Why aren't more universities using open source?

[Spartrekus]

guys, you should really try Mathematica, it is great piece of software. To get things done !

Maybe a good alternative is to use tinyexpr, to be used maybe with gnuplot, it is more powerful and you will learn C++.

GitHub - codeplea/tinyexpr: tiny recursive descent expression parser, compiler, and evaluation engine for math expressions

tiny recursive descent expression parser, compiler, and evaluation engine for math expressions - codeplea/tinyexpr

github.com

 

[Nicola Mingotti]

Maybe a good alternative is to use tinyexpr, to be used maybe with gnuplot, it is more powerful and you will learn C++.

GitHub - codeplea/tinyexpr: tiny recursive descent expression parser, compiler, and evaluation engine for math expressions

tiny recursive descent expression parser, compiler, and evaluation engine for math expressions - codeplea/tinyexpr

github.com

oh Lord, no, we are far Spartrekus, in Mathamtica you can do things you can't dream of with the tools you are proposing. I have seen physicists working on Mathematica full day, for years ! Except the time to write the papers, for that LaTex is the tool (it is the best and it is open source).

I have worked with it, for a long time. For testing a new idea that has something to do with applied mathematics it is the best. Because it is 'holistic tool', Wolfram has stolen the fundamental idea of Lisp (without giving credit) and built a fanstastic Lisp with a beautiful interface.

I find a computer which still has a license i will write a short demo of what you can do

 

[Spartrekus]

oh Lord, no, we are far Spartrekus, in Mathamtica you can do things you can't dream of with the tools you are proposing. I have seen physicists working on Mathematica full day, for years ! Except the time to write the papers, for that LaTex is the tool (it is the best and it is open source).

I have worked with it, for a long time. For testing a new idea that has something to do with applied mathematics it is the best. Because it is 'holistic tool', Wolfram has stolen the fundamental idea of Lisp (without giving credit) and built a fanstastic Lisp with a beautiful interface.

I find a computer which still has a license i will write a short demo of what you can do

It's true, mathematica is an excellent companion. What about open source alternative for it?
Even the CAS is not from Wolfram.

It depends the level of complexity, when it starts to be really hard, non linear dynamical systems with about >= 50-60 variables, then, mathematica will be useless. Then, C++ comes in!

 

[Nicola Mingotti]

It's true, mathematica is an excellent companion. What about open source alternative for it?
Even the CAS is not from Wolfram.

As of today JupyterLab has a good momentum. I see people using it in my building, it is cool. It is promising. I guess as symbolic calculus capabilites we are still far from Mathematica/Maple. Also, forget the "holistic stuff", here there are separate blocks that do separate things (GUI, the server web, the browser and the computation engine [python or R or Ruby or else] ).

It depends the level of complexity, when it starts to be really hard, non linear dynamical systems with about >= 50-60 variables, then, mathematica will be useless. Then, C++ comes in!

Well, it is true that once you want speed nothing can stand near a good written C or Fotran code. Once you write in C the problem must be definitely well understood. But i must say, for what i see from my collegues, now a big part of hard numerical code is made in Python first, then if it turns out it is not fast enough, go for C. Then it can go for FPGA or GPU for further optimizations.

 

[badbrain]

Err it seemed both Maple and Mathematica were written in Java...

 

[Nicola Mingotti]

Err it seemed both Maple and Mathematica were written in Java...

No, who told you that ? The UI is Java for Maple and C/C++ for Mathematica... And you can well see the difference when you run them. (p.s. my data may be a bit old, it is a few years i don't use each of the sofware .... i should need to check in an old computer with the licenses, if you are really interested on this detail let me know.)

In general Wolfram has spent a lot of effort on the interface, Maple tries to close the gap but has put considerably less resources in that.

If you check wikipedia you can see that both project are older than Java, wikipedia, and the Web.

 

[ShelLuser]

Apologies but I didn't read the entire thread.

My take: they use what works best for them. Just because something is "open source" doesn't automatically make it superior to other products. A few months ago I did a thorough comparison between the latest Libre Office and Microsoft Office 2016, while carefully taking my bias into consideration.

So basically I compared a rather recent LibreOffice build (6.16 iirc, so released this year) with Microsoft Office 2016 which was a lot older; released in 2015. For general purposes Libre Office was just fine. But the very moment you started to demand a little bit more then things got quickly turned around; for a professional user LibreOffice left a lot to desire.

For example: Ever since MS Office 2010 (even older) I can set up text snippets and save them as text blocks which I can then quickly insert in my document. I can even save these blocks as part of a template. I can't begin to tell you how much time this can save... especially in combination with some macros. I used to set up invoices solely using these building blocks, and they got inserted by pressing specific key combinations.

A feature MS Office has supported since 2012 but which LibreOffice still doesn't have in 2019.

When writing feature requests, documentation and even reviews it has become important to be very careful with your sources. As such it's not uncommon to set up a collection of resources which you either quoted, used as an information resource or which influenced you. Fully supported by MS Office 2016 but an unknown feature in LibreOffice.

In Office (Word) the language I use is detected automatically thus allowing me to mix Dutch, English and German in a document while my spell checker will have no issues with either of these languages. Fully automatic. In LibreOffice I have to specifically designate such areas.

If I'm writing about a specific topic and I need some global details about it I can use Wikipedia as an automated source to look up the topic I'm working on. So: I select a keyword, click the wikipedia icon and I get a side window which shows the wikipedia page about that topic, allowing me to look into general aspects.

.. you guessed it.


Now.. don't get me wrong here. I'm pretty sure that it looks as if I'm talking down on LibreOffice, but I'm not. It's an impressive project and most definitely a solid alternative for MS Office. However... I am saying that for powerusers which need to get work done and where time = money there's a really big chance that LibreOffice simply won't cut it.

To put it very blunt and direct: MS Office has expanded and improved their feature set over the years. With the latest 2019 release (which I have on my laptop, but my experience is still limited) you can even translate pieces of text within your editor "just like that". But LibreOffice didn't have any of this. There are improvements on existing features, sure, but there really isn't that much new in comparison to a version from, say, 2 years ago. And with 'new' I'm not referring to features which got introduced and pushed by the devs but which are actually useful for powerplayers.

With all due respect, but when I see a comparison chart between LibreOffice & MS Office (on the LibreOffice website) then I cannot help but grin when I notice that they even highlight the fact that LibreOffice still supports ancient graphic formats such as .PCX whereas MS Office does not. Yes, it's a difference in features but... is it really that much important to include it into a comparison?

And that, in my opinion, also scratches the whole problem at hand. I can imagine that a developer could be proud of the fact that LibreOffice supports a ton of graphical formats, I really do. But... if you stop to think as a developer and approach the situation as a user... then seriously: is it really all that important? Ton's of Office users don't even know what kind of format they're working with because Microsoft is stupid enough to hide file extensions by default on Windows.


Just because it's "open source" doesn't mean it's superior by design. Germany has taught us as much when a city (which name I forgot) decided to move the entire administration to Open Source solutions. They lasted 4 or 5 years and then it went downhill. fast.

Much, of not most, open source software is made by developers, for developers. Most developers have a really hard time imagining what a user might want, and if they can imagine as much then it's also often met with a bit of disdain because... "Just let them press F1 for help and read up!". Wake up call: it doesn't work that way in the commercial business.

... which is exactly the kind of business universities are preparing their students for, is it not?

</rant>

 

[rootbert]

Just because it's "open source" doesn't mean it's superior by design. Germany has taught us as much when a city (which name I forgot) decided to move the entire administration to Open Source solutions. They lasted 4 or 5 years and then it went downhill. fast.

It was Munich. However, the reason for failure was not a technical one. Politics, bad management and lobbyism was the reason they let the project down. To sum up: Microsoft promised the mayor to build a central office and generating lots of jobs so the mayor let down the project. Not all programs were migrated from Windows os to a modern OS independent web applicatiin architecture. Microsoft managers sitting in the expertise group of course suggested to go back to their product. All in all it was a shame because it was so obvious.

The thing about MS Office vs Libreoffice is true. I never propose open source software is better per se, but IMHO being open source is a major win.

To me this is just like the environmental discussion among countries: governments have to commit to protection of the environment although there are negative effects on productivity, but all have to set the same goal. I am sure if governments started implementing open source, especially libreoffice, we would see a boost in features and catching up with commercial offers sooner.

 

[sidetone]

I would like to see Apache Open Office take off. The last time I looked, they relied on Java, and they were working on getting rid of that. Because of licensing, LibreOffice is allowed to take improvements from Apache Open Office, but Apache Open Office can't take improvements from LibreOffice. I'll continue using Libreoffice, until Apache Open Office improves and loses Java as a required dependency.

 

[Deleted member 9563]

It was Munich. However, the reason for failure was not a technical one. Politics, bad management and lobbyism was the reason they let the project down. To sum up: Microsoft promised the mayor to build a central office and generating lots of jobs so the mayor let down the project. Not all programs were migrated from Windows os to a modern OS independent web applicatiin architecture. Microsoft managers sitting in the expertise group of course suggested to go back to their product. All in all it was a shame because it was so obvious.

The whole idea of ethics in business and administration does not appeal to everybody.

Like I said earlier. And I'd add to that by saying that these people have no backbone and are morally reprehensible.

The thing about MS Office vs Libreoffice is true. I never propose open source software is better per se, but IMHO being open source is a major win.

I'd also agree there, particularly in the context that ShelLuser detailed above. That said, he's probably talking about 1% of users. Many others could just as easily use plain text and it would actually be an improvement to the final product.

To me this is just like the environmental discussion among countries: governments have to commit to protection of the environment although there are negative effects on productivity, but all have to set the same goal. I am sure if governments started implementing open source, especially libreoffice, we would see a boost in features and catching up with commercial offers sooner.

Open source is indeed a win, but to me the overreaching thing is not so much whether it is open source, free, or proprietary (people deserve to be paid, and a buisness needs to survive), but rather whether the company producing the software is harming the rest of the software ecosystem or not (or worse). Despicable business practices simply do not fly in my world, and I am very willing to do without in order to not support such things.

 

[sidetone]

It would be cool if more colleges used FreeBSD. The IT department would have to upgrade it every or every other semester.

 

[fernandel]

I am LibreOffice user and for mwhat I need is enough. At work we use Adobe PDF writer but there are many genetics analyzers which run on Linux. Also Python and R are very common in genetics research. Also in microbioligy was antibiogram tester which run on Unix with CDE . I think now is everything on Windows.
Both GIMP, Blender which I know more are IMO very good opensource applications.

 

[Nicola Mingotti]

in the end, it is not really opensource that makes the difference, and less then less the free (as in beer) status. if a software is useful you will buy it.

IMHO, the greatest hurt is made by closed (proprietary) file format. That is, I can’t write a program that works perfectly well with Word files because Words encodes stuff in a way that is not fully disclosed. (afaik)

This i think is a political battle worth the effort: Nothing producing third party undecodable files should ever enter the public administration. Also, should not be thaught in school, because it will undermine freedom.

Freedom of opening the same document tomorow, with another program, and not be doomed a priori to get a suboptimal result.

 

[Spartrekus]

Hej, guys, who cares about MS Office? Because ...
There is Tex and Latex, it is far much better, it compiles after 30 years, and it is also stable ; it is as old as BSD. To say, that you can rely on it.

Err it seemed both Maple and Mathematica were written in Java...

"were"...?

younger version of maple were not on java, it was cool and fast.

 

[Spartrekus]

As of today JupyterLab has a good momentum. I see people using it in my building, it is cool. It is promising. I guess as symbolic calculus capabilites we are still far from Mathematica/Maple. Also, forget the "holistic stuff", here there are separate blocks that do separate things (GUI, the server web, the browser and the computation engine [python or R or Ruby or else] ).


Well, it is true that once you want speed nothing can stand near a good written C or Fotran code. Once you write in C the problem must be definitely well understood. But i must say, for what i see from my collegues, now a big part of hard numerical code is made in Python first, then if it turns out it is not fast enough, go for C. Then it can go for FPGA or GPU for further optimizations.

I don't know what is so exciting to run supercomputer simulations on python. Famous universities will get funding for that. It seems that they discover python, but it is as you said slow, very slow compared to C++.

If the code is released to public. It is fine. However, US or EU funding will give only publications, without any source code. Basically, you can then buy the funded close-source software. Fundamental research, funded by taxes, can then contribute to close source software, no source code, and ... again business.

 

[Lamia]

Seriously, if universities cannot be in control of their data, how else can they embrace FOSS? We all know that the !FOSS lobbyists are the friends and allies of the executive management of these universities and are more willing to continuously provide funding for them as long as they continue to use their systems and software. A very large percentage of the higher institutions use Office360/SAP/etc needless to same corporations too. We live in a world were 98% of the people online use Uncle Sam services - email, phone, etc. We are all online!

 

[Spartrekus]

Seriously, if universities cannot be in control of their data, how else can they embrace FOSS? We all know that the !FOSS lobbyists are the friends and allies of the executive management of these universities and are more willing to continuously provide funding for them as long as they continue to use their systems and software. A very large percentage of the higher institutions use Office360/SAP/etc needless to same corporations too. We live in a world were 98% of the people online use Uncle Sam services - email, phone, etc. We are all online!

Office360 is at all univs yeah, indeed.
tex could be the only allowed format from universities for thesis or books. many scientific journals allow tex and some of them recommend it.

 

[Lamia]

Office360 is at all univs yeah, indeed.
tex could be the only allowed format from universities for thesis or books. many scientific journals allow tex and some of them recommend it.

This argument is not unrelated to the confiscation of scientific intellectual property in Big Brother's DMZ. That is why the BB and her allies are fighting tooth and nail to quench the blazing fire from Scihub and related efforts - libgen, etc.