Investment of time and money is at a premium in this economic climate and wasting effort and budgets on unnecessary acclimatisation to less efficient software does not rank highly in the list of priorities any organisation will have. We will be reinstalling 2012a and hoping that at some point a reversion to the “classic” form will be available in future updates (but I’m not holding my breath on that…).

I note that the annual maintenance fees have increased – to help cover the cost of this debacle?

As said by others, there seems to have been a lack of consideration of what many scientific/technical users need, a straight-forward, functional, efficient system – of which continuity plays a substantial role. The flim-flam of ribbons is rather like some of the fashion trends of yesteryear, thought of as the pinnacle of style at the time, but merely an embarrassment with hindsight.

I hope there is a reversion back to the classical style that provides effective functionality for users rather than pandering to misguided trends of software purveyors. The development should be aligned with adding more algorithms and true functionality, not providing unwarranted interface changes.

The ribbon interface might be appropriate for people who are totally new to a command-prompt style interface… but for the rest of us, who structure our code and know what functions we’ve built and don’t need shortcuts to silly things like “comment this line”, the ribbon is totally pointless.

After a solid month of trying to like this new system, I’m jumping ship. The UI is sluggish, memory-hungry, and poorly laid-out. Mathworks, I challenge you to demonstrate that ACTUAL product testing was performed for this “feature”. If by “passed muster within our development organization” you mean “we asked around the office and people didn’t want to disagree with the bosses”, that doesn’t count.

To echo some good suggestions from above for future upgrades:
1. Line numbers for errors in executed code blocks,
2. Break points in code blocks,
3. A decent GUI generator,
4. The ability to view and rotate large 3D point clouds without MATLAB dying,
5. Faster start-up time.

What moron could have possibly thought this was a GUI “improvement”? Its just absolutely ghastly and inexcusable! And believe it or not, I’m a pretty open minded guy (a Mac user for example).

But most importantly, it PROVES what I have been claiming now for a couple of years, that the Mathworks has lost their way, and that Microsoft C++ programers have taken over the Matlab product line and apparently also the company! These C++ programmers have no more idea than a goat what a REAL engineer needs for doing their work, and just like the fabled rhyme that “to one with a hammer, everything is a nail”, now too “to a bored C++ Programmer everything is an Object… and now also a Ribbon!”.

Mathworks, you need to take your company BACK from the C++ programmer minions and get control back in the hands of engineers! REAL engineers! Fire these people because they obviously have nothing productive to do or contribute. Get some real engineers and mathematicians back in there and let THEM program Matlab, and not these perpetrators of malice, inefficiency and confusion.

I have no other way of expressing my gross displeasure with the shear stupidity of Mathworks direction than to simply not renew my maintenance license. Maybe money will talk. I will stay with 2012a, and cherish it as long as possible as the last real Matlab.

Cleve, please help! Your company has been taken over by C++ zombies!

1. I don’t like the new interface. It was better to develop new core numerical/math/compsci functionality. However this is not that big problem for me. I use emacs interface to matlab and as long as this works, I am OK.

2. I am worried by the fact that the new version does not support 32 bit linux. This worry started some time ago when Solaris disapeared from the list of supported OS, but now things got really acute. This is an indicator for bad things and sloppy coding practices in Mathworks. Well written code should compile on any UNIX quite easy in 32/64 bit. The fact that decision was made to remove 32 bit linux indicates that it was too expensive to maintain the code for 32/64 bit. And this is an indicator for bad code.

While I understand that most Linux users are big server clients and they have migrated to 64 bit, I am quite sure there are quite a few users, who would prefer to have 32 bit linux version available.

In short – Mathworks starts to behave not like engineering company, but like a company run by MBA-s…. I am too old to swich the product, but I will advise any students reading this to consider and invest effort to learn alternatives to Matlab. There are alternatives.

Looks like I will either have to downgrade or go back to ‘matlab -nodesktop -nosplash’ again. Shame.