“‘Siag, it sucks less!’ This is the slogan for Siag Office. This and the self-effacing name for the Siag Office Word Processor, Pathetic Writer, might leave you thinking that this office suite is a mere plaything, a university student’s cobbled-together programming assignment. But don’t be fooled by first impressions. Siag Office is a lightweight suite of applications which might be just the right set of office tools for you, especially if you have older hardware.”
I remember Siag from my first Linux distro, SuSE 5.2 back in 1998. It was pretty much the only half-way decent free office suite for Unix back then. There were a couple of “word-processors” but most of them were no better then Wordpad on Windows. SuSE came with a 60-day trial of some commerical one, called Adabas or something.
However a year or two later StarOffice became a free “non-commerical” download, and then the free OpenOffice, and it blew pretty much all the competition out of the water, including KOffice and the old “Gnome Office” (Abiword, Gnumeric and some other stuff). Despite the commendable work that’s being done, it’s really only of historical interest these days. Much more interesting is the new upcoming KOffice, which would appear to be a free office suite, that the average developer can contribute too (by all accounts OpenOffice’s code-base is frighteningly complicated).
I really wish that Gnome’s Office would take off more. I really like Abiword’s feel, though admittedly it’s strange to me that we have OpenOffice.org that is open source, yet the import of word documents for Abiword isn’t quite up to it’s quality.
By that I mean that with the source code being open, couldn’t Abiword or other office suites just take some of OpenOffice.org’s code and put it into their own project? Abiword is under GPL and OpenOffice.org is under LGPL.
Well if OpenOffice.org has such an incredible mess under the hood it would explain why projects don’t use their code. Gnumeric for one already is probably the best program around for the job it does. But the rest of “Gnome Office” is quite lacking.
How I’ve longed for a native office suite for Gnome.
Just a ‘me too’ comment. Happy with Gnumeric and Abiword, but wish there were a presentation app to go along with them. There was a project called Criawips started a couple of years ago, but it seems to be dead in the water now.
Sorry for the off-topicness. But the review, even while praising Siag, doesn’t inspire much confidence.
KeyJnote is supposed to be a nice presentation program, although I think it’s a Qt app.
http://linuxappfinder.com/package/keyjnote
KeyJnote is supposed to be a nice presentation program, although I think it’s a Qt app.
No it is not a QT program, nor a program for generating presentations. It is a program for displaying PDF presentations with fancy transitions and stuff. It is written in Python and is cross platform.
So you need pdfLaTeX and or Beamer, or else have a template for generating PDF presentations from a word processor.
I tried to create such a template for Abiword recently but came up with a number of obstacles and got nowhere – and I thought it would be an easy job. Anyone out there made one or know of one ?
Am I wrong if I say that essential features have not added for years ? I think that project is not active.
“Happy with Gnumeric and Abiword, but wish there were a presentation app to go along with them.”
I’m usually using pdflatex and xpdf -fullscreen for presentation purposes – or a HTML composition running in an Opera fullscreen (PF11) session.
“””
By that I mean that with the source code being open, couldn’t Abiword or other office suites just take some of OpenOffice.org’s code and put it into their own project?
“””
That’s a really good question. And by that, I don’t mean that they should because its a slam dunk. I mean it’s a really good question to ask.
It seems to me that code reuse is a realy, really hard problem, which many people dismiss as trivial.
People worry that if their license is not iron-clad enough then someone will just steal their code and put it in their project… as though code were currency that can be found on the street and put into one’s pocket.
Transplanting code has a very real cost.
For the last couple of years, I have been trying to settle on a Python web framework. The major contenders are Django and Turbogears.
Turbogears is the pinnacle of the Open Source philosophy. They bring together the “best of breed” of all the projects Python has to offer, and provide glue code to integrate them.
Meanwhile Django is a one stop shopping destination. They roll their own everything.
Turbogears has all of the Python universe (well, the cheeseshop, anyway) to draw from. And Django mostly (but not entirely) sticks to their own stuff.
And yet whereas Turbogears had the limelight in late 2005, and early 2006, Django is now, arguably, the leader.
How can that be?
The answer is complex. But I believe that one of the factors in Django’s favor is that code reuse is hard. And even the lesser task of writing glue code to bind disparate modules (ORM, templating system, application server) together sometimes takes longer than writing the whole frecking stack from scratch with the intent of having the parts integrate. Especially when the project gets to the point of wanting to polish and those criss-cross Frankensteinian suture marks start to become a bit of an embarrassment.
I don’t know if OpenOffice has “a mess under its hood” or not. I imagine they do. Most feature complete software is considered to be a mess by its creators. ( http://www.joelonsoftware.com/articles/fog0000000069.html )
But don’t discount plain, old fashioned, impedance mismatch as the reason for the “islands of code” that we have floating about in the FOSS world.
Edited 2007-07-11 02:26
Turbogears has all of the Python universe (well, the cheeseshop, anyway) to draw from. And Django mostly (but not entirely) sticks to their own stuff.
And yet whereas Turbogears had the limelight in late 2005, and early 2006, Django is now, arguably, the leader.
How can that be?
Fashion.
“””
Fashion.
“””
That’s not true anymore. *Fashion* has *changed*!
Of course, the huge disparity between the two projects’ attitude towards documentation might have made a difference, too. It’s what prompted me to switch.
That’s not true anymore. *Fashion* has *changed*!
Of course, the huge disparity between the two projects’ attitude towards documentation might have made a difference, too. It’s what prompted me to switch.
So TurboGear’s docs are rubbish?
“””
So TurboGear’s docs are rubbish?
“””
Well, saying that would be unfair to the dedicated people who do the hard work of writing TG docs. Here is the problem as I see it.
The TG devs love to write code but they hate to document. So what we see is endless commits of code and the occasional post by this or that developer acknowledging that they really need to write some documentation one day.
So it is left to the users to band together to write the docs.
This is a *hugely* inefficient process, because they have to dig through the code and guess what it does, which is about 90% of the work of documenting it, and then actually write the docs, which is the remaining 10%.
For the 12 months from the initial release up to the TG 1.0 release, the lead developer had a hard and fast rule that if it wasn’t documented it would not be advertised as a TG feature. And 1.0 would not be released until the documentation was complete. (Wink! Wink! Knudge! Knudge! Use the development version instead of 0.8 or 0.9 because it has all the good stuff!)
And then, out of the blue, 1.0 gets released, with hardly any documentation at all for entire areas of the framework, accompanied by some explanation about how it’s hard to release a beta of a beta, so 1.0 had to be released (which I never quite understood), and handed the reigns to a new lead developer at the same time.
It’s been another 10 months now, and docs are still not up to snuff.
So I want to make it very clear that I have a huge respect for the people doing the docs. And at the same time criticize the completeness of the docs. Not being the sort of person who can read code as if it were documentation, as some more talented people seem to be able to do, I followed the situation for 20 months and bailed. (It’s a nice framework, though.)
I’m not sure what policies the Django guys have in place, but whatever they are, they seem to be effective in maintaining a very good standard for documentation.
I love Python. And I love the Python community. But I swear that for a community which goes on about “explicitness” and “doing things right” as much as we do, Python projects can be some of the most poorly documented projects in existence.
Edited 2007-07-11 17:39
That’s a really good question. And by that, I don’t mean that they should because its a slam dunk. I mean it’s a really good question to ask.
It seems to me that code reuse is a realy, really hard problem, which many people dismiss as trivial.
People worry that if their license is not iron-clad enough then someone will just steal their code and put it in their project… as though code were currency that can be found on the street and put into one’s pocket.
Transplanting code has a very real cost.
Yeah, in the general sense I’d agree. But we’re talking just import filters. I’ll admit, I’m not a coder myself, but I would tend to think that something that just interprets a file and then presents it in it’s native format couldn’t be all that difficult to port from one word processor to another, unless of course that word processor just doesn’t support all of the features of the other one.
But for example, I recently received a .doc file that had a logo embedded into it. Open Office opened it up without problems. Abiword didn’t even bother with the logo at all.
The code reuse issue has been a long time question for me. A good example of “Why?” is with instant messengers. GyachE has had voice and video support for Yahoo messenger forever, and yet Gaim (now Pidgin) and Kopete didn’t (I think Kopete finally has Webcam support, but not voice). I’ve always wondered why they couldn’t just get the code and put it in.
I know interface-wise it could be a problem, but we’re talking basically filters and codecs. Mplayer has been able to take binary .dll files from Windows and create a generic interface around them. Same with gstreamer. There are a lot of projects out there that are simply ‘re-inventing the wheel’ as a lot of people say, but with open source, that’s supposed to be one of the advantages, being able to build upon or use code from another project in one you’ve already started. I guess the real question should be, “is the cost to implement our own version of a filter less than or greater than the cost to take code from another project?” And of course the beauty of open source, is that if for example Abiword borrowed code from OpenOffice.org, they could then possibly fix some of the bugs in it, and then give it back to the OO.o guys.
Sorry for going this far off-topic. A bit more on topic… being a Debian user, I just checked to see if siag was still in the repositories. Debian Sid no longer has it in, and I think the last release that may have included it was Woody.
Translation: “I’ll admit, I don’t know what I’m talking about, but despite my own admission of ignorance I’m going to offer a view anyway…”
Reminds me why I stopped posting on this site.
Incidentally, the reasons were
(a) OpenOffice is built on its own custom platform independence library, which means it’s practically written in its own programming language
(b) There are (or at least used to be) very little comments
(c) All the comments are (or at least used to be) written in German
(d) The code for reading in files was very tightly coupled with OpenOffice’s method for handling documents: consequently you had to be an expert on OpenOffice’s file format, and how it managed documents to understand how it read in Microsoft files
People who were coders made a major attempt to “rip out the import filters” back when OpenOffice was opened up, and they all had to admit defeat, even those who spoke German were stumped by the alien, tightly coupled (a bad thing) code-base
I think the main problem is that developers don’t think in terms of libraries. If they put functionality (like Word-ODF import/export) into a library then other developers could use that with their own code.
Libraries like that would also add contributers to the original project, because those who use the library will improve it and give code back
Gnumeric is fine, but there are a couple of things about it that make it hard to use compared to OO in something demanding. One is speed on large files. I’ve found both KOffice and Gnumeric unusably slow to open a file with a few thousand rows.
Second, Gnumeric seems not to be compatible with Excel in terms of array formulas. So, you do an array in OO, it works fine in Excel. You then try to open it in Gnumeric, and nothing computes.
OO has speed problems as well, but it makes a better job of it than either of the other two. Of course if your files are small, any of them will do perfectly well.
There were very few comments in OpenOffice, at least a few years ago, they were all in German, and the code-base was very dense and difficult to understand, being built on an entirely custom platform-independence library (also documented in German). Lastly, the code for reading in Word files seemed to be fairly tightly coupled with OpenOffice’s way of handling documents themselves. Consequently, despite a lot of effort by a lot of people, it just wasn’t possible to simply “copy” it out, or even use it for hints (again due to absence of comments and comments being in German).
Edited 2007-07-11 20:16
ADABAS D was the database program, I think there was Xess for spreadsheet work, and Maple was a mathematics suite
I built it from source on Mandrake 8.2 a few years ago. Quite a fun exercise at the time.
As I remember it it looked quite nice but you kept hitting unexpected gotchas when you tried to use it as a real productivity tool.
I just got a new computer for only 260 CAD and OOo Writer opens in 2 seconds flat in Feisty, so I certainly don’t need a lightweight office suite. On my old system it took 60 – yes 60 seconds.
The Unix world needs a better OpenOffice. OO is big, slow and not reliable (many import/export from/to various formats are not always good). Don’t think bad, OO is a titanic effort but we need something in the Unix spirit, lightweight fast and reliable.
Of course, the suit mentioned in this article is not the question.
Maybe KOffice, but KDE won’t fit into geeks or tech Unix people.
Or finally… maybe the “Office Suite” concept is f***n wrong
“Maybe KOffice, but KDE won’t fit into geeks or tech Unix people.”
I had to laugh…the usual battle of the GUI’s tends to end up with people saying Gnome is nice and simple and KDE is aimed at the geeks and techy Unix people
I find myself drawn to KDE because it seems more responsive with better tools than Gnome, but I like the Gnome layout better than KDE. Maybe KDE and Gnome should get to gether. I bet they would have pretty and smart children.
“Maybe KOffice, but KDE won’t fit into geeks or tech Unix people. ”
I beg to differ.
Well, neither will Gnome, or any other huge “Desktop” environment for that matter. Your hardcore “geeks” and “Unix(TM) nuts”, when they even sue a GUI at all, tend to go with much more simplistic, but equally flexible options like WindowMaker, BlackBox, etc. For that crowd, an X server is simply an efficient way to have a whole bunch of xterms visible at once.
And really they’ll tend to have little need for an “Office Suite”. Need to write up documents and papers, LaTeX does the trick beautifully, and all you need is either vi or emacs (no arguments on which is better, to each their own) to make a professional looking PDF. Spreadsheets for these folks are often only for crunching statistics or making plots. But there’s R and gnuplot for that.
Now many people are very uncomfortable using these tools and want a nice office suite instead, but the geeks and hardcore Unit(TM) nuts tend to not be said people.
Edited 2007-07-11 19:31 UTC
Finding a nice word processor is pretty tricky for Linux. My favorite so far is Abiword – but I find it’s table support lacking (same with Open Office). Tables are pretty simple things of course, it’s just that manipulating them in Abiword is quite a bit more awkward than in Microsoft Word. There also seems to be no way to import a Gnumeric or other table into Abiword, which is rather unfortunate. I suppose I could take a screenshot of the spreadsheet I want, but that’s pretty inconvenient.
OpenOffice is far too monolithic for my use, and as I mentioned above it’s table support is also not great.
Copy and paste works (standard Fedora7 packages here) but it copies only the values; functions over cells won’t be preserved. Not perfect, but better than a screenshot.
Anyway, I love Abiword and Gnumeric. They are both small, fast a fit nicely in my Gnome-centric environment.
Try WYMIWYG for a change: what you mean is what you get.
I have largely abandoned OO for LyX (and the text editor). All change takes some time and effort and the “getting used to” thing, which we humans generally don’t like.
But it is absolutely very rewarding. The word processor paradigm is easy to understand, but it keeps getting in your way.
Damnit ! We have (La)tex ! (and better yet Xe(La)TeX…)
Why use Word or open office to write stuff, when you can type these kind of documents in a breeze with TeX :
http://www.lakedaemon.org/uploads/media/Kanji.pdf
(a little French-Japanese dictionnary)
Notice the index (against Keys, number of Stroke, Lectures) at the end. Can you do that in Word ?
http://www.lakedaemon.org/uploads/media/Mathematicon.pdf
(a math course in french with pictures)
and when with just _1_ compilation (with Tex4ht),
you can convert it into XHTML+Mathml+SVG as you can see it here (works with firefox):
http://www.hekathomb.com/Mathematicon.xml
Now…the only times a real Geek really needs word or open office, are when someone sends him a .doc document (pdf or regular txt files were made for that purpose damnit), when he stumbles on a .doc document on the net or when he wants to write a very short letter with a model (cv, …)
The thing is…(La)TeX documents are way more beautifull/professionnal looking that word documents .
And as TeX is a programmation language
(kinda twisted but easy and fun to use, bug free and powerful)…
1) there is (nearly) no limit to what you can typeset
2) all the commands you learn in TeX brings you benefice for years (decades for the realy adopters)…
My conclusion :
if you use Word, Open Office or another wysiwyg word processor to write complicated stuff, you are NOT a true geek (if you use LYX, you are only half a geek ^_^).
ps : 1) with XeTeX, you can now type (and program your macro) in unicode (power to the masses).
Three TeX for the American with their Ascii charset
Seven TeX for the Asian with their (Shift)-Jis charset
Nine Tex for the european with their accented-charset
But One XeTeX to bind them,
One XeTeX to rule them all and to bind them in the charset of Unicode, where the Shadows lie.
2) with PGF/Tikz, you have a powerful and sane library to program your (svg or pdf) pictures directly inside TeX (no need for postscript, metapost or external programs)…
see examples here :
http://www.fauskes.net/pgftikzexamples/all/
3) with luatex coming (lua integrated in tex)even better stuff comes :
3d-picture (heavy on the calculation side) inside TeX
databases (that would be nice)
…
Edited 2007-07-11 09:14
You misunderstood the whole point of office suites. It’s not to fit the needs of geeks, but the needs of office workers.
Just out of curiosity, when you suggest to use TeX to fit office needs, did your cost estimation include the cost to train regular office staff in a programming language as complex as TeX?
1) I dind’t suggest to use TeX for office needs, I just said that true Geeks use TeX.
2) let me answer to your question :
I) In my university, there was a time when the secretaries were doing all the TeX-Typing work.
This shows that the average worker can do it
come on…
a) to type unformated text in TeX, you don’t need a single TeX command
b) if you want to format it with Sections, Chapters, Paragraphs and do the basic stuff that people using word processor do, you only have to know a few very basic TeX commands
II) TeX is free. Once it has been set up (a process that a distribution such as MikteX makes very easy/user-friendly),
it has few bugs (no bug know for years),
it doesn’t crash,
it is portable,
it runs on all architecture.
These makes the cost of maintenance very low.
III) Now, about the cost of training regular office staff (which is the only real point).
It takes less than 30 minutes to explain and show how TeX works (once it is set up) :
a) you type a .tex file (a regular text file with the .tex extension) with contains TeX commands (those usually starts with a “backslash” char like “backslash-par” to begin a new paragraph, “backslash-eject” to jump to the next page, …
b) you compile it by clicking on a button…it turns the .tex file into a .pdf file (with pdftex)
c) you can see the .pdf file with acrobat
d) if the compilation process b) produced errors, correct them by going back to a) (here it is nice to have a people knowleable with TeX at first)
People using TeX learn fast : is is faster and easier to explain someone that to typeset a section named “Gah” he should just type
backslash-section{Gah}
than to explain it in Word.
Now, I don’t think that the cost of training would be that high. And to get them going, you just have to have a few people proficient with TeX that answer to Email in your company.
(besides TeX support on the internet is great)
After a time, there would be some (wondrous) gains of productivity too : TeX code can be reused.
Copy, Paste and don’t worry… TeX works on the formatting
If you force people to use TeX, they’ll use it (and they’ll become good at it).
People use/are familiar with Word because that’s the first app that they were made to use.
Now, humans are mostly lazy. After they become 14 years old, They just loathe to learn new things.
Even if they are told by many people that they will be much much better afterwards.
Now…I’m not saying that TeX should be used by office workers, I’m only saying that they could use it…mind you ^_^ -(and I’m ranting a bit too)
ps : sorry for the backslash-, OSnews filter the backslash char
Edited 2007-07-11 10:30
I think you overestimate the average person’s ability to deal with anything remotely resembling a programming language.
True ^_^
> 1) I dind’t suggest to use TeX for office needs, I just said that true
> Geeks use TeX.
No problem with that. Actually for some people TeX is a real improvement in productivity, which is justification enough to use it.
> a) to type unformated text in TeX, you don’t need a single TeX command
Actually… no. Even something as simple as an unformatted text using umlaute, accents etc. requires you to use special character combinations such as “a instead of just typing the character you wanted (if you do the latter, you are already asking for trouble because TeX doesn’t handle that correctly). Remember, I’m talking about regular characters, not about formatting.
To enable the umlaut / accent combinations (where you have to wonder why this must be enabled at all), you already have to use the include command (can’t remember the exact name) *and* the package name to use. There are actually at least three different ways to produce umlaute, using different add-on packages and character combinations, for no good reason, and incompatible in a way that may crash TeX with an obscure error message in a totally unrelated package just for enabling umlaute. Trust me, I got all that when I tried it.
If you intent to use *some* formatting, you are probably asking for LaTeX since that includes more useful stuff. BAM! Now you have to start with a documentclass command and more, not to mention the problems due to overfull / underfull boxes.
> b) if you want to format it with Sections, Chapters, Paragraphs and
> do the basic stuff that people using word processor do, you only
> have to know a few very basic TeX commands
Agreed. Once you make it work at all, experimenting and simple formatting is easy.
> Now, I don’t think that the cost of training would be that high.
Is this based on theory or practice?
Sorry, I lied there…to typeset an unformated text in xetex, you need just 1 command : backslash end to tell xetex where the end of your document is (but it wont display your fancy characters).
If you want it to display fancy characters, you have to use two more commands at the beginning of your document say
1)
backslash-font backslash-myfancyfont=”name of your font in your windows font repertory” at 12pt
(to define the backslash-myfancyfont command)
and
backslash-myfancyfont
(to make xetex use your font from now on)
But of course, you’ll want more stuff (sectioning, using a template for your documents) and you will have to load packages/macro files…
But this is part of the set up. You make it once and you never ever think about it.
So…to typeset Japanese, German, French, Indhi and Arabic in the same XeTex, you need 5 fonts and 10 commands…and switching from one langage to another is a breeze…^_^
I say it’s not that bad.
Actually, this is wrong.
If you use xe(la)tex,
you can write in german, english, japanese, chinese, whatever language you like,
XeTeX will work _out of the box_ and will display it all just fine as long as you provide it with a font that includes the characters :
for exemple :
“ms gothic” for Japanese
“arial” for french
Who knows what font for sanscrit…etc…
This is only true for xetex and xelatex though as the older flavour of tex/latex don’t work that well with unicode.
Hell no !
The people who want to have complete control over their document write in TeX
The people who don’t care and who just want to use things made for them and that work out of the box (think Apple) write in LaTeX
(LateX is a layer of extended macros (abstraction, bloat ?) on TeX)
On a side note. You can try xetex now on windows
(linux and mac os X too) with the W32TeX distribution (it looks scary but, surprisingly, it’s quite easy to set up) :
http://www.fsci.fuk.kindai.ac.jp/kakuto/win32-ptex/web2c75-e.html
and as soon as miktex 2.7 is out (2.7beta should be out in august 2007) it’ll be a breeze to set up xetex (as miktex installation is very easy with an installer and a good updater)
> If you want it to display fancy characters, you have to use two more
> commands at the beginning of your document say
> 1)
> backslash-font backslash-myfancyfont=”name of your font in your
> windows font repertory” at 12pt
First, this is already a point of failure: How many employees are going to forget to do this? How much time is wasted by this, including searching for the scribbled note how exactly the file name of the font file is? What about the bad impression to somebody who gets a text with characters simply missing? Consider that all this must be done only to solve a problem that doesn’t exist in word processors at all.
Secondly, your solution sounds very much as if the ability to place umlaute/accents in a text is coupled with using a specific font.
> Actually, this is wrong.
If you try to argue against empiric data, you have no point. Fact is, some totally unrelated package spit completely obscure error messages in my face if and only if I loaded the “nbabel” package to get umlaut support. This never happened with MS Office or OpenOffice.
This problem had cost me 1-2 hours and I could only resolve it because I had the knowledge and the nerves to understand what went wrong. This isn’t the case for most users.
Please, just explain where the use of TeX in an office had increased productivity. Then I’m interested.
> If you want it to display fancy characters, you have to use two more
> commands at the beginning of your document say
> 1)
> backslash-font backslash-myfancyfont=”name of your font in your
> windows font repertory” at 12pt
First, this is already a point of failure: How many employees are going to forget to do this?
This is as it should be ^_^.
He doesn’t have to remember the command to set the font if it is set once and forever for his need by the set up.
How much time is wasted by this, including searching for the scribbled note how exactly the file name of the font file is? What about the bad impression to somebody who gets a text with characters simply missing?
1 minute. The employee opens his macro file where all his macro are defined (once and for all),
the command he is looking for lies in it.
(he can find it by doing a query for “font”)
Consider that all this must be done only to solve a problem that doesn’t exist in word processors at all.
Man…try to write in sanscrit in Word or with a language that type vertically or from right to left.
Can you ?
I can do that in XeTeX in one command to load the font
(you would have to load the font in word too) and one command to change the font. Finished.
Secondly, your solution sounds very much as if the ability to place umlaute/accents in a text is coupled with using a specific font.
Nope. You type tour .tex file in whatever (unicode)-word processor that you like, as you usually do.
Then you compile the .tex file with xetex
If you told XeTeX to use your german font (with 1 command), the pdf produced displays all your fancy stuff…
Fact is, some totally unrelated package spit completely obscure error messages in my face if and only if I loaded the “nbabel” package to get umlaut support.
If you don’t like errors,
You definately aren’t cut out for TeX… ^_^
(you aren’t a geek, sorry ^_^;)
Besides…
error in TeX come from the human (remember that TeX is bug free). No harm done, you correct it, recompile and you have the thing you wanted
errors in word or open office cdon’t come from the human, but from the software. And they tend to crash your word processor (before you saved), to crash your system or to corrupt your data…
This never happened with MS Office or OpenOffice.
LoL…being able to convert in less than a minute
a 200+ pages book full of complicated math formulas and vectorial pictures into xhtml+svg+mathml
(the result being nearly a clone of the printed result)
never happened to me with word or open office
It happens to me in TeX every day ^_^
Come on…TeX and Word have a different set of features.
Word being “user-friendly and usefull for writing letters/reports”
TeX being “quality oriented” for the power user.
> Now, I don’t think that the cost of training would be that high.
Is this based on theory or practice?
I concede the point. This is pure speculation on my part. I guess that you could spend years before you found someone who had experience training people to TeX and who know the answer in practice to your question.
Tex spreads through word of mouth.
What happens is that Texnicians
train beginner in TeX out of love for TeX.
Word is much too popular. It killed everything else.
I’m angry when I see that nowadays the mesure of a word processor is not it’s features but that fact that it can or cannot import .doc documents.
> Word is much too popular. It killed everything else.
And still you don’t ask, “why?”
I agree with the LaTeX pamphlet (KNUTH!)
Not only for geeks thou. Everyone should have grown up with LaTeX. LaTeX on elementary school!
Work-like programs have all the same deflect, its hard to tell the program how you text should be saved. Like a newbie trying to webdesign in dreamweaver without knowing HTML.
>>This and the self-effacing name for the Siag Office Word Processor, Pathetic Writer, might leave you thinking that this office suite is a mere plaything, a university student’s cobbled-together programming assignment.
Yes, and frankly, I still think that. Sorry, but MS Office 6.x runs circles around this. (Yes, I have used Siag Office and most versions of MS Office.)
I think the currently available free office suites for Linux are a train wreck. OpenOffice is powerful, reasonably compatible with MS Office, somewhat stable, but dreadfully, unusably slow on all but the newest hardware. Gnome Office is unstable, quirky, lacking some important features, has poor MS Office compatibility, and is slow on some operations. And Siag Office? No MS Office compatibility? I have to program for myself the most common missing features? Sorry, but no.
If they make KOffice 2 run without KDE dependancies and if they improve the MS Office compatibility, it might be an option. I sure hope it can be, because MS Office and Windows cost too much.
If they make KOffice 2 run without KDE dependancies
That would be brilliant! Then they could reimplement all the stuff they get from using KDE libraries. (since this is the internet, I’d better make it clear that was sarcasm) A kdelibs dependency isn’t as bad as a KDE dependency.
OOo had the bright idea of using their own toolkit (GSL/VCL/etc) and such. Is creating, maintaining and extending a toolkit really something people need to do when they could be creating an office suite? You might say Koffice could keep the Qt dependency and drop the kdelibs dependency, but they don’t have it for no reason. KDE has a lot of convenience classes, tweaks etc over Qt, and installing kdelibs (or having a copy of it shipped with koffice on Windows say) isn’t too high a price IMHO.
As you say though, Koffice really need some help with Office compatibility though (that right there is the answer if anyone ever wonders why MS didn’t consider using ODF). Kspread could use some help loading large files: from 1.4 to 1.5 loading time seemed to quintuple at least. Kchart has some retarded years old bugs dealing with creating pie charts.
I keep hoping it will get some love as an alternative lightweight office suite, but it isn’t forthcoming. If it does get some attention, I wonder if it will turn out like khtml. Khtml was a light html engine, but the love it got was a company forking it and then a bunch of people clamoring to replace khtml with the fork
Some people apparently only have 50 megabytes of harddisk space and simply cannot afford to have KDE libraries installed. As such, they are stuck with OpenOffice or GNOME Office. Poor them.
>>Some people apparently only have 50 megabytes of harddisk space and simply cannot afford to have KDE libraries installed. As such, they are stuck with OpenOffice or GNOME Office. Poor them.
It´s not the space, it´s the kdeinit processes and other related KDE bloat that make a KDE program on a lightweight desktop feel like a fish out of water. KDE apps take FOREVER to start up when they are used outside of KDE. I imagine that most of this will change with KDE 4, though. I hope it does.
With all the talk about TeX, shouldn’t you be writing everything in XML and then processing it into TeX, so you can use your document in other areas (web, etc)?
Perhaps http://getfo.org/texml/ would make more sense?
a well written TeX document
(with sane macro definitions and names, tabs and space)
is easily read and debugged by a human.
It is fun to write, easy (once you understand it’s principle) and powerful
Besides, with TeX, I can produce Xhtml+svg+mathml that looks like my printed documents
In my experience, it is really horrible to look at xml code.
Why should we write Xml code ?
For the sake of flexibility, we lose too much (speed, maintanability of the code, etc…).
Not every thing has to be XML…
Edited 2007-07-11 16:32
a TeX command is not the same as LaTeX command
I was always a fan of siag from early in its life. It was with great sadness that I watched attention shift to koffice, a mythical gnome office and the open source star office when it was released. Siag had the potential, had the architecture and was *here now* at a time when Linux needed an office suite *now*. Years later siag is still where it was–ready to be infused with the features that would make it complete.
I think it might have been a traditional fear of lisp-type languages that prevented siag from ever becoming what it had the potential to be. I’ve seen the same fear drive excellent systems written in guile into disuse… and let us not forget sawfish, the gnome window manager that even today wipes the floor with metacity but has the frightening down side of being written in a lisp-based language.
By now it’s much too late for siag. It has no momentum and has too much catching up to do. But, at one time, it could have been the next great thing in Free software.
The article’s title is “Siag office is far from pathetic”, but ironically, the author lists all kinds of caveats and shortcomings in the article, seemingly blind to their severity or how annoying they can be.
But I’m not bashing him at all for that. It just reminds me of us people that still use the DOS version of WordPerfect. Yeah, WP 5.1 DOS is ancient and lacks a lot of functionality, but we still keep hacking it into working on modern operating systems. We see potential and use in it, even though the rest of the world’s left it behind. I find something immensely comforting about its blue screen and completely unintrusive, minimalist interface. It’s like coming home to an old friend.
I remember Siag from my early Linux days. Very fast and capable little set of programs for its time. StarOffice/OpenOffice was dog slow back then, to the point of being unusable for me, and AbiWord didn’t have tables in those days. Then again, neither did Pathetic Writer, but I had no trouble doing the spreadsheet hack to make pseudo-tables.
Edited 2007-07-11 19:00
Reveal Codes.
I’m a WordPerfect user and I’m still waiting for a good open source replacement. Well either that or a good Linux version of WordPerfect.
Anyway, I love all these open source office suites, but for me none are anywhere near good enough to replace WordPerfect. That includes OpenOffice (which I keep around because it does a much better job of opening Word docs).
However, regardless of my feelings I’m glad to see people out there at the very least trying. WordPerfect is my office suit of choice, but it’s a dinosaur and lacks modern features like Unicode support which will probably continue to make it a less attractive program to a world that knows no boundaries. I only hope that some open source program will be able to step up and take it’s place.