home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.ms-windows.apps
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!The-Star.honeywell.com!umn.edu!lynx!news
- From: (Tim Keitt)
- Subject: re: Seeking Opinions on Mathcad
- Message-ID: <rnnqh_g@lynx.unm.edu>
- Date: Fri, 20 Nov 92 18:22:32 GMT
- Organization: University of New Mexico, Albuquerque NM
- References: <BxzC5L.Bqq@uxa.ecn.bgu.edu>
- Reply-To: tkeitt@algodones.unm.edu
- Lines: 78
-
- In article <BxzC5L.Bqq@uxa.ecn.bgu.edu> mfprf@uxa.ecn.bgu.edu (Paul R. Falzer) writes:
- >I just received an offer from the company to purchase the Mathcad
- >program for about $80 and would appreciate your assistance in
- >deciding whether I should buy it at that price.
-
- That's interesting because MathSoft wants $99 for an upgrade
- from 3.0 to 3.1! I would press them a bit about their upgrade
- policy before you commit.
-
- >
- >My needs are for an equation editor that is more powerful and
- >flexible than AmiPro's and for a statistical analysis tool that
- >would give me the flexibility to write linear models and analyze
- >data accordingly. My first question is, how well would Mathcad
- >meet my needs relative to other programs on themarket--especially
- >in comparison with other stat packages such as SPSS?
-
- MathCAD is not an equation editor. I don't think it will
- produce the output quality you want. My guess is that MathCAD
- can't touch SPSS for statistics (I have only used 3.0, not
- 3.1). You can extend it however with add on packages from
- MathSoft (which tend to be overpriced as well). Also be aware
- that, at least in 3.0 there are some pretty severe limits on
- the size of arrays (about 8k numbers) which makes some Monte
- Carlo techniques imposible.
-
- >
- >Second, Mathcad seems to resemble an electronic chalkboard.
- >Besides its working with mathematical equations and models, what
- >(if any) are its other capabilities? For instance, does it have
- >an drawing, charting, or graphing capabilities? How effectively
- >does it perform word processing and layout functions?
-
- It does produce resonable plots, but I wouldn't rely on it for
- anything very sophisticated.
-
- >
- >Third, what else should I know about the program's overall and
- >specific strengths and weaknesses? I am wondering about the
- >following items, among others: how fast is the program compared
- >with other significant applications? How well does it work with
- >other programs? How effectively does it function as an OLE
- >server? What's its memory and system resource demands?
-
- The visual interface in version 3.0 is increadibly slow! You
- will often find the cursor 3-4 steps behind your key inputs,
- which makes editing a pain. I would try a demo of 3.1. The
- numerical part is reasonably fast. v3.0 doesn't support much in
- the way of interprocess communication (DDE, etc) but I haven't
- really explored this. It'll run fine on a 4MB windows box (you
- will *want* a math co-processor).
-
- >
- >Thanks in advance for your comments and opinions.
- >
- > Paul R. Falzer
- > Internet: mfprf@uxa.ecn.bgu.edu
- > Bitnet: mfprf@ecnuxa.bitnet
-
- Taken in parts MathCAD is pretty kludgy in version 3.0. This is
- true however of many first generation windows programs. Perhaps
- they've patched it up in 3.1. Taken as a whole however, its
- really a pretty slick program. For simple modelling and
- equation solving, it can be a lot faster than programming (if
- you've used it for a while). It doesn't support loops, which
- means you have to store results on an iteration in an array.
- This often isn't the most natural way to do things. But the
- visual clipboard metaphore is really nice sometimes. I guess I
- would say that the bottom line is that MathCAD is great for
- certain kinds of applications, but I find when the going gets
- tough, I use other tools (IMSL/IDL and C on our UNIX systems).
-
- --
- Timothy Keitt
- tkeitt@algodones.unm.edu
-
-
-
-