NT, XP, or Win2000, etc. for power users

[codyswan]

I'm going to get a new PC soon, stocked with at least 1 gig of ram and 100 gig on the hard drive. I use a lot of spreadsheets in the 10-100 meg range. Sometimes I would like to have several open at once, though that has made my current PC run slowly.

Will it make any difference which operating system, version of Excel, etc. is loaded on it? If so, which do you recommend?

 

[AJ]

I can't really back this up with hard figures but just from experience in my company and from talking with other geeks(!), Win2K is the best of the bunch from a power user point of view.
I wouldn't touch XP with a barge pole until they've upped it with a few more service packs. Win2K is pretty robust now and is a noticeable improvement on NT for personal use (don't forget to think about gaming and such-like that your bound to want to use your PC for!).
As to the version of Excel, there's nowt wrong with Excel2000 that I've found yet, although I bow to the experience of the rest of the board if they think different!

Finally, as a adjunct to this, if your spreadsheets are in the 10-100MB range, your much probably better off using a database instead.
If you've got a spreadsheet this big I'll bet that a great deal of it is just plain old rows and columns of data. If Excel is your preferred tool to work with then stick your data part in an Access/SQL/etc. database and you can still use Excel to query it, either by MSQuery or with the xlodbc.xla addin (all standard facets of the Excel install)

Rgds
AJ

 

[codyswan]

AJ thanks for your help. I gave up gaming after a shoulder problem developed years ago from aggressive use of joysticks.

Win2000 would be great, as it is what I have installed on my current PC, which I will keep and use as well.

About half of the size of the spreadsheets are pure data. The rest consists of various manipulations and formulas. I've tried various alternatives and gotten it to run somewhat faster.

I think I may have to go with SQL. My primary reason for using Excel has been that names are used inconsistently in the database (e.g., ABC corp; ABC, Inc.; ABC Inc; ABC&DEF Inc.). Old premerger names show up a lot too. If it wasn't for the massive data cleanup required, I would just use SAS because the manipulations are statistics and curve fitting.

Inconsistent names also created a problem with pivot tables. The names vary from one quarter to another and it is a terrible task to match everything up in the pivot table and make hundreds of new merged categories. So far, I am still using dsums because they tolerate large differences in spelling using wildcards in the criteria.

 

[shades]

Database sounds like the way to go.

I have a similar situation with about 300 company names, and various incarnations with quartlery reports involved 10,000 data points. However, going to MS Access has actually solved many of those problems. Rather than change them, I just add one Access query ahead of all others to determine combined status. Then after all the data manipulation, I export to Excel for further refinement and preparation for sending out (to several hundred people who only have Excel).