I know about TeX. I've used batch formatters since GML didn't have an S, TeX, troff, etc...
They are powerful, wonderful packages but
do not allow certain types of presenations
or require extra programming, macro or running things through filters, to get the output you want. Image handling in batch formatters is
typically painful. Frame isn't perfect but it
wraps a lot of things you want (other than
bibligraphic db) into a GUI package that
is very easy to drive once you understand
some fundamentals. Achieiving the same with
TeX, if possible, is far more difficult
other than for simple documents or those
that simply use a pre-defined set of macros.
I'm a long time Frame user (12 years) and also desire either a Linux (really FreeBSD) FrameMaker (for real this time guys) or an alternative that is at least as functional.
But there aren't any.
And before people say, no LyX is not an alternative, KWord isn't there by a long shot (although it's heart is in the right place) and StarOffice is too busy trying to be Word for my liking (I haven't tried 6, is it any different other than being in its own window?). $EDITOR isn't in the running
for the type of documents you use Frame for.
The important thing is the structured approach to documents with, as you say, consistent formatting and a reasonable, though not great, GUI (Unix frame is better than Windows in this regard, the Windows one is pretty sucky). The Frame feature set is
pretty good, or at least provides the types of
things technical documents require. And it
works, pretty much, reliably and consistently if a little quirky at times.
The approach taken by a lot of the people desiring the Linux version is to get a cheap Sun box and a multilicense install of Solaris Frame running with remote X sessions. The
machine only runs Frame so doesn't need to be
super fast, give it a bunch of memory and
it should be happy.
There's opportunity here. To the comments that said this is another "re-ivent the wheel" it's only because Adobe don't seem to care too much about FrameMaker. The feature
set and approach to document formatting
suits many types of technical documents
far better than typical "word processors".
People who do these types of documents
and have experienced Frame mostly agree
there is no current alternative.
But it doesn't get advertised along with their other stuff
in the things I see, they try hard to pretend
there isn't a version for Unix and some
annoying problems have persisted for years.
People (like me!) are crying out for a GUI-based, document preparation system for technical documents and others that require reliable formatting, integrate graphics, footnotes, tables, want an index, TOC, LOF, etc... We don't need talking parrots. Just copy Frame and add some biblography support to it:) May be fix some of the other stuff too (reference pages, uggh, process the Frame data structure and you'll realise). Oh, and the types of people who really want a tool
like this pay $$$ for it. We depend on it,
which is why we don't use Word (Word Perfect
sucks too:))
Excellent point. And to allow this to be the case AND get a healthy, free market we need standard, open interfaces, i.e, what Unix is these days. We shouldn't have it in the hands of a corp, especially one that behaves in the manner of Microsoft.
The market has essentially polarized into Windows vs. (some form of) Unix. Which
basically boils down to "Microsoft" vs.
"everyone else". Some play both sides
but if they're smart they know what'll
happen to them if they hang around MS too long. Some exit strategy is required.
So we get Unix, with all its warts. But we
also get a "People's implementation" via Linux which is great ("pimp" linux:) and BSD - pick your political party. There's the "Mack Truck" versions via Solaris/AIX/et al on honking big mothers of machines. And
things like QNX, various real-time Unicies, etc... A free market. (I'm purposefully avoiding the Linux everywhere thing, sure do it but other systems have often have their own advantages and it may not be too healthy in the long term, we have to wait and see)
While I agree with "Where did you pull this assertion out?" the follow up doesn't go far enough. People are forgetting that languages have models of data and regardless of translator and execution style you need to implement the language's data model. For Java this means objects with GC. VM or real-M.
I think it'll go the way of the car industry with hot rods looked down on. Machines (h/w + OS) will need to be certified before they can be "driven" on the public roads ('net). People who drive (admins) need licenses (MSCE, oh god no!) before hooking the machine to the 'net. Cops look out for drivers (probe open relays etc...) and eat donuts while reading/.
USP#4555775 basically patents the idea of multiple graphical virtual screens (called ?windows?) to represent the multiple physical screens.
That's Rob Pike's layers patent. It isn't windows per se (even though claim 1 reads that way) but the layers representation of backing store (or backing store in general if your AT&T lawyers).
For the Lindows folks.... He even states, in a patent filed in 1982,
It is well known to break the bitmap, and hence the display, into a plurality of regions for separate displays. Each separate display is called a "window" and the prior art has the ability to display multiple windows simultaneously, with several if not all windows overlapping, leaving one window fully visible and the others partially or wholly obscured. Windows are overlapping rectangles each of which can be considered an operating environment, much like sheets of paper on a desk.
So at that stage someone skilled in the art considered "windows" to be a well-known term.
However he did use it in a more restricted way
(read the patent).
Correct the correction. NeXTSTEP ran on those architectures (I used to do fat binaries for all of them). Then the OpenStep stuff happened to support other base OS's - Solaris, Windows, others. NeXTSTEP changed along with it. That's about the time I left the NeXT world but watched from the sidelines (the people I worked for owned a chunk of NeXT).
But you're right about the poor integration. Here's a comment from some code I wrote
when NeXTSTEP for Intel was first shipped,
#ifdef NSFIP /*
* memcmp (as per ANSI definition) was shipped broken on
* NeXTStep for Intel Processors. This should be removed
* sometime after it is fixed.
*/
Oh boy. When memcmp doesn't work what else is broken:)
The C library version of memcmp assumed big-endian [68K] and did word accesses to compare ints rather than bytes, IIRC it messed up the end case, comparing the wrong bytes of the last word. There was no problem when optimizing, gcc uses a (good) builtin, but turn off -O and things broke.
But it all lives on...
I bought a Mac a while ago to run Mac OS X. Immediately after getting it I install OS X and reboot. And a big smile comes to my face. It was like stepping back in time. The blue background, the spinning color wheel, the boot graphics, the login panel. The look's updated a bit but it's NeXTSTEP. Services, NetInfo,.app's, Interface Builder. Finally, its all out there again.
Re:Looks dorky, but makes a great hat!
on
New iMac Announced
·
· Score: 2
Ha! We were talking about that at work today. Wear it as a hat. Flip the screen over. Hop on your SHT and whoosh... Get me off this crazy thing!
Rather than thinking of it as a collection of bytes it is worthwhile to consider the bytes as a single, multi-precision number. Just as that prime that unzips to DeCSS is a number, the number posted can be interpreted as an image. Isn't it great that copyright law in the digital age lets people own numbers. Most are pretty big (and there's quite a few to go round:) - an average MP3 being around 5 million base 256 digits long - but there are some that aren't so human-unfriendly. What I want to know is what is the shortest number I can own?
But if you think these Tumbleweed patents aren't great leaps forward take a look at US6,061,448, "Method and system for dynamic server document encryption" which sounds familiar.
Read the patent please. They're not patenting UNIX-style symlinks. Here's their
description from the patent text
The present invention provides facilities in the file system of an operating system or other program for creating and maintaining dynamic links among file-system objects. These links differ from conventional static links in that they come into being only when needed to access an object, and they cease to exist whenever a user's access ceases. At the same time, however, dynamic links are more than mere ad hoc relationships that need to be defined anew for every file access. Although the links themselves go away between successive accesses, a persistent rule or definition recreates the links automatically for every access. Therefore, changes to files or directories between accesses are automatically reflected whenever a new access occurs, without any user action in changing or recreating the links.
Now this in itself may or may not be entirely new
but is definitely something a little more than
your ln -s. BSD portals and the Plan 9 file server approach would allow similar systems to be implemented and such systems were likely discussed, people looking for prior art should be able to find something.
If people are interested I reccomend reading some of the other Microsoft patents. You'd be surprised what they claim in some of them. Others are so specific ("bit-3 of MS-DOS FAT field XYZ in the root directory of the C drive" sort of thing) they appear solely to exist to get their patent
numbers up. (Did someone implement patent quotas at MS in the 90s?) In fact a full review of MS patents could keep/. going for quite
some time. It'd really get the blood pressure up too:)
Re:Not willing to go to jail to prove a point?
on
DMCA 2, Freedom 0
·
· Score: 2
When the 8086 was released Intel supplied an 8080 assembler translator. The instruction set was obviously not that different. Many of the first IBM PC programs - BASIC, Wordstar, etc... were simply the CP/M versions run through the translator and then patched up to make them actually work. MS-DOS 1 was like my hacked up CP/M system with the renamed commands and a bit of Z system thrown in (but not enough). The work in MS-DOS was in the CCP - COPY and DEL, woohoo!
I don't know where you get this "CP/M compatibility" thing, it was pretty much a direct copy and some even say it was a direct copy thanks to the Intel assembler translator and CP/M source access. I recall the old QDOS ads in the top right corner of some mag I was reading at the time, may have been Kilobaud or maybe something a little more techie, can't recall).
Oh, Kildall said something about the use of '$' as the sentinel in the output call (9) as being special and that only he could explain it. Anyone know something about this?
Quite a few years ago I recall seeing a documentary on security guys. One used to get into the buildings, usually dressed as phone tech or similar, and just leave a biz card on their patch panel "If you find this you need me." or somesuch. Very cool way
to advertise (although some ahole would slap a charge eventually).
Re:Release Often?
on
Linux 2.4.13
·
· Score: 1, Offtopic
Ha! No teams from the northern hemisphere can play rugby.
The IP virus, looks for documents containing trade secrets, and quietly posts them to random usenet groups.
We've already had that one. It's called SirCam. It's been mailing out corporate IP for a whlie now. Unfortunately I mustn't appear in enough Outlook address books, the most interesting thing I got was some poetry that had been submitted to a publisher (in a western state of the USA). After conducting some free "market research" for them my advice is, "Don't publish that crap poetry!"
Slashdot Mirror
True, but it's better than nothing. Wilde summed it up quite well,
I know about TeX. I've used batch formatters since GML didn't have an S, TeX, troff, etc... They are powerful, wonderful packages but do not allow certain types of presenations or require extra programming, macro or running things through filters, to get the output you want. Image handling in batch formatters is typically painful. Frame isn't perfect but it wraps a lot of things you want (other than bibligraphic db) into a GUI package that is very easy to drive once you understand some fundamentals. Achieiving the same with TeX, if possible, is far more difficult other than for simple documents or those that simply use a pre-defined set of macros.
And before people say, no LyX is not an alternative, KWord isn't there by a long shot (although it's heart is in the right place) and StarOffice is too busy trying to be Word for my liking (I haven't tried 6, is it any different other than being in its own window?). $EDITOR isn't in the running for the type of documents you use Frame for.
The important thing is the structured approach to documents with, as you say, consistent formatting and a reasonable, though not great, GUI (Unix frame is better than Windows in this regard, the Windows one is pretty sucky). The Frame feature set is pretty good, or at least provides the types of things technical documents require. And it works, pretty much, reliably and consistently if a little quirky at times.
The approach taken by a lot of the people desiring the Linux version is to get a cheap Sun box and a multilicense install of Solaris Frame running with remote X sessions. The machine only runs Frame so doesn't need to be super fast, give it a bunch of memory and it should be happy.
There's opportunity here. To the comments that said this is another "re-ivent the wheel" it's only because Adobe don't seem to care too much about FrameMaker. The feature set and approach to document formatting suits many types of technical documents far better than typical "word processors". People who do these types of documents and have experienced Frame mostly agree there is no current alternative. But it doesn't get advertised along with their other stuff in the things I see, they try hard to pretend there isn't a version for Unix and some annoying problems have persisted for years. People (like me!) are crying out for a GUI-based, document preparation system for technical documents and others that require reliable formatting, integrate graphics, footnotes, tables, want an index, TOC, LOF, etc... We don't need talking parrots. Just copy Frame and add some biblography support to it :) May be fix some of the other stuff too (reference pages, uggh, process the Frame data structure and you'll realise). Oh, and the types of people who really want a tool
like this pay $$$ for it. We depend on it,
which is why we don't use Word (Word Perfect
sucks too :))
"They Live". What a film.
"We've got one who can see!"
(Kind of on topic actually, people are starting to get through the hype associated with some things and adopt stuff that works, like PostgreSQL).
Well, for starters. 95% of it is brown and sticky.
The market has essentially polarized into Windows vs. (some form of) Unix. Which basically boils down to "Microsoft" vs. "everyone else". Some play both sides but if they're smart they know what'll happen to them if they hang around MS too long. Some exit strategy is required.
So we get Unix, with all its warts. But we also get a "People's implementation" via Linux which is great ("pimp" linux :) and BSD - pick your political party. There's the "Mack Truck" versions via Solaris/AIX/et al on honking big mothers of machines. And
things like QNX, various real-time Unicies, etc... A free market. (I'm purposefully avoiding the Linux everywhere thing, sure do it but other systems have often have their own advantages and it may not be too healthy in the long term, we have to wait and see)
Where do MS fit in then?
While I agree with "Where did you pull this assertion out?" the follow up doesn't go far enough. People are forgetting that languages have models of data and regardless of translator and execution style you need to implement the language's data model. For Java this means objects with GC. VM or real-M.
It obliterates the file (just read the source).
Doesn't check st_nlink even though its got a statbuf at hand. Hmmm, bug, feature or another switch?
I think it'll go the way of the car industry with hot rods looked down on. Machines (h/w + OS) will need to be certified before they can be "driven" on the public roads ('net). People who drive (admins) need licenses (MSCE, oh god no!) before hooking the machine to the 'net. Cops look out for drivers (probe open relays etc...) and eat donuts while reading /.
For the Lindows folks.... He even states, in a patent filed in 1982,
So at that stage someone skilled in the art considered "windows" to be a well-known term. However he did use it in a more restricted way (read the patent).But you're right about the poor integration. Here's a comment from some code I wrote when NeXTSTEP for Intel was first shipped,
#ifdef NSFIP
/*
* memcmp (as per ANSI definition) was shipped broken on
* NeXTStep for Intel Processors. This should be removed
* sometime after it is fixed.
*/
Oh boy. When memcmp doesn't work what else is broken :)
The C library version of memcmp assumed big-endian [68K] and did word accesses to compare ints rather than bytes, IIRC it messed up the end case, comparing the wrong bytes of the last word. There was no problem when optimizing, gcc uses a (good) builtin, but turn off -O and things broke.
But it all lives on... I bought a Mac a while ago to run Mac OS X. Immediately after getting it I install OS X and reboot. And a big smile comes to my face. It was like stepping back in time. The blue background, the spinning color wheel, the boot graphics, the login panel. The look's updated a bit but it's NeXTSTEP. Services, NetInfo, .app's, Interface Builder. Finally, its all out there again.
Ha! We were talking about that at work today. Wear it as a hat. Flip the screen over. Hop on your SHT and whoosh... Get me off this crazy thing!
Rather than thinking of it as a collection of bytes it is worthwhile to consider the bytes as a single, multi-precision number. Just as that prime that unzips to DeCSS is a number, the number posted can be interpreted as an image. Isn't it great that copyright law in the digital age lets people own numbers. Most are pretty big (and there's quite a few to go round :) - an average MP3 being around 5 million base 256 digits long - but there are some that aren't so human-unfriendly. What I want to know is what is the shortest number I can own?
But if you think these Tumbleweed patents aren't great leaps forward take a look at US6,061,448, "Method and system for dynamic server document encryption" which sounds familiar.
Oh, and the claims seem quite reasonable in attempting to claim what they describe. There's
no broad claim of symbolic links per se (that I can see).
If people are interested I reccomend reading some of the other Microsoft patents. You'd be surprised what they claim in some of them. Others are so specific ("bit-3 of MS-DOS FAT field XYZ in the root directory of the C drive" sort of thing) they appear solely to exist to get their patent numbers up. (Did someone implement patent quotas at MS in the 90s?) In fact a full review of MS patents could keep /. going for quite
some time. It'd really get the blood pressure up too :)
But they don't own USENIX.
I programmed an Imlac and used to get audible sound out of the monitor when my lines got redrawn too fast. I always thought it was going to blow up.
I don't know where you get this "CP/M compatibility" thing, it was pretty much a direct copy and some even say it was a direct copy thanks to the Intel assembler translator and CP/M source access. I recall the old QDOS ads in the top right corner of some mag I was reading at the time, may have been Kilobaud or maybe something a little more techie, can't recall).
Oh, Kildall said something about the use of '$' as the sentinel in the output call (9) as being special and that only he could explain it. Anyone know something about this?
NASA already have a cluster of these.
Quite a few years ago I recall seeing a documentary on security guys. One used to get into the buildings, usually dressed as phone tech or similar, and just leave a biz card on their patch panel "If you find this you need me." or somesuch. Very cool way to advertise (although some ahole would slap a charge eventually).
Ha! No teams from the northern hemisphere can play rugby.
Damm, there's that "post from the future" bug again.