It's not even slightly scriptable. It's a file format.
That's funny. TeX is a Turing complete language. You can have interactive programs (like makebst). Every single one of my files does something differently (by design) if the target format is PS or PDF. I have definitions that can be replaced right-up-until compile-time.
LaTeX doesn't have a macro format for dynamic changes as you edit it or embedded spreadsheets. OpenDocument does.
Once again, this is just divergent philosophies; OpenDocument doesn't (yet) have definitions about how to typeset documents through different filters. As an XML format, I hope & expect these tools to be built. Why typeset DocBook XML & not OpenDocument?
LaTeX could certainly be part of some monolithic Office Suite, but it is already very good at what it does.
It could be part of it, in the same way DocBook could be a part of it.
Or the same way OpenDocument is part of it...
But LaTeX doesn't have good software surrouding it - where "good" is defined as mass market word processing software.
Neiter are there barriers for this software.
Remember, mass market word processing software is the topic. We need office suite software for a whole government - LaTeX isn't ready for anything more than geeks.
Whatever. I work with people who manage to typeset beautiful documents without knowing any code. They use Scientific Workplace or LyX or even XML formats.
Working software with soon-to-be ISO standards vs a proof of concept.
My point was about the flexibility of the file format. The software does work.
And users insert these graphs in the way they're accustomed to, right?;)
Sure--any TeX editor has a file menu command to insert a jpg/png/pdf image.
OpenOffice.org does PDF presentations, and Flash presentations too.
The PDF presentations done by OO.o's Impress aren't in the same league. The LaTeX presentations can have movies, working links (including auto-generated Outlines and updated per-slide tables of contents) & transition effects. It also allows you to edit the metadata. Of any of the OO.o apps, Impress has the longest way to go.
Does LaTeX excel at any of these? Probably not. But why not do, as others do, and choose tools which DO excel at them.
Good idea. Lets use OpenOffice.org.
Because OO.o excels at NONE of them, with the possible exception of Writer (though I prefer AbiWord for most things other than complicated MS Office files). Gnumeric is a FAR better spreadsheet than Calc; Inkscape better than Draw. Impress is "O.K.," but only because ALL of the alternatives suck.
Look- these are simple things that users want and there's no technical reason why it shouldn't happen and you're brushing off by telling them to manually follow a process the software chould do for them.
Absolutely not. I am saying that the FILE FORMAT (which you protested against) does nothing to prevent this. OpenDocument does nothing to enforce good applications either. I appreciate both file formats. They are both flexible enough and good tools exist for both.
LaTeX just isn't mature enough to have pretenses of being something for anything more than 1% of users.
No. Your can make an argument that the existing tools don't appeal to a broad enough customer base. But the file format itself is mature. And there are both powerful tools for hardcore geeks & GUIs which make it easy (though they are probably aesthetically unpleasing (ugly)) for everyone else.
I don't think you've used datasources in an office package before...
I suspect that OpenDocument adoption is more wide-spread than LaTeX adoption, mostly because OO.o groks both that and MS Word format (the de facto standard).
So the real lesson is that you need to make tools that are good enough & people won't care about the technical merits of the file format. So long as they can work with other people & use their old documents, they won't moan too much.
This is the minimum of what's needed in an office suite
Right--you are arguing different philosophies. LaTeX could certainly be part of some monolithic Office Suite, but it is already very good at what it does. It may even be better than you give it credit for.
Spreadsheets
See the EMACS file as a proof of concept. Something similar could be written in TeX.
graphs
PSTricks & other packages let you add graphs which are generated on the fly.
presentations
I actually like LaTeX Beamer quite a bit--the PDF presentations are fantastic.
Does LaTeX excel at any of these? Probably not. But why not do, as others do, and choose tools which DO excel at them.
a single-file container format so exchange is easy. OpenDocument has it. HTML and LaTeX fails it;
Just zip the needed files together, as OpenDoc does....
* a user interface that regular users can migrate to. OpenDocument has it. HTML has it. LaTeX fails it;
These are file formats. Not interfaces. There are friendly HTML and LaTeX authoring tools.
* macro language (admitedly not standardised in OpenDocument). OpenDocument has it. LaTeX fails it;
This is laughable. LaTeX is VERY scriptable.
* integration with other office formats such as OleDB datasources. OpenDocument has it. LaTeX fails.
No, again--the programs that grok OpenDoc have it. Not the format itself. There are LaTeX tools which can pull data from a database.
Then you generate RTF, PS or HTML files from the LaTeX source. It works like a charm.
I'm a LaTeX fan. But post-process XML file formats with XSLT works FAR better than the LaTeX tools. DocBook XML converted to RTF/HTML look a lot better than starting from LaTeX in many cases. I wish that OpenDoc would leverage this more--using (as the DocBook people do) TeX to typeset nice PDFs. I also wish they were more hand-hackable, but I'd take a GOOD authoring tool which was well-adopted & had enough functionality. OpenDoc is close, but I'll use LaTeX until then.
Indeed, being able to read the various formats is a problem, especially when they're some proprietary binary format. That's why using plaintext, LaTeX and PDF files works so well: they're well documented, non-obfuscated, and are easily transmittable.
Forgot the link to the article...have included all relevants links in this one.
Linux.com ran a story about web development tools.They approach it as "web development tools for Linux," but most are available for win32 and OS X. I have almost no experience with commercial web development tools (except when trying to tidy up their ugly code). I use content management systems/wikis/etc. where possible (so others can add content & no one need worry about the code or an editor) & a text editor () when not. That being said, Bluefish,Quanta, and Nvu are all nice. All of these options are discussed in the article, as is Screem, which I haven't seen first-hand.
Newsforge ran a story about web development tools.They approach it as "web development tools for Linux," but most are available for win32 and OS X. I have almost no experience with commercial web development tools (except when trying to tidy up their ugly code). I use content management systems/wikis/etc. where possible (so others can add content & no one need worry about the code or an editor) & a text editor (vim!) when not. That being said, Bluefish, Quanta, and Nvu are all nice. All of these options are discussed in the NF article, as is Screem, which I haven't seen first-hand.
We tried Mozilla Calendar/Sunbird with a WebDAV server (even though it deleted two calendars upon upload and barfed on a third, my office loves Sunbird's interface)
I'm a huge fan of WebDAV+iCal & I suggest you try again & solve some of the problems you encountered. If needed, automatically backup your WebDAV content and/or choose a better WebDAV module. It is too bad that WebDAV doesn't have true versioning, but there are implementations which do DeltaV versioning, which would solve a lot of this.
Also look into the fledgling CalDAV implementations & projects like Hula (server) and Chandler (client). Very recent binaries of Sunbird also sport CalDAV support.
If you have a 5 gallon jug and a 3 gallon jug of water, and a hose so u can refill any as u please. What are the steps to get exactly 4 gallons of water?
Fill the 5G jug. Pour it into the 3G jug, so you have 2G in the 5G jug. Empty the 3G jug & pour the 2G from the 5G jug into the 3G jug. Refill the 5G jug & finish filling the 3G jug. It will only take 1G, so you will have 4G in the 5G jug.
The cheapest are any of the handfull of non-raided appliances that are out there. They don't cost much more than putting together your own system, unless you cruise Fatwallet (or similar) for greatly-below-MSRP deals. If DIY isn't your thing (or you value your time significantly), you can go with a SNAP server or something similar that does have RAID.
I personally preferred having a standard rack-mountable server which didn't have proprietary software. The best "bang for my buck" a year ago was, as mentioned, eRacks.
Why not use TeX as your typesetter in a GUI front end? Using LyX or a similar program would allow people to organize their ideas without knowing the underlying LaTeX syntax. I know a lot of people use TeX to get output from DocBook XML. Why not OpenDoc too?
They were on the OASIS committee that formulated OpenDocument (MS wasn't). Another member was Corel, makers of WordPerfect. If we can get OpenDocument into Pages, WordPerfect, and Lotus, I think that would be great.
The main loop looks slightly different in every c app. How would you craft a main loop that will work in most every app and not break its functionality?
You might be able to run your virus code on the first execution but then that app won't get run again.
And the damage could be done on first execution (which is why I advocated replacing the main loop). However, you should also be able to just put a call to a malicious function in a predictable place. So, at the start of main or right before you return from it. You can even rename main to mainORIGINAL & call it (passing argv, argc, etc.) from your new malicious main.
but why do that when you can write an ELF virus and spread perfectly to all apps?
If you can compromise the source & not the binaries, or can hit a lot of people who use the source, why not?
I was making a comment from the context of the post you replied to. Most installers/packaging systems require (or at least permit) root to do SOMETHING other than just copying executables.
Yes, moz is bad for suggesting the first use of firefox be as root. Yes, individuals and distros should avoid this step where possible.
No--not all the people who could be infected by this manually and deliberately ran mozilla as root.
Almost all packaging systems allow the execution of arbitrary code. I'll grant that some only use root privileges to copy the programs to the final destination directory. But they're a minority--most allow, for example, at least allow testing the deployed app.
Most packaging systems also allow the installation of setuid programs.
I would think it would be fairly easy to look for *.c and *.cpp & simply replace the main loop.
Regardless of the feasibility self-replicating source code-targetting viruses, I would think targetted attacks against the source of a single application would be just as easy (easier?) than against a binary. You can download the source of Firefox, which someone has infected with the source of this same virus & then it will infect all the binaries on your system.
Both isos and source can suffer the same injection attacks as single binaries. However, I was referring to downloading updated versions of programs on the iso.
The original CD can't have the signatures for software that was released after the CD was pressed (including any updates). Most distros keep signatures on a server. Hopefully a different server than the mirrors which keep the binaries you need to verify (because then two sites would need to be compromised (signature server & the mirror) or the attacker would need to find a collision in the signature algorithm).
Slashdot Mirror
That's funny. TeX is a Turing complete language. You can have interactive programs (like makebst). Every single one of my files does something differently (by design) if the target format is PS or PDF. I have definitions that can be replaced right-up-until compile-time.
Once again, this is just divergent philosophies; OpenDocument doesn't (yet) have definitions about how to typeset documents through different filters. As an XML format, I hope & expect these tools to be built. Why typeset DocBook XML & not OpenDocument?
Or the same way OpenDocument is part of it...
Neiter are there barriers for this software.
Whatever. I work with people who manage to typeset beautiful documents without knowing any code. They use Scientific Workplace or LyX or even XML formats.
My point was about the flexibility of the file format. The software does work.
Sure--any TeX editor has a file menu command to insert a jpg/png/pdf image.
The PDF presentations done by OO.o's Impress aren't in the same league. The LaTeX presentations can have movies, working links (including auto-generated Outlines and updated per-slide tables of contents) & transition effects. It also allows you to edit the metadata. Of any of the OO.o apps, Impress has the longest way to go.
Because OO.o excels at NONE of them, with the possible exception of Writer (though I prefer AbiWord for most things other than complicated MS Office files). Gnumeric is a FAR better spreadsheet than Calc; Inkscape better than Draw. Impress is "O.K.," but only because ALL of the alternatives suck.
Absolutely not. I am saying that the FILE FORMAT (which you protested against) does nothing to prevent this. OpenDocument does nothing to enforce good applications either. I appreciate both file formats. They are both flexible enough and good tools exist for both.
No. Your can make an argument that the existing tools don't appeal to a broad enough customer base. But the file format itself is mature. And there are both powerful tools for hardcore geeks & GUIs which make it easy (though they are probably aesthetically unpleasing (ugly)) for everyone else.
I suspect that OpenDocument adoption is more wide-spread than LaTeX adoption, mostly because OO.o groks both that and MS Word format (the de facto standard).
So the real lesson is that you need to make tools that are good enough & people won't care about the technical merits of the file format. So long as they can work with other people & use their old documents, they won't moan too much.
Does LaTeX excel at any of these? Probably not. But why not do, as others do, and choose tools which DO excel at them.Just zip the needed files together, as OpenDoc does....These are file formats. Not interfaces. There are friendly HTML and LaTeX authoring tools.This is laughable. LaTeX is VERY scriptable.No, again--the programs that grok OpenDoc have it. Not the format itself. There are LaTeX tools which can pull data from a database.
Forgot the link to the article...have included all relevants links in this one.
Linux.com ran a story about web development tools.They approach it as "web development tools for Linux," but most are available for win32 and OS X. I have almost no experience with commercial web development tools (except when trying to tidy up their ugly code). I use content management systems/wikis/etc. where possible (so others can add content & no one need worry about the code or an editor) & a text editor () when not. That being said, Bluefish, Quanta, and Nvu are all nice. All of these options are discussed in the article, as is Screem, which I haven't seen first-hand.
Newsforge ran a story about web development tools.They approach it as "web development tools for Linux," but most are available for win32 and OS X. I have almost no experience with commercial web development tools (except when trying to tidy up their ugly code). I use content management systems/wikis/etc. where possible (so others can add content & no one need worry about the code or an editor) & a text editor (vim!) when not. That being said, Bluefish, Quanta, and Nvu are all nice. All of these options are discussed in the NF article, as is Screem, which I haven't seen first-hand.
Also look into the fledgling CalDAV implementations & projects like Hula (server) and Chandler (client). Very recent binaries of Sunbird also sport CalDAV support.
There's a good write up of this on MathWorld.
13112221
The cheapest are any of the handfull of non-raided appliances that are out there. They don't cost much more than putting together your own system, unless you cruise Fatwallet (or similar) for greatly-below-MSRP deals. If DIY isn't your thing (or you value your time significantly), you can go with a SNAP server or something similar that does have RAID.
I personally preferred having a standard rack-mountable server which didn't have proprietary software. The best "bang for my buck" a year ago was, as mentioned, eRacks.
See if any of the suggestions in Data Storage For Home? apply to you.
I have done DIY and bought a rig from eRacks & am happy with both.
Why not use TeX as your typesetter in a GUI front end? Using LyX or a similar program would allow people to organize their ideas without knowing the underlying LaTeX syntax. I know a lot of people use TeX to get output from DocBook XML. Why not OpenDoc too?
They were on the OASIS committee that formulated OpenDocument (MS wasn't). Another member was Corel, makers of WordPerfect. If we can get OpenDocument into Pages, WordPerfect, and Lotus, I think that would be great.
I use ImageMagick's convert to generate both gif and MPG from multiple still images, but that is because my stills are programatically generated.
Animated GIFs in GIMP
We recently purchased a NAS from eRacks, which we are happy with & is slightly less expensive than the options you posted.
Ther are MANY tutorials on building your own RAIDed NAS on the net. Some have been slashdotted. Here's one. Google for others.
In make.conf FEATURES="userpriv" will "Allow portage to drop root privledges and compile packages as portage:portage"
However, a make install must be run with escalated privs & if that is malicious, you could be hosed.
I was making a comment from the context of the post you replied to. Most installers/packaging systems require (or at least permit) root to do SOMETHING other than just copying executables.
Yes, moz is bad for suggesting the first use of firefox be as root. Yes, individuals and distros should avoid this step where possible.
No--not all the people who could be infected by this manually and deliberately ran mozilla as root.
Almost all packaging systems allow the execution of arbitrary code. I'll grant that some only use root privileges to copy the programs to the final destination directory. But they're a minority--most allow, for example, at least allow testing the deployed app.
Most packaging systems also allow the installation of setuid programs.
Really? Why?
I would think it would be fairly easy to look for *.c and *.cpp & simply replace the main loop.
Regardless of the feasibility self-replicating source code-targetting viruses, I would think targetted attacks against the source of a single application would be just as easy (easier?) than against a binary. You can download the source of Firefox, which someone has infected with the source of this same virus & then it will infect all the binaries on your system.
Both isos and source can suffer the same injection attacks as single binaries. However, I was referring to downloading updated versions of programs on the iso.
1)Would you feel better if the virus infected /usr/bin and /usr/local/bin too?
/bin?
2)Do you have to boot into single-user mode to upgrade the apps in