I know that you prefixed this with "in general," but I still think you're wrong.
In general open source programmers scratch an itch.
Ah...But who is to say what itch you have?
If you really love engineering, you'll typically want to spend more time engineering..ing than, say, programming. Therefore you'll likely never write your own tools if there's something available that you can use out of the box.
Ah...but writing your own programs can be quite helpful (many mechanical engineers do some machining (sometimes making their own tools with which to make the final product) & many electrical engineers do some electronics (sometimes programming a chip to do what they want it to do in the finished product). MANY people, particularly in academic or government employment, write their own code. It is often a prerequisite to do so for even the very good FE software, such as ABAQUS. So it CAN scratch an itch.
This is almost always done when no comparable code exists. It is sometimes done when there is an expense associated with the comparable code. It is also often done to just give the coder a deeper understanding of what a program is doing. Many DON'T want a blackbox, and that is all they have with most proprietary software.
You don't see many open source knitting programs. Same for cat breeding and many other areas where programming doesn't naturally flow.
I actually see about as many open source programs which are meant for these as proprietary applications--those niches are much smaller than engineering programs.
If anything, however, engineering is one of the closest disciplines to programming, and there is a lot of OS engineering software out there.
Agreed, but not enough! I think fulltime programmers were the first to embrace open source & a lot of tools other people (including engineers) write end up being just for themselves. When released, they're often not released with a license. If there IS a license, it often isn't a F/OSS license.
A lot of this is culture. Some is culture that should be addressed--many feel that their code is "ugly" (a lot is, but ugly code is often better than no code--at worst, you ignore both, but at best ugly code can be refactored). Some feel that if they release their code, others may modify the code & not contribute back to the F/OSS community. This seems especially true in academia, where people who make models or analysis routines are reluctant to let others test their model and/or to improve their model or analysis routines without letting anyone else have those improvements. I don't really know how/if this subtle point can be addressed.
I posted previously on CAD/CAM/CAE/FE. Now for more genereal purpose (numerical) math software.
As has been suggested, GNU Ocatve is great. It is mostly compatible with MatLab, and even more powerful if you use Octave Forge. It satisfies all of your requirements: it uses "standard" matlab-like files, and is stand-alone. Some of the apps which say they require Matlab may work under Octave. If they don't, a few bug reports might urge the developers to try to make sure their programs work on Octave too (many F/OSS developers have at least sympathy for those who were looking to create a complete F/OSS foundation). If not, you might have found a good project to take on: it could be used by you and others & will also expose what the programs are doing.
Python is also quite nice. Particularly when you extend it with scipy, scientific python, numeric or numarray, ipython, matplotlib, and other libraries and tools for numerical analysis. Some of my past work was in Octave. Now much of it is in python. I use ipython as my calculator & my saved python code interfaces with Grace to make beautiful 2D plots, etc. Octave and Python are both able to call external programs/scripts, so you don't have to be tied to one language.
Mostly from http://en.wikipedia.org/wiki/List_of_open_source_s oftware_packages. I've used most, usually for some part of a design or analysis. You could do engineering with only these (people used to not use computers at all), but you are correct that they aren't always "polished." They do, however, work fine for the patient, idealistic hobbyist who doesn't want to spend much money:
CAE Linux - A LiveCD which lets you try out SALOME and Code-Aster
I'd mod you up, but just spent my last point. So I'll give the common response instead:
"I just looked up what programs seem to be dedicated to doing what "obscure expensive niche software that is only of interest to a tiny audience" does. I used Wikipedia's List of open source software or a google search or something similar. If I actually cite my source, it is only to ridicule 1)you for asking in the first place 2)Slashdot editors for actually posting your garbage. I haven't actually used the software, but it rules."
If you have knowledgable IT, Amanda is nice--it will let you spend money on a nice tape changer and media, rather than expensive backup software that is often flakier than Amanda. If you don't have knowledgable IT, I'd actually say the next-best would be to out-source the backups.
I agree with your points, but would also add that an aggregator also gives you some things that a web browser doesn't.
For one, you can save locally-cached copies of posts. Yes, a web browser also has a cache, but you can't typically have both easy and fine-grained control of the content you keep or throw away. Some sites that have feeds have mediocre connectivity (and feeds were originally promoted partly as a bandwidth saver--you don't download as much content at once). Some authors have a nasty habit of deleting the best content. By archiving it in an aggregator, you can save the best stuff.
Aggregators also let you search over all relevant feeds and only those feeds. No more dealing with separate search engines, with their separate "advanced search" syntax (or, worse, very basic or non-existent searches).
Finally, an aggregator lets you apply filters so that the best, most relevant content sees your eyes & bad/spammy content doesn't. I keep my feeds in Thunderbird, and treat some blogs as email--I apply Bayesian filters to particularly noise-filled feeds (such as comment feeds), and sort content topically. Some aggregators eliminate or group related posts that come from different feeds. Some let you push these posts (which have the most "buzz") to the top, so you don't miss it.
TWIS is a somewhat entertaining college radio show that is podcast. Not the most thorough of science reporting, but digestable & there is some good stuff. Berkeley Groks is in the same vein, but far nerdier.
I've read Asa Dotzler and you, sir, you are no Asa Dotzler. For one thing, he actually had experience with what he was talking about. This isn't to say I agreed with Asa on most points (I do run desktop Linux), but his arguments didn't seem to be from 10 years ago. Has Greg actually run Linux? Which distro? Where are his credentials?
I am a strong believer in F/OSS. Not only do I run it, but I contribute code and money (I also submit bug reports, patches, contribute answers to forums, and am an advocate of F/OSS to my peers). In short, I drink the Kool-Aid. I use only Linux and FreeBSD on machines I own (and I bought the machine on my desk at work, so I can use it there). I am a zealot. I think that an all (or mostly) F/OSS stack is something to strive for & that a lot of F/OSS software does work better on a free OS (usually because that is what the developers write it for & where it gets the most complete and knowledgable testing).
That being said, I do still sometimes have to use Windows & I am happy to have F/OSS on that platform. I patch my own code to work around bugs that only impact Windows users & I have financially supported projects on that platform. I have even given money to good F/OSS software which is only on that platform. I am certainly not alone. Just look at the top projects on sourceforge. Most run on Windows. Some run only on Windows.
So...some of the best Windows-only/Windows-mostly F/OSS: Filezilla--great (S)FTP client/server. Hopefully a *nix port soon. 7-zip--excellent compression software. p7zip is there for the rest of us, but updates take a while to reach us. PuTTYFor your ssh/scp/sftp needs.
I've given money to these projects & carry them around on a USB key (along with Thunderbird, Firefox, and vim). Cygwin is another handy thing to have if you have to be on win32 for very long.
Screen is great & not hard to use. See articles on kuro5hin or Wikipedia.
For a brief synopsis.... screen has a number of useful features, most of which you can access by prefixing a command with a command key, which by default is 'Ctrl-A' ('C-A' or '^A' if you prefer). To see some of the things screen can do, press 'Ctrl-A' and then '?'.
screen allows you to have multiple terminal windows in a single terminal session ("multiplexing"). This is REALLY useful if you spend a lot of time in a term (for example, when you're SSHing to other boxes). To create a new window, 'Ctrl-A' and then 'C'. To switch between windows, 'Ctrl-A' and then 'N' or 'P' for next & previous windows, or a number to switch to a specific window.
screen allows you to recover a session, so you can gracefully detach (or be rudely interrupted by a network outage) & reattach to see all of your processes again. To reattach to an existing screen, start screen with 'screen -r'. You can also forcibly detach an already-attached session by 'screen -Dr'. If you have multiple screen sessions, you can tell it which session to reattach to by typing in the process number after the '-r'.
You can also lock the screen if you are stepping away for a shor period ('Ctrl-A' and then 'X'), and even have a bigger scrollback buffer which is copy/pastable from the keyboard! ('Ctrl-A' and then '[' to enter copy mode. Use vi-movement keys, cursor keys, etc. to move the cursor, hit 'enter' to start a selection, use the same movement keys to expand it, hit 'enter' again to put it in the buffer. 'Ctrl-A' and then ']' to paste).
screen allows a lot of other useful things, such as sharing a session between two users (which makes for interesting pair programming or terminal-style "presentations").
It is definitely one of the first apps I install on any system.
that liability is $EXPERT_HOURLY_RATE * $HOURS_TO_REINSTALL_ENTIRE_NETWORK + $COST_TO_ORGANISATION_OF_TIME_LOST_DOING_SO, because once you're compromised, nothing less is safe.
Competent network admins should be able to protect against internal attacks, as well as external ones. Indeed, targetted malicious attacks are more likely to come from within the firewall & usually do the most damage. One rooted box should never be capable of making it so you have to reinstall the entire network.
If you aren't prepared/able to underwrite such a sum, you have no business ignoring IT's policy and using a non-standard set-up, end of story.
IT policy in this case should be based in risk management. The figure-of-merit should be COST_OF_COMPROMISE * LIKELIHOOD_OF_COMPROMISE. Both numbers are typically lower for *nix distros (the later because *nix boxes typically don't spread win32 worms). ANY genuine need for the OS should merit the risk.
Contrary to popular opinion, not all sysadmins are stupid, draconian power freaks.
Not all are as smart or capable as they fancy themselves. The same is true of the staff, but IT is a service and should help people do their job, rather than hinder them. I have had the privilege of using the services of IT departments who understood this. I have also put up with those who don't.
Some of them just take security seriously. (Not saying that's necessarily the case here, but don't assume a policy like this is unreasonable just because it's inconvenient.)
Given that other IT departments in the corporate, public, and educational sectors all manage to run heterogeneous networks that require more security than this one probably has, this IT department would probably be hard-pressed to actually demonstrate security gains by excluding all *nix boxes.
that liability is $EXPERT_HOURLY_RATE * $HOURS_TO_REINSTALL_ENTIRE_NETWORK + $COST_TO_ORGANISATION_OF_TIME_LOST_DOING_SO, because once you're compromised, nothing less is safe.internal attacks, as well as external ones. Indeed, targetted malicious attacks are more likely to come from within the firewall & usually do the most damage. One rooted box should never be capable of making it so you have to reinstall the entire network.
If you aren't prepared/able to underwrite such a sum, you have no business ignoring IT's policy and using a non-standard set-up, end of story.
IT policy in this case should be based in risk management. The figure-of-merit should be COST_OF_COMPROMISE * LIKELIHOOD_OF_COMPROMISE. Both numbers are typically lower for *nix distros (the later because *nix boxes typically don't spread win32 worms). ANY genuine need for the OS should merit the risk.
Contrary to popular opinion, not all sysadmins are stupid, draconian power freaks.
Not all are as smart or capable as they fancy themselves. The same is true of the staff, but IT is a service and should help people do their job, rather than hinder them. I have had the privilege of using the services of IT departments who understood this. I have also put up with those who don't.
Some of them just take security seriously. (Not saying that's necessarily the case here, but don't assume a policy like this is unreasonable just because it's inconvenient.)
Given that other IT departments in the corporate, public, and educational sectors all manage to run heterogeneous networks that require more security than this one probably has, this IT department would probably be hard-pressed to actually demonstrate security gains by excluding all *nix boxes.
There are a number of ways to get around arbitrary rules. Either overtly or covertly.
if you ask the school to put machine on your desk, you get a Windows box. Faculty who want to run... Linux have had to provide their own machines
You can ask for permission to dual-boot. Or, if you already have permission to install your own software, you can do it covertly. I would not advise wiping the Windows partition--you can boot into it when IT starts snooping around & also some might have a problem with you removing licensed software. Failing this, run from a LiveCD/USB key. Or run coLinux or run it under QEMU, VMWare, or similar.
Great consternation has ensued in the faculty senate
Cause greater consternation & bring it over IT's heads. Bring it to the President of the school or the trustees. An army of pissed off faculty will beat a lazy IT head any day.
Any suggestions on how to deal with this?
In addition to the above, you can probably ask for a special exception & say you are willing to take the blame if your FreeBSD box gets rooted. Once you show minimum competency & need, as well as the willingness to put your ass on the line instead of theirs, IT will probably cave.
Effective arguments to use?
The most effective argument is you can't otherwise do your job. Show that you need FreeBSD. Another good argument is obviously to point out the past infections of campus-maintained machines. Tell them you'll firewall your machine from the University network, both to protect you from it & it from you.
Corel WordPerfect and, to a lesser extent, Lotus SmartSuite, are both mature products that are already being using by blind and visually impaired users. Corel supports the transition to OpenDocument. IBM/Lotus supports the transition too. Now is the time to actual implement OpenDocument. Both suites already have XML support & using a new schema shouldn't be a hard feature to add. Either or both companies could then contest the claims (and increase sales of their product(s) to boot).
I do realize that Sun's StarOffice and OpenOffice both support various 3rd-party tools to aid the visually impared (including some of those which ship with MS Windows), but Corel's support is almost as strong as MS Office's support. Having a few more companies pushing this would help against the Redmond political wing. It is ridiculous that MS is the only major office suite vendor that has no plans to give OpenDocument support.
I use GNU R for my heavy lefting. If Excel isn't meant to do statistics right, then maybe it shouldn't have statistical functions! That Gnumeric and others can get much of this right is a sign that spreadsheets can be decent at this--maybe not as good as SPSS/Stata/SAS/R/S+, but still better than Excel.
If people complain about OO.o Calc or Gnumeric not importing their Excel spreadsheets (as OP had done), they are almost certainly using Excel beyond the basics.
I'm not complaining that my hammer can't handle screws. I'm complaining that MS sells toolboxes that claim to be able to deal with screws, but that none of the tools that appear to be screwdrivers actually work. Only it is worse than that: if I had screws, I'd know they had failed--I would see they were stripped. When Excel @#%@#$s up, you haven't a clue.
Who said it was a "scientific article?" Take up concerns over phrasing with Drexel Assoc. Prof. B. D. McCullough. But please point out places where he defended Gnumeric's mistakes. In most cases, he noted Gnumeric fixed an error & MS Excel didn't. This isn't saying the Gnumeric errors were good. But that it is good they were fixed.
Indeed, he criticizes Gnumeric: "On this basis alone, the RNG (random number generator) in Gnumeric can be judged as unacceptable for statistical purposes."
Early versions of Dia did suck. It has gotten a lot better. I do like kivio more. But agree that OO.o Draw makes a good, basic vector graphics program (and Inkscape more so).
I assume I was modded troll by someone who didn't realize something from Redmond can contain mistakes. F/OSS also has errors, but one hopes they can get fixed. Which is what the first link said--Gnumeric replicated errors of Excel and, when statisticians complained, Gnumeric got fixed & Excel didn't.
For those interested in Excel errors, here are other sources:
Can someone please tell me what the big deal is with Visio? On the commercial win32 front, SmartDraw seems just as capable & usually cheaper. On the F/OSS front, Kivio and dia rock.
While I run Linux on all of my machines, I must maintain win32 machines at work. You can use teTeX and LyX both natively and under cygwin. You can use JabRef on any platform.
You can run most popular windows apps under Wine, but I'd suggest using the fine native or online software.
I love GnuCash--especially for stock management (you can make a cron script to download quotes as periodically as you want to track, which Quicken & MS Money both lack). It takes some work to do auto-downloads of US financial institutions, but it is possible. I'd actually advise against it--if you record transactions as you make them (you can do this when watching TV), you can spot bank errors or improper charges easily & then reconcile when you reconcile with your monthly statement. Both MS Money and Quicken encourage this too, but few people take advantage of it.
Taxes are trickier. There is F/OSS tax software, but it just makes the process of manually doing your taxes a little easier. I'd suggest cruising over to IRS's taxfreedom site--there are scores of online tax preparation utilities which have interfaces similar to TurboTax. The IRS can direct most people to one that won't cost them anything. E-filing is fast, easy & gets your refund quicker. You can actually track your accounts online with an aggregator such as yodlee, but I find Gnucash or others to be so handy when making customized reports.
And you can't defend your statement that the FILE FORMAT is bad? You don't even mention XSLT transformations of XML, or that TeX is TOO scriptable (which can allow for insecure documents).
I didn't post screenshots because I thought we were arguing file formats, rather than applications. But perhaps you stopped arguing file formats when you realized that TeX was certainly powerful enough to be the backend for documents. Similarly, you stopped arguing about (X)HTML when you realized that there were apps which people would almost universally agree were both easy and versatile (though I didn't even mention the spreadsheets & presentation software available for the format or that javascript allows most of the scriptability and dynamic content that you crave). I feel you have no arguments against the file formats because you don't understand the file formats.
But, if it makes you feel better, here's what my peers used (most were mentioned in my post, but I hadn't linked them for you & apparently you can't use a search engine):
AbiWord (F/OSS, not the best TeX editor (but I had none of the problems you had))
None of these apps are Motif based (which allows native widgets anyway). LyX used to be, but now uses XForms and Qt, with a GTK interface in development). Easiest/best looking apps are probably SWP and LyX, depending on your platform.
Slashdot Mirror
This is almost always done when no comparable code exists. It is sometimes done when there is an expense associated with the comparable code. It is also often done to just give the coder a deeper understanding of what a program is doing. Many DON'T want a blackbox, and that is all they have with most proprietary software.I actually see about as many open source programs which are meant for these as proprietary applications--those niches are much smaller than engineering programs.Agreed, but not enough! I think fulltime programmers were the first to embrace open source & a lot of tools other people (including engineers) write end up being just for themselves. When released, they're often not released with a license. If there IS a license, it often isn't a F/OSS license.
A lot of this is culture. Some is culture that should be addressed--many feel that their code is "ugly" (a lot is, but ugly code is often better than no code--at worst, you ignore both, but at best ugly code can be refactored). Some feel that if they release their code, others may modify the code & not contribute back to the F/OSS community. This seems especially true in academia, where people who make models or analysis routines are reluctant to let others test their model and/or to improve their model or analysis routines without letting anyone else have those improvements. I don't really know how/if this subtle point can be addressed.
So why not use Octave, which IS free as in libre?
Octave seems to have a bigger developer and user community (with Octave-Forge & similar repositories).
I posted previously on CAD/CAM/CAE/FE. Now for more genereal purpose (numerical) math software.
As has been suggested, GNU Ocatve is great. It is mostly compatible with MatLab, and even more powerful if you use Octave Forge. It satisfies all of your requirements: it uses "standard" matlab-like files, and is stand-alone. Some of the apps which say they require Matlab may work under Octave. If they don't, a few bug reports might urge the developers to try to make sure their programs work on Octave too (many F/OSS developers have at least sympathy for those who were looking to create a complete F/OSS foundation). If not, you might have found a good project to take on: it could be used by you and others & will also expose what the programs are doing.
Python is also quite nice. Particularly when you extend it with scipy, scientific python, numeric or numarray, ipython, matplotlib, and other libraries and tools for numerical analysis. Some of my past work was in Octave. Now much of it is in python. I use ipython as my calculator & my saved python code interfaces with Grace to make beautiful 2D plots, etc. Octave and Python are both able to call external programs/scripts, so you don't have to be tied to one language.
I'd mod you up, but just spent my last point. So I'll give the common response instead:
"I just looked up what programs seem to be dedicated to doing what "obscure expensive niche software that is only of interest to a tiny audience" does. I used Wikipedia's List of open source software or a google search or something similar. If I actually cite my source, it is only to ridicule 1)you for asking in the first place 2)Slashdot editors for actually posting your garbage. I haven't actually used the software, but it rules."
If you have knowledgable IT, Amanda is nice--it will let you spend money on a nice tape changer and media, rather than expensive backup software that is often flakier than Amanda. If you don't have knowledgable IT, I'd actually say the next-best would be to out-source the backups.
I agree with your points, but would also add that an aggregator also gives you some things that a web browser doesn't.
For one, you can save locally-cached copies of posts. Yes, a web browser also has a cache, but you can't typically have both easy and fine-grained control of the content you keep or throw away. Some sites that have feeds have mediocre connectivity (and feeds were originally promoted partly as a bandwidth saver--you don't download as much content at once). Some authors have a nasty habit of deleting the best content. By archiving it in an aggregator, you can save the best stuff.
Aggregators also let you search over all relevant feeds and only those feeds. No more dealing with separate search engines, with their separate "advanced search" syntax (or, worse, very basic or non-existent searches).
Finally, an aggregator lets you apply filters so that the best, most relevant content sees your eyes & bad/spammy content doesn't. I keep my feeds in Thunderbird, and treat some blogs as email--I apply Bayesian filters to particularly noise-filled feeds (such as comment feeds), and sort content topically. Some aggregators eliminate or group related posts that come from different feeds. Some let you push these posts (which have the most "buzz") to the top, so you don't miss it.
TWIS is a somewhat entertaining college radio show that is podcast. Not the most thorough of science reporting, but digestable & there is some good stuff. Berkeley Groks is in the same vein, but far nerdier.
I've read Asa Dotzler and you, sir, you are no Asa Dotzler. For one thing, he actually had experience with what he was talking about. This isn't to say I agreed with Asa on most points (I do run desktop Linux), but his arguments didn't seem to be from 10 years ago. Has Greg actually run Linux? Which distro? Where are his credentials?
I am a strong believer in F/OSS. Not only do I run it, but I contribute code and money (I also submit bug reports, patches, contribute answers to forums, and am an advocate of F/OSS to my peers). In short, I drink the Kool-Aid. I use only Linux and FreeBSD on machines I own (and I bought the machine on my desk at work, so I can use it there). I am a zealot. I think that an all (or mostly) F/OSS stack is something to strive for & that a lot of F/OSS software does work better on a free OS (usually because that is what the developers write it for & where it gets the most complete and knowledgable testing).
That being said, I do still sometimes have to use Windows & I am happy to have F/OSS on that platform. I patch my own code to work around bugs that only impact Windows users & I have financially supported projects on that platform. I have even given money to good F/OSS software which is only on that platform. I am certainly not alone. Just look at the top projects on sourceforge. Most run on Windows. Some run only on Windows.
So...some of the best Windows-only/Windows-mostly F/OSS:
Filezilla--great (S)FTP client/server. Hopefully a *nix port soon.
7-zip--excellent compression software. p7zip is there for the rest of us, but updates take a while to reach us.
PuTTYFor your ssh/scp/sftp needs.
I've given money to these projects & carry them around on a USB key (along with Thunderbird, Firefox, and vim). Cygwin is another handy thing to have if you have to be on win32 for very long.
Screen is great & not hard to use. See articles on kuro5hin or Wikipedia.
For a brief synopsis....
screen has a number of useful features, most of which you can access by prefixing a command with a command key, which by default is 'Ctrl-A' ('C-A' or '^A' if you prefer). To see some of the things screen can do, press 'Ctrl-A' and then '?'.
screen allows you to have multiple terminal windows in a single terminal session ("multiplexing"). This is REALLY useful if you spend a lot of time in a term (for example, when you're SSHing to other boxes). To create a new window, 'Ctrl-A' and then 'C'. To switch between windows, 'Ctrl-A' and then 'N' or 'P' for next & previous windows, or a number to switch to a specific window.
screen allows you to recover a session, so you can gracefully detach (or be rudely interrupted by a network outage) & reattach to see all of your processes again. To reattach to an existing screen, start screen with 'screen -r'. You can also forcibly detach an already-attached session by 'screen -Dr'. If you have multiple screen sessions, you can tell it which session to reattach to by typing in the process number after the '-r'.
You can also lock the screen if you are stepping away for a shor period ('Ctrl-A' and then 'X'), and even have a bigger scrollback buffer which is copy/pastable from the keyboard! ('Ctrl-A' and then '[' to enter copy mode. Use vi-movement keys, cursor keys, etc. to move the cursor, hit 'enter' to start a selection, use the same movement keys to expand it, hit 'enter' again to put it in the buffer. 'Ctrl-A' and then ']' to paste).
screen allows a lot of other useful things, such as sharing a session between two users (which makes for interesting pair programming or terminal-style "presentations").
It is definitely one of the first apps I install on any system.
Corel WordPerfect and, to a lesser extent, Lotus SmartSuite, are both mature products that are already being using by blind and visually impaired users. Corel supports the transition to OpenDocument. IBM/Lotus supports the transition too. Now is the time to actual implement OpenDocument. Both suites already have XML support & using a new schema shouldn't be a hard feature to add. Either or both companies could then contest the claims (and increase sales of their product(s) to boot).
I do realize that Sun's StarOffice and OpenOffice both support various 3rd-party tools to aid the visually impared (including some of those which ship with MS Windows), but Corel's support is almost as strong as MS Office's support. Having a few more companies pushing this would help against the Redmond political wing. It is ridiculous that MS is the only major office suite vendor that has no plans to give OpenDocument support.
I use GNU R for my heavy lefting. If Excel isn't meant to do statistics right, then maybe it shouldn't have statistical functions! That Gnumeric and others can get much of this right is a sign that spreadsheets can be decent at this--maybe not as good as SPSS/Stata/SAS/R/S+, but still better than Excel.
If people complain about OO.o Calc or Gnumeric not importing their Excel spreadsheets (as OP had done), they are almost certainly using Excel beyond the basics.
I'm not complaining that my hammer can't handle screws. I'm complaining that MS sells toolboxes that claim to be able to deal with screws, but that none of the tools that appear to be screwdrivers actually work. Only it is worse than that: if I had screws, I'd know they had failed--I would see they were stripped. When Excel @#%@#$s up, you haven't a clue.
Who said it was a "scientific article?" Take up concerns over phrasing with Drexel Assoc. Prof. B. D. McCullough. But please point out places where he defended Gnumeric's mistakes. In most cases, he noted Gnumeric fixed an error & MS Excel didn't. This isn't saying the Gnumeric errors were good. But that it is good they were fixed.
Indeed, he criticizes Gnumeric: "On this basis alone, the RNG (random number generator) in Gnumeric can be judged as unacceptable for statistical purposes."
Early versions of Dia did suck. It has gotten a lot better. I do like kivio more. But agree that OO.o Draw makes a good, basic vector graphics program (and Inkscape more so).
For those interested in Excel errors, here are other sources:
Can someone please tell me what the big deal is with Visio? On the commercial win32 front, SmartDraw seems just as capable & usually cheaper. On the F/OSS front, Kivio and dia rock.
Both do a better job at QIF import than MS Money. See my FAQ (linked elsewhere in this thread) for which F/OSS apps have QIF import.
While I run Linux on all of my machines, I must maintain win32 machines at work. You can use teTeX and LyX both natively and under cygwin. You can use JabRef on any platform.
I maintain a FAQ on F/OSS finance applications over on FatWallet.
You can run most popular windows apps under Wine, but I'd suggest using the fine native or online software.
I love GnuCash--especially for stock management (you can make a cron script to download quotes as periodically as you want to track, which Quicken & MS Money both lack). It takes some work to do auto-downloads of US financial institutions, but it is possible. I'd actually advise against it--if you record transactions as you make them (you can do this when watching TV), you can spot bank errors or improper charges easily & then reconcile when you reconcile with your monthly statement. Both MS Money and Quicken encourage this too, but few people take advantage of it.
Taxes are trickier. There is F/OSS tax software, but it just makes the process of manually doing your taxes a little easier. I'd suggest cruising over to IRS's taxfreedom site--there are scores of online tax preparation utilities which have interfaces similar to TurboTax. The IRS can direct most people to one that won't cost them anything. E-filing is fast, easy & gets your refund quicker. You can actually track your accounts online with an aggregator such as yodlee, but I find Gnucash or others to be so handy when making customized reports.
I didn't post screenshots because I thought we were arguing file formats, rather than applications. But perhaps you stopped arguing file formats when you realized that TeX was certainly powerful enough to be the backend for documents. Similarly, you stopped arguing about (X)HTML when you realized that there were apps which people would almost universally agree were both easy and versatile (though I didn't even mention the spreadsheets & presentation software available for the format or that javascript allows most of the scriptability and dynamic content that you crave). I feel you have no arguments against the file formats because you don't understand the file formats.
But, if it makes you feel better, here's what my peers used (most were mentioned in my post, but I hadn't linked them for you & apparently you can't use a search engine):
None of these apps are Motif based (which allows native widgets anyway). LyX used to be, but now uses XForms and Qt, with a GTK interface in development). Easiest/best looking apps are probably SWP and LyX, depending on your platform.