeld on my Tivo.
Apr 23 2003
prev sibling next sibling parent reply Jan Knepper <jan smartsoft.us> writes:
 2) I'm working on a linux version. dmd is working under linux now, but the
 runtime library needs much work. Work is frustratingly slow on it, mainly
 because of my inexperience with linux and the dearth of good information on
 how the internals of file formats, etc., actually work.


While you're at it... Any change you might have a DMC++ for Linux with it too?

Jan
Apr 23 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Jan Knepper" <jan smartsoft.us> wrote in message
news:3EA70C58.34099300 smartsoft.us...

 2) I'm working on a linux version. dmd is working under linux now, but




the

 runtime library needs much work. Work is frustratingly slow on it,




mainly

 because of my inexperience with linux and the dearth of good information




on

 how the internals of file formats, etc., actually work.


 While you're at it... Any change you might have a DMC++ for Linux with it


too?

They do share quite a bit.
Apr 23 2003
parent reply Garen Parham <nospam garen.net> writes:
Walter wrote:


 They do share quite a bit.


If I could use DMC++ on Linux I would do so right away.  You'd also have a 
big Gentoo advocate over here.  A common complaint is that compilation takes
so long.  Another is the often crappy code that GCC generates.  DMC++ doesn't
seem to have those two problems IME.  

I currently have a distributed compilation system setup with distcc that
uses separate windows systems as 'compilation nodes' with gcc cross compilers
just to speed it up because it's so damn slow. :)

/me dreams of building/testing the whole system in under 24hrs without
requiring 5+ computers.
Apr 24 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Garen Parham" <nospam garen.net> wrote in message
news:b88t3n$1r84$1 digitaldaemon.com...

 Walter wrote:


 They do share quite a bit.


 If I could use DMC++ on Linux I would do so right away.  You'd also have a
 big Gentoo advocate over here.  A common complaint is that compilation


takes

 so long.  Another is the often crappy code that GCC generates.  DMC++


doesn't

 seem to have those two problems IME.

 I currently have a distributed compilation system setup with distcc that
 uses separate windows systems as 'compilation nodes' with gcc cross


compilers

 just to speed it up because it's so damn slow. :)

 /me dreams of building/testing the whole system in under 24hrs without
 requiring 5+ computers.


GCC is pretty slow. One big problem with putting DMC on linux is it would
need to be binary compatible with the way GCC does things. There's a lot of
undocumented stuff there.
Apr 24 2003
next sibling parent reply Charles Banas <greywolf greyfade.net> writes:
On Thu, 24 Apr 2003 11:10:18 -0700, Walter <walter digitalmars.com> wrote:


 GCC is pretty slow. One big problem with putting DMC on linux is it would
 need to be binary compatible with the way GCC does things. There's a lot 
 of
 undocumented stuff there.


i would have to disagree.  all of the source code is there to look at, the 
ABI GCC uses is compatible with the Intel compiler ABI on linux, and linux 
is a mature and well-documented platform.

i really don't see why you're saying there's a lot that's undocumented.

-- 
Charles "grey wolf" Banas
Apr 24 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Charles Banas" <greywolf greyfade.net> wrote in message
news:oprn5ilrrt8ctebf news.digitalmars.com...

 On Thu, 24 Apr 2003 11:10:18 -0700, Walter <walter digitalmars.com> wrote:

 GCC is pretty slow. One big problem with putting DMC on linux is it




would

 need to be binary compatible with the way GCC does things. There's a lot
 of
 undocumented stuff there.


 i would have to disagree.  all of the source code is there to look at, the
 ABI GCC uses is compatible with the Intel compiler ABI on linux, and linux
 is a mature and well-documented platform.
 i really don't see why you're saying there's a lot that's undocumented.


If you could point me to a url which describes the format of the stabs
section in the ELF file for 386 linux, I'd appreciated it. Also, a url to
the format of the exception handling tables. I've spent hours with google
looking, and I find hints, but nothing definitive. I have several ELF
"specs", none of which describe even half the sections I find in the map
file generated by ld.

Examining the source code isn't really documentation, unless there's a
document embedded in the comments.

It's not that it's impossible to figure this stuff out, it's just time
consuming. I've been through the mill before on the PC with undocumented
formats.
Apr 24 2003
parent reply Charles Banas <greywolf greyfade.net> writes:
FWIW, i'm no expert on compilers.  i'm more a casual user of compilers for 
OS development. :)

of course, once i get an OS going, i'll need to port GCC....  that'll be 
fun.

see below:

On Thu, 24 Apr 2003 19:08:36 -0700, Walter <walter digitalmars.com> wrote:


 "Charles Banas" <greywolf greyfade.net> wrote in message
 news:oprn5ilrrt8ctebf news.digitalmars.com...

 i would have to disagree.  all of the source code is there to look at, 
 the
 ABI GCC uses is compatible with the Intel compiler ABI on linux, and 
 linux
 is a mature and well-documented platform.
 i really don't see why you're saying there's a lot that's undocumented.


 If you could point me to a url which describes the format of the stabs
 section in the ELF file for 386 linux, I'd appreciated it. Also, a url to
 the format of the exception handling tables. I've spent hours with google
 looking, and I find hints, but nothing definitive. I have several ELF
 "specs", none of which describe even half the sections I find in the map
 file generated by ld.


it looks like ./gcc/except.c shows how GCC generates the EH tables, and 
what their form is.  a few nested structures and a couple unions....  maybe 
you can just "steal" the code.  of course, that would require a GNU license 
on the linux port, but you know.


 Examining the source code isn't really documentation, unless there's a
 document embedded in the comments.


as i've said before, i'm not an expert when it comes to compilers, but at 
first glance, ./gcc/stab.def seems to contain a lot of well-commented 
information on the STABS section.  (it is also referenced by gstab.h.)  so, 
logically, anything that references gstab.h probably contains some useful 
code and comments.

again, i'm not an expert.


 It's not that it's impossible to figure this stuff out, it's just time
 consuming. I've been through the mill before on the PC with undocumented
 formats.


oh, i understand that.  of course, about 75% or more of windows in general 
seems to be extremely well undocumented.  sometimes, i'm rather amazed at 
how many things there are.  hence some of the many problems i've had 
writing OS kernel code in windows.  i'm seriously considering moving away 
from the win32-cygwin platform to a linux or BSD environment just for my OS 
development work.

of course, i'm pretty much a beginner with this, but i've learned a lot of 
things. :)  (some which, of course, may be wrong.)

after a couple quick google glances, i see what you mean by finding just 
clues.  but the clues were enough for me to find possibly useful 
information.  (i stress _possibly_.)

i hope i gave what you asked for.  smack me around with a large blunt 
object if i didn't.

-- 
Charles "grey wolf" Banas
Apr 24 2003
parent reply "Walter" <walter digitalmars.com> writes:
If you could email me those two files, I'd appreciate it. I don't have the
gcc source. Anyhow, despite having the C++ compiler running on linux, there
is many months of work with all those compatibility issues. Reworking the
way inheritance works is particularly horrible.
Apr 25 2003
parent reply Charles Banas <greywolf greyfade.net> writes:
On Fri, 25 Apr 2003 00:17:22 -0700, Walter <walter digitalmars.com> wrote:


 If you could email me those two files, I'd appreciate it. I don't have 
 the
 gcc source. Anyhow, despite having the C++ compiler running on linux, 
 there
 is many months of work with all those compatibility issues. Reworking the
 way inheritance works is particularly horrible.


the GCC source is available at http://gcc.gnu.org for download at any time. 
 it's actually not that big.

i'll send you the files ASAP, though.

-- 
Charles "grey wolf" Banas
Apr 25 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Charles Banas" <greywolf greyfade.net> wrote in message
news:oprn7a4py08ctebf news.digitalmars.com...

 the GCC source is available at http://gcc.gnu.org for download at any


time.

  it's actually not that big.


It's 26Mb compressed!


 i'll send you the files ASAP, though.


Thanks!
Apr 25 2003
next sibling parent Charles Banas <greywolf greyfade.net> writes:
On Fri, 25 Apr 2003 13:21:15 -0700, Walter <walter digitalmars.com> wrote:


 "Charles Banas" <greywolf greyfade.net> wrote in message
 news:oprn7a4py08ctebf news.digitalmars.com...

 the GCC source is available at http://gcc.gnu.org for download at any


 time.

 it's actually not that big.


 It's 26Mb compressed!


i guess cable and DSL have warped my mind, eh?

i actually have to think about what it decompresses to.  with only 270MB 
free space on a drive with 4kb blocks, GCC decompresses to over 350MB total 
(150MB actual, i think) and the linux kernel decompresses to a little over 
300MB.  but i somehow manage to keep it all straight. :)


 i'll send you the files ASAP, though.


 Thanks!


done.  enjoy.

-- 
Charles "grey wolf" Banas
Apr 25 2003
prev sibling parent Garen Parham <nospam garen.net> writes:
Walter wrote:


 It's 26Mb compressed!


Yeah it's huge.  However not all of the stuff in there is source code.  There
is a lot of documentation and testsuite stuff in there.  Plus libary support 
for C++, and Java/Objc/ADA compilers and test stuff for all those too.  There
are zillions of little test cases for C and C++ in there.  I'd think they'd 
be good to run through DMC++ for someone who felt like going bughunting.
Apr 26 2003
prev sibling parent reply Garen Parham <nospam garen.net> writes:
Walter wrote:

 GCC is pretty slow. One big problem with putting DMC on linux is it would
 need to be binary compatible with the way GCC does things. There's a lot of
 undocumented stuff there.



There is a multi-vendor ABI standard that GCC, ICC and HP follows thats
available.  A general overview is here:

http://www.codesourcery.com/publications_folder/a_common_vendor_abi_for_cplusplus

Complete docs here:

http://www.codesourcery.com/cxx-abi/
Apr 24 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Garen Parham" <nospam garen.net> wrote in message
news:b8a5vn$2ltp$1 digitaldaemon.com...

 There is a multi-vendor ABI standard that GCC, ICC and HP follows thats
 available.  A general overview is here:


http://www.codesourcery.com/publications_folder/a_common_vendor_abi_for_cplu
splus

Thanks, but it's in postscript. Any way to view ps in a browser?


 Complete docs here:

 http://www.codesourcery.com/cxx-abi/


That's a spec for the 64 bit Itanium. I need it for the 32 bit x86
processors.
Apr 24 2003
next sibling parent reply "Walter" <walter digitalmars.com> writes:
Never mind, I found a ps viewer.
Apr 24 2003
parent reply Andy Friesen <andy ikagames.com> writes:
Walter wrote:

 Never mind, I found a ps viewer.
 
 


If you didn't know, Ghostscript has a nifty little ps2pdf script. (I use 
the cygwin version in conjunction with Adobe's PS printer driver to 
create PDFs all the time)
Apr 24 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Andy Friesen" <andy ikagames.com> wrote in message
news:b8aie2$2u14$1 digitaldaemon.com...

 Walter wrote:

 Never mind, I found a ps viewer.


 If you didn't know, Ghostscript has a nifty little ps2pdf script. (I use
 the cygwin version in conjunction with Adobe's PS printer driver to
 create PDFs all the time)


I frankly don't understand why anyone uses ps to publish on the web. What's
the matter with html? Can't get the font just so? Who cares? PS is a
nuisance, and the viewer is what, 20 megabytes? Argh.

On top of that, in 20 years, who's going to be able to read those files?
Apr 25 2003
next sibling parent reply Charles Banas <greywolf greyfade.net> writes:
On Fri, 25 Apr 2003 00:10:58 -0700, Walter <walter digitalmars.com> wrote:


 "Andy Friesen" <andy ikagames.com> wrote in message
 news:b8aie2$2u14$1 digitaldaemon.com...

 Walter wrote:

 Never mind, I found a ps viewer.


 If you didn't know, Ghostscript has a nifty little ps2pdf script. (I use
 the cygwin version in conjunction with Adobe's PS printer driver to
 create PDFs all the time)


 I frankly don't understand why anyone uses ps to publish on the web. 
 What's
 the matter with html? Can't get the font just so? Who cares? PS is a
 nuisance, and the viewer is what, 20 megabytes? Argh.

 On top of that, in 20 years, who's going to be able to read those files?


printers.  plotters.  a few other things.

PostScript is the standard format that is used to send text and graphics to 
most printers.  i believe it's still used today.

and, since it's so darned common, Unix developers decided to use it as the 
main format for formatted text.  (although many use TeX and some derivative 
of SGML instead.)

but it's nice because it not only acts as an editable semi-permanent 
document, but it also encapsulates everything that a document may contain - 
charts, images, etc.

it's basically the open-source PDF.

i think in 20 years, there may still be a PostScript derivative in fairly 
wide usage.

-- 
Charles "grey wolf" Banas
Apr 25 2003
parent reply Helmut Leitner <leitner hls.via.at> writes:
Charles Banas wrote:

 but it's nice because it not only acts as an editable semi-permanent
 document, but it also encapsulates everything that a document may contain -
 charts, images, etc.
 
 it's basically the open-source PDF.


I think it is the other way round. PDF is a proprietary encapsulation of PS
with 
hyperlinks and data compression (and perhaps a few other things) added....

-- 
Helmut Leitner    leitner hls.via.at
Graz, Austria   www.hls-software.com
Apr 25 2003
parent reply "J. Daniel Smith" <J_Daniel_Smith HoTMaiL.com> writes:
Both the PostScript language and the PDF format are document by Adobe; I
don't think either one could be classified as "open source", but
"proprietrary" seems a bit of a strong word too given the extensive
documentation available.

PostScript is a pain as an archival format because it requires a fair amount
of effort to process; PDF was specifically designed for storing documents
electronically, it is much simplier (few programming language constructs)
and as a result is better suited this purpose.

   Dan

"Helmut Leitner" <leitner hls.via.at> wrote in message
news:3EA9046A.32A6FECB hls.via.at...

 Charles Banas wrote:

 but it's nice because it not only acts as an editable semi-permanent
 document, but it also encapsulates everything that a document may




contain -

 charts, images, etc.

 it's basically the open-source PDF.


 I think it is the other way round. PDF is a proprietary encapsulation of


PS with

 hyperlinks and data compression (and perhaps a few other things) added....

 -- 
 Helmut Leitner    leitner hls.via.at
 Graz, Austria   www.hls-software.com
Apr 25 2003
parent Charles Banas <greywolf greyfade.net> writes:
On Fri, 25 Apr 2003 15:10:02 -0400, J. Daniel Smith 
<J_Daniel_Smith HoTMaiL.com> wrote:


 Both the PostScript language and the PDF format are document by Adobe; I
 don't think either one could be classified as "open source", but
 "proprietrary" seems a bit of a strong word too given the extensive
 documentation available.

 PostScript is a pain as an archival format because it requires a fair 
 amount
 of effort to process; PDF was specifically designed for storing documents
 electronically, it is much simplier (few programming language constructs)
 and as a result is better suited this purpose.

 Dan

 "Helmut Leitner" <leitner hls.via.at> wrote in message
 news:3EA9046A.32A6FECB hls.via.at...

 Charles Banas wrote:

 but it's nice because it not only acts as an editable semi-permanent
 document, but it also encapsulates everything that a document may




 contain -

 charts, images, etc.

 it's basically the open-source PDF.


 I think it is the other way round. PDF is a proprietary encapsulation of


 PS with

 hyperlinks and data compression (and perhaps a few other things) 
 added....

 -- Helmut Leitner    leitner hls.via.at
 Graz, Austria   www.hls-software.com




i stand corrected


-- 
Charles "grey wolf" Banas
Apr 25 2003
prev sibling parent reply Olaf Rogalsky <olaf.rogalsky theorie1.physik.uni-erlangen.de> writes:
Walter wrote:

 I frankly don't understand why anyone uses ps to publish on the web. What's
 the matter with html? Can't get the font just so? Who cares? PS is a
 nuisance, and the viewer is what, 20 megabytes? Argh.


Postscript is like an API for printing under UNIX and Mac-OS. All 
professional printers and type setting machines understand postscript
natively. PDF essentialy is a cut down postscript language with only a
few new added features. Postscript nowadays is so widespread, that I fear
it will be still around when you and I are not.

The real nuisance are DOC files. But that would open up new bag of fleas.

Conclusion: Get a postscript viewer (adobe distiller/ghostscript).
-- 
+----------------------------------------------------------------------+
I Dr. Olaf Rogalsky                         Institut f. Theo. Physik I I
I Tel.: 09131 8528440                       Univ. Erlangen-Nuernberg   I
I Fax.: 09131 8528444                       Staudtstrasse 7 B3         I
I rogalsky theorie1.physik.uni-erlangen.de  D-91058 Erlangen           I
+----------------------------------------------------------------------+
Apr 25 2003
next sibling parent "Walter" <walter digitalmars.com> writes:
"Olaf Rogalsky" <olaf.rogalsky theorie1.physik.uni-erlangen.de> wrote in
message news:3EA92405.D919DF4 theorie1.physik.uni-erlangen.de...

 Postscript is like an API for printing under UNIX and Mac-OS. All
 professional printers and type setting machines understand postscript
 natively. PDF essentialy is a cut down postscript language with only a
 few new added features. Postscript nowadays is so widespread, that I fear
 it will be still around when you and I are not.

 The real nuisance are DOC files. But that would open up new bag of fleas.

 Conclusion: Get a postscript viewer (adobe distiller/ghostscript).


I have a viewer now. But if ps is so common, why can't browsers display it?
(My HP printer can't do ps either.)
Apr 25 2003
prev sibling parent reply "J. Daniel Smith" <J_Daniel_Smith HoTMaiL.com> writes:
PostScript is a full-blown programming language; it's primary function is as
a page description language.  It's no more of an API than C.

PDF and PostScript share a lot of common features such as the basic imaging
model; this makes it "easy" to move between the two.  But PDF does not have
programming language features such as loops; the result is a simplier
(although perhaps not as efficient) description of a printed page.

These days I would say that PDF is *MUCH* more widespread than raw PS files;
the Adobe Acrobat Reader is the way to go (and it even works as a "plug in"
in IE).

   Dan


"Olaf Rogalsky" <olaf.rogalsky theorie1.physik.uni-erlangen.de> wrote in
message news:3EA92405.D919DF4 theorie1.physik.uni-erlangen.de...

 Walter wrote:

 I frankly don't understand why anyone uses ps to publish on the web.




What's

 the matter with html? Can't get the font just so? Who cares? PS is a
 nuisance, and the viewer is what, 20 megabytes? Argh.


 Postscript is like an API for printing under UNIX and Mac-OS. All
 professional printers and type setting machines understand postscript
 natively. PDF essentialy is a cut down postscript language with only a
 few new added features. Postscript nowadays is so widespread, that I fear
 it will be still around when you and I are not.

 The real nuisance are DOC files. But that would open up new bag of fleas.

 Conclusion: Get a postscript viewer (adobe distiller/ghostscript).
 -- 
 +----------------------------------------------------------------------+
 I Dr. Olaf Rogalsky                         Institut f. Theo. Physik I I
 I Tel.: 09131 8528440                       Univ. Erlangen-Nuernberg   I
 I Fax.: 09131 8528444                       Staudtstrasse 7 B3         I
 I rogalsky theorie1.physik.uni-erlangen.de  D-91058 Erlangen           I
 +----------------------------------------------------------------------+
Apr 25 2003
parent reply "Walter" <walter digitalmars.com> writes:
"J. Daniel Smith" <J_Daniel_Smith HoTMaiL.com> wrote in message
news:b8c0du$r6f$1 digitaldaemon.com...

 These days I would say that PDF is *MUCH* more widespread than raw PS


files;

 the Adobe Acrobat Reader is the way to go (and it even works as a "plug


in"

 in IE).


That's true, but pdf has its own bugs. You can't cut bits of text to the
clipboard. You can't convert it to text without emailing it to an Adobe
service, which produces raw text that is nearly unusable. (That said,
Google's translater is much better.) The IE plugin renders PDF files in a
blurry font. The page format just doesn't fit well on a typical computer
display.

HTML doesn't suffer from any of these problems. HTML automatically adjusts
to your display size (bring up some of the DMC online documentation, and
resize the window!), it's rendered crisply, it's easy to use without any
proprietary tools (I do all my HTML with a simple text editor), it can be
read by practically any machine, and if that fails, you can still read it
with an ordinary text editor.

The DMC++ documentation was originally all in pdf. I converted it all to
HTML, and am well convinced of the superiority of the latter. Just compare
the two side by side!
Apr 25 2003
next sibling parent "J. Daniel Smith" <J_Daniel_Smith HoTMaiL.com> writes:
It sounds like some of your PDF problems are caused from scanned images;
I've had no problem copying text from a PDF file or seeing clear fonts on my
screen.  Take a look at some of the files at
http://partners.adobe.com/asn/developer/technotes/acrobatpdf.html.
Remember, PDF is intended to be an electronic piece of paper; you're right
in that PDF files sometimes don't work well on a computer screen, but that's
mostly the fault of the document's creator, not the PDF format itself.

I was trying to clarify the differences between PDF and PostScript, not
HTML.  I agree that for many things HTML is the way to go for the reasons
you site.

   Dan

"Walter" <walter digitalmars.com> wrote in message
news:b8c657$uud$1 digitaldaemon.com...

 "J. Daniel Smith" <J_Daniel_Smith HoTMaiL.com> wrote in message
 news:b8c0du$r6f$1 digitaldaemon.com...

 These days I would say that PDF is *MUCH* more widespread than raw PS


 files;

 the Adobe Acrobat Reader is the way to go (and it even works as a "plug


 in"

 in IE).


 That's true, but pdf has its own bugs. You can't cut bits of text to the
 clipboard. You can't convert it to text without emailing it to an Adobe
 service, which produces raw text that is nearly unusable. (That said,
 Google's translater is much better.) The IE plugin renders PDF files in a
 blurry font. The page format just doesn't fit well on a typical computer
 display.

 HTML doesn't suffer from any of these problems. HTML automatically adjusts
 to your display size (bring up some of the DMC online documentation, and
 resize the window!), it's rendered crisply, it's easy to use without any
 proprietary tools (I do all my HTML with a simple text editor), it can be
 read by practically any machine, and if that fails, you can still read it
 with an ordinary text editor.

 The DMC++ documentation was originally all in pdf. I converted it all to
 HTML, and am well convinced of the superiority of the latter. Just compare
 the two side by side!
Apr 25 2003
prev sibling next sibling parent reply Ilya Minkov <midiclub 8ung.at> writes:
Walter wrote:


 That's true, but pdf has its own bugs. You can't cut bits of text to
 the clipboard. You can't convert it to text without emailing it to an
 Adobe service, which produces raw text that is nearly unusable.


Oh really? I've had similar problems with so far very few files. There
is a text selection tool in acrobat reader, which allows you to select
text and then copy it into clipboard. And if this method works, there
are also free tools which would extract the whole text for you.

Now, if it doesn't work and letters appear blocky if you zoom closely
in, that probably means it was created with a mistake. Some distros of
TeX prefer rasterising single letters into bitmaps when converting into
PostScript, to ensure high printing speed and quality. Then this
document might get converted into PDF and stays illegible on screen, but
if you print it it might look very good. It also becomes a collection of
bitmaps and looses text meaning, so you cannot search in such documents
or extract text. My MikTeX DVI2PS files without such a problem. There
also exists DVI2PDF which always gets high-quality PDF output, but
sometimes fails to work at all.



 (That said, Google's translater is much better.) The IE plugin
 renders PDF files in a blurry font. The page format just doesn't fit
 well on a typical computer display.


Right.


 HTML doesn't suffer from any of these problems. HTML automatically
 adjusts to your display size (bring up some of the DMC online
 documentation, and resize the window!), it's rendered crisply, it's
 easy to use without any proprietary tools (I do all my HTML with a
 simple text editor), it can be read by practically any machine, and
 if that fails, you can still read it with an ordinary text editor.


Right. But some people prefer writing TeX - it is very easy to read, and
it is very well structured and need not impose any formatting. Thus it
converts perfectly well to HTML, and of course to printed documentation.


 The DMC++ documentation was originally all in pdf. I converted it all
 to HTML, and am well convinced of the superiority of the latter. Just
 compare the two side by side!


:)

-i.
Apr 25 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Ilya Minkov" <midiclub 8ung.at> wrote in message
news:b8cab3$11ki$1 digitaldaemon.com...

 That's true, but pdf has its own bugs. You can't cut bits of text to
 the clipboard. You can't convert it to text without emailing it to an
 Adobe service, which produces raw text that is nearly unusable.


 Oh really? I've had similar problems with so far very few files.


I read a lot of pdf files I find on the web, and I find it with all of them.
You can't even save the file - you need to back up and right-click on the
hyperlink to it.


 There
 is a text selection tool in acrobat reader, which allows you to select
 text and then copy it into clipboard.


Hmm. I was using the click & drag method, which is how you copy text with
everything else. Is there something I missed?
Apr 25 2003
parent reply "Daniel Yokomiso" <daniel_yokomiso yahoo.com.br> writes:
"Walter" <walter digitalmars.com> escreveu na mensagem
news:b8cjh4$19s9$1 digitaldaemon.com...

 "Ilya Minkov" <midiclub 8ung.at> wrote in message
 news:b8cab3$11ki$1 digitaldaemon.com...

 That's true, but pdf has its own bugs. You can't cut bits of text to
 the clipboard. You can't convert it to text without emailing it to an
 Adobe service, which produces raw text that is nearly unusable.


 Oh really? I've had similar problems with so far very few files.


 I read a lot of pdf files I find on the web, and I find it with all of


them.

 You can't even save the file - you need to back up and right-click on the
 hyperlink to it.



There's a toolbar available in the Adobe Acrobat plugin, so you can click in
the save button. AFAIK its default is 'on', but you can hide it at will.



 There
 is a text selection tool in acrobat reader, which allows you to select
 text and then copy it into clipboard.


 Hmm. I was using the click & drag method, which is how you copy text with
 everything else. Is there something I missed?


Use the tooltips Luke ;-)
There's an icon with a T and a dotted square. It's the text selection tool.
Enable it and the mouse cursor will enter in text selection mode, allowing
selection, copy & paste, etc.


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.476 / Virus Database: 273 - Release Date: 24/4/2003
Apr 25 2003
parent "Walter" <walter digitalmars.com> writes:
"Daniel Yokomiso" <daniel_yokomiso yahoo.com.br> wrote in message
news:b8clh6$1b37$1 digitaldaemon.com...

 Use the tooltips Luke ;-)
 There's an icon with a T and a dotted square. It's the text selection


tool.

 Enable it and the mouse cursor will enter in text selection mode, allowing
 selection, copy & paste, etc.


You're right, that's the trick. Thanks!
Apr 25 2003
prev sibling next sibling parent reply Helmut Leitner <helmut.leitner chello.at> writes:
Walter wrote:

 HTML doesn't suffer from any of these problems. HTML automatically adjusts
 to your display size (bring up some of the DMC online documentation, and
 resize the window!), it's rendered crisply, it's easy to use without any
 proprietary tools (I do all my HTML with a simple text editor), it can be
 read by practically any machine, and if that fails, you can still read it
 with an ordinary text editor.


Disadvantages of HTML for printed documentation:
   - you can't create a neat printout
   - the toc is pretty useless without page numbering
   - you can't create an index
   - ...

A printed documentation is IMHO somewhat better than an online documentation
in certain situations:
   - if you want to annotate what you read
   - if you want to get a real overview (read from A-Z)

--
Helmut Leitner    leitner hls.via.at   
Graz, Austria   www.hls-software.com
Apr 25 2003
parent Alix Pexton <Alix thedjournal.com> writes:
Helmut Leitner wrote:

 Disadvantages of HTML for printed documentation:
    - you can't create a neat printout


If you use CSS to create a style sheet for your document, you can use 
the  page media selector to add special formatting for printing, 
including page numbering and per page headers and footers.


    - the toc is pretty useless without page numbering


Creating the pagenumbers is not that hard, but linking them to your toc 
is. I think you could probably do it using xsl:fo[1] and xsl:t[2] .


    - you can't create an index


with xsl:t that is a sinch.


    - ...
 
 A printed documentation is IMHO somewhat better than an online documentation
 in certain situations:
    - if you want to annotate what you read


The w3c has a working group that is developing a system call annotea, 
which is like a digital sticky note crossed with wiki-like community 
features, so you can annotate a webpage, and set the annotation to be 
public if you like, you also get to put hyperlinks and cross-references.


    - if you want to get a real overview (read from A-Z)


Agreed, I can't read from a monitor for more than a few minutes at a 
time, and you can't keep a finger in a page that you want to flick back 
to latter (yet)...

I'm a web-guy, it's my job to advocate [X|HT]ML...

Alix...

-- 
             Alix Pexton
Webmaster - http://www.theDjournal.com

             Alix theDjournal.com
Apr 26 2003
prev sibling next sibling parent Richard Krehbiel <rich kastle.com> writes:
Walter wrote:

 "J. Daniel Smith" <J_Daniel_Smith HoTMaiL.com> wrote in message
 news:b8c0du$r6f$1 digitaldaemon.com...
 

These days I would say that PDF is *MUCH* more widespread than raw PS


 
 files;
 

the Adobe Acrobat Reader is the way to go (and it even works as a "plug


 
 in"
 

in IE).


 
 
 That's true, but pdf has its own bugs. You can't cut bits of text to the
 clipboard. You can't convert it to text without emailing it to an Adobe
 service, which produces raw text that is nearly unusable. (That said,
 Google's translater is much better.) The IE plugin renders PDF files in a
 blurry font. The page format just doesn't fit well on a typical computer
 display.
 
 HTML doesn't suffer from any of these problems. HTML automatically adjusts
 to your display size (bring up some of the DMC online documentation, and
 resize the window!), it's rendered crisply, it's easy to use without any
 proprietary tools (I do all my HTML with a simple text editor), it can be
 read by practically any machine, and if that fails, you can still read it
 with an ordinary text editor.
 
 The DMC++ documentation was originally all in pdf. I converted it all to
 HTML, and am well convinced of the superiority of the latter. Just compare
 the two side by side!


So, what does the DMC++ documentation look like on the printed page? 
Yeah, you don't really know, and you don't particularly care - and I'm 
not saying you should.  But if you DID care, then you'd be better served 
by PDF.

PostScript was designed during the bad old days of DOS and the million 
WordPerfect printer drivers.  It's a language for describing a printed 
page as accurately as possible.  (The original intent was to create a 
higher-level abstraction and API for printed output, so that printer 
interfaces and languages did not continue to proliferate.  Though it was 
a staple of the Apple/Adobe desktop publishing platform, the Windows 
GDI, with vendor-written printer drivers, kinda killed the idea on the PC.)

PS is not intended to be human-authored or editable; it's to be 
program-generated and shipped to the printed page.  And, as PDF is 
mostly just encapsulated PS, it has the same characteristic.  You're 
supposed to author and edit your documents in some other tool, then 
render to PDF for distribution to millions who *intend* *to* *print* *it*.

HTML is mostly derived from (I think; at the very least, inspired by) 
SGML, which *was* designed for human consumption as plain text.  So it's 
easy to author and edit, it's easy to read and revise, but you can't say 
what it will look like when printed, except in the most general terms.

(BTW I am able to copy selected text from a PDF to the clipboard. 
What's the problem there?)
Apr 28 2003
prev sibling parent reply Karl Bochert <kbochert copper.net> writes:
 The DMC++ documentation was originally all in pdf. I converted it all to
 HTML, and am well convinced of the superiority of the latter. Just compare
 the two side by side!
 


Ever considered using Help&Maual?  Table of contents, cross references etc, etc.
Outputs document in HTML, PDF, WORD(RTF),  .HLP, .CHM and e-books.
Provides a gui interface for designing document.
May 03 2003
parent reply "Walter" <walter digitalmars.com> writes:
"Karl Bochert" <kbochert copper.net> wrote in message
news:1103_1051984547 bose...

 The DMC++ documentation was originally all in pdf. I converted it all to
 HTML, and am well convinced of the superiority of the latter. Just




compare

 the two side by side!


 Ever considered using Help&Maual?  Table of contents, cross references


etc, etc.

 Outputs document in HTML, PDF, WORD(RTF),  .HLP, .CHM and e-books.
 Provides a gui interface for designing document.


I'm not familiar with it. But I will suggest that once one is familiar with
how html works, having a gui interface is not likely to help much. Right
click and 'view source' on some of the html files on the web site.
May 03 2003
parent reply Karl Bochert <kbochert copper.net> writes:
On Sat, 3 May 2003 14:20:59 -0700, "Walter" <walter digitalmars.com> wrote:

 
 "Karl Bochert" <kbochert copper.net> wrote in message
 news:1103_1051984547 bose...

 The DMC++ documentation was originally all in pdf. I converted it all to
 HTML, and am well convinced of the superiority of the latter. Just




 compare

 the two side by side!


 Ever considered using Help&Maual?  Table of contents, cross references


 etc, etc.

 Outputs document in HTML, PDF, WORD(RTF),  .HLP, .CHM and e-books.
 Provides a gui interface for designing document.


 
 I'm not familiar with it. But I will suggest that once one is familiar with
 how html works, having a gui interface is not likely to help much. Right
 click and 'view source' on some of the html files on the web site.
 


 
Your manual is an excellent example of clear and straight-forward html.
(ISort of illustrates why you are good at writing  compilers!)

The gui interface is more than just a WYSIWYG editor. It has a separate
window for editing the TOC, and menus with many tools to assist with things
like keywords, links, outlines, booklet printing, etc.
Just the thing for a documentation putz like me.

see <www.ec-software.com>
May 04 2003
parent "Walter" <walter digitalmars.com> writes:
"Karl Bochert" <kbochert copper.net> wrote in message
news:1103_1052037017 bose...

 Your manual is an excellent example of clear and straight-forward html.
 (ISort of illustrates why you are good at writing  compilers!)


Thanks!


 The gui interface is more than just a WYSIWYG editor. It has a separate
 window for editing the TOC, and menus with many tools to assist with


things

 like keywords, links, outlines, booklet printing, etc.
 Just the thing for a documentation putz like me.


If you find it useful, I won't argue with that.


 see <www.ec-software.com>
May 06 2003
prev sibling next sibling parent reply Garen Parham <nospam garen.net> writes:
Walter wrote:



 
 That's a spec for the 64 bit Itanium. I need it for the 32 bit x86
 processors.
 


IIRC it was originally targeted for Itanium/funded by Intel&HP, but it's now
used for all arch's.  Checkout the ACCU overview link.
Apr 24 2003
next sibling parent "Walter" <walter digitalmars.com> writes:
"Garen Parham" <nospam garen.net> wrote in message
news:b8ahum$2tqv$1 digitaldaemon.com...

 Walter wrote:



 That's a spec for the 64 bit Itanium. I need it for the 32 bit x86
 processors.


 IIRC it was originally targeted for Itanium/funded by Intel&HP, but it's


now

 used for all arch's.  Checkout the ACCU overview link.


You're right.
Apr 25 2003
prev sibling parent Mark T <Mark_member pathlink.com> writes:
 That's a spec for the 64 bit Itanium. I need it for the 32 bit x86
 processors.
 


IIRC it was originally targeted for Itanium/funded by Intel&HP, but it's now
used for all arch's.  Checkout the ACCU overview link.


what link could you post it here

thanks
May 03 2003
prev sibling parent "Sean L. Palmer" <palmer.sean verizon.net> writes:
If you can get Google to find it, it can display postscript as html.  This
one, unfortunately, does not appear to be cached as html yet.

btw a Google search on "Digital Mars" gets lots of appropriate hits.  ;)

Sean

"Walter" <walter digitalmars.com> wrote in message
news:b8abmv$2pq5$1 digitaldaemon.com...

 "Garen Parham" <nospam garen.net> wrote in message
 news:b8a5vn$2ltp$1 digitaldaemon.com...

 There is a multi-vendor ABI standard that GCC, ICC and HP follows thats
 available.  A general overview is here:




http://www.codesourcery.com/publications_folder/a_common_vendor_abi_for_cplu

 splus

 Thanks, but it's in postscript. Any way to view ps in a browser?


 Complete docs here:

 http://www.codesourcery.com/cxx-abi/


 That's a spec for the 64 bit Itanium. I need it for the 32 bit x86
 processors.
Apr 25 2003
prev sibling parent "Matthew Wilson" <matthew stlsoft.org> writes:
Oh no! It's the what-if-you-get-run-over-by-a-bus conundrum.

Perhaps we need to start cloning little compiler walters?

:)

"Walter" <walter digitalmars.com> wrote in message
news:b86psp$eum$3 digitaldaemon.com...

 You're probably all wondering why I haven't done an update in a while.


There

 are two reasons:

 1) I was sick for a couple weeks. Ugh.

 2) I'm working on a linux version. dmd is working under linux now, but the
 runtime library needs much work. Work is frustratingly slow on it, mainly
 because of my inexperience with linux and the dearth of good information


on

 how the internals of file formats, etc., actually work.
Apr 23 2003