Livid.

[latemodel]

The short version is that I've just lost about six hours of work because Apple COMPLETELY BROKE one of the programs I really, really need for school.

Comments

[pir]

Eh, they have their oddities and they're a bit annoying sometimes if you want to do things in the way Steve didn't intend but certainly with all the UNIX level stuff if you don't like it you can just go off and either fix it or replace it. I certainly wouldn't go as far as "painful".

I have a Mac because it's a supported UNIX box and I can do everything I need to do on it. Not always in the way it was intended but it works fine.

[perligata]

Do python libraries not come with the ability to change the default installation directory?

[latemodel]

The installers for the compiled binaries don't have that option, no.

[halo]

this might be a stupid question, but what if you delete the old version first?

[latemodel]

Oh, that would work fine. It's not that it's a hard problem to fix once you've found what's wrong, it's that it's hard to find the problem because it breaks silently. Or rather, breaks by giving you error messages that don't get at the problem.

[halo]

ah, yeah. It would certainly not occur to me to manually remove something before installing a new version.

As a nearly lifelong mac user, I think there are some very lovely innovations in the newer systems, but they are getting so complicated in some ways that the simplicity that made macs great is falling by the wayside, sometimes. Things on macs used to be easy to fix because you could find every system setting in very intuitive, predictable places. It took me longer than I'd care to admit recently to figure out how to change my Airport network password, which really ought to be so straightforward, and isn't. (In general, I think the control software for Airport is appallingly badly designed.)

[atalanta]

Ugh, good luck.
I know Mathematica's now trying to sell itself as "we can do everything MATLAB can do and more" but I haven't explored much yet ...

[latemodel]

The reason I started using Python was really SQLAlchemy. Matlab's database toolkit is a farce (they want $$ to do something you can get a free C library for), and they aren't even pretending to do a real ORM. I wanted an easy ORM, so it was either Python or Ruby; and between the general consensus behind python in the scientific community, and SQLA's configurability, it won out.

A lot of folks think NumPy + SciPy + MatPlotLib + iPython makes a good Matlab-killer. For the price it certainly can't be beat, the graphics are easier to manipulate, a real OOP framework is great, and you can use real coding tools like an IDE and source control. But things like matrix math and basic plotting don't work quite as fluidly, and Matlab's debugger is still phenomenal.

[structurefall]

as somebody who literally started using a mac yesterday: wouldn't darwin ports solve this problem? i had an issue with 10.5.6 inexplicably running a massively outdated version of curl earlier today, so i just did an /opt/local/bin/port upgrade curl and it fixed it.

[scullin]

"Solve" is a strong word, but yeah, macports will allow you to install newer versions of many things that OSX ships with. Just make sure you keep /opt/local/bin in your path before the usual system paths. macports is also really fussy and is prone to forgetting dependencies, though, and I have spent a few afternoons uninstalling and reinstalling random bits to get something to install.

[netik]

macports does that by default, adding /opt/local/bin to your paths.

Generally, it's not recommended to use the pre-installed versions of ruby, python, or perl on mac.

[latemodel]

Generally, it's not recommended to use the pre-installed versions of ruby, python, or perl on mac.

Man, I wish somebody had told me this. The only things I saw about having to upgrade the preinstalled stuff was to get IDLE to work, and I hate IDLE.