apps choice for Ubuntu and installation choices [Was: cross-platform virus]

[Alan McKinnon]

On Wednesday 12 April 2006 13:41, Alexander Jacob Tsykin wrote:
> On Wednesday 12 April 2006 20:33, Alan McKinnon wrote:
> > So we come back to my original statement. What are these apps
> > that these people do and don't need? Please provide a list. And
> > when you do, please make sure it is a researched list that
> > accurately reflects the needs and wishes of the actual user base,
> > as opposed to your own personal opinion.
>
> I will do that, but here are a few simple ones to start (if anybody
>  disagrees, please let me know):

I don't think you are getting what I'm saying. I asked for a 
*researched* list and I have very good reasons for asking that. By 
researched I mean statistically analysed answers from a 
representative large group of users. As a wild guess I'd say this 
would be 10,000 of them minimum but intuition tells me 100,000 is 
more like it.

> A home user most definitely does not need the Gimp. It is overkill,
> it has far too many features, is hard to learn, and difficult to
> use. 

No, you think those things about The Gimp. You might not be 
representative.

> Gnucash would be useful to almost any body who owns their own 
> business 

I happen to agree with you here, household budgets would benefit too. 
But we'd have to ask the devs why it was excluded.

> No multimedia apps need to be included in a business 
> install because chances are the user will need word processing,
> email, and web browsing only. 

That's a huge assumption you are making. How do you know that's what 
business users need? Did you ask them?

> A tool to convert measurements is 
> useful for almost anybody. 

OK

> A desktop choice for kids would be good. 
> It could include mainly games, some educational and some not.

Isn't there a spin off distro that specifically targets that market?

> Bittorrent most definitely does not need to be included by default.
> Only users with a fairly good understanding of computers have even
> heard of it. 

Says who? What about projects that use torrents to distribute data?
Is this statements based on any real observation at all, or are you 
just quoting your own personal use-case and assuming that it 
therefore applies to all?

> Gnash should be included in the base install, so that 
> people aren't forced to sue Macromedia's proprietary plugin, and it
> works for 64-bit too. 

That's a flash replacement right? I avoid flash where possible but 
AFAIK there's a problem with the flash format - it's free to write 
but not to read. Correct me if I'm wrong, but that looks like a 
similar category of thing to mp3 and dvd encryption - legal issues

> At the moment it is not even installable. 
> The various printer drivers should not be installed by default. A
> user is very unlikely to use more than 2, and yet he has all of
> them. It is unnecessary.

User upgrades printer. Now what?
A company leases Nashua printers. At the end of the contract they get 
a better deal from another supplier using Canons. No user can print. 
Now what?
Will you volunteer to be the one to hand-upgrade ever desktop in the 
company to install new printer drivers?

> These are just a few which rapidly come to mind. Please add more or
> correct me as you see fit. I think this is a good idea, so I would
> love to discuss it. 

But you are making a classic error, of assuming that your own 
viewpoint represents something other than you. And you are doing it 
in many threads lately. This isn't a flame, but you need to get this 
point:

You cannot assume that what works for you is best for all. You cannot 
assume that because you saw say three people have the same problem 
that all users will have that problem (all three could have the same 
misinformation from the same mistaken source).

A sure fire way to kill any good project is to make assumptions about 
what the users want from it. There is only one way to answer this 
question, and that is by doing a real survey of the users, and making 
sure your sample size is representative of the user base. You have to 
take steps to avoid false positives and negatives, and the human 
tendency to not mention things that are not problematic.

Your idea isn't bad in principle, but you have absolutely no data to 
work with currently. In fact you have tainted data - your own 
opinions, biases and preferences. And you can't back any of them up 
with facts.

> As for who would choose apps, the devs would, 
> who else?

Wrong. Ideally, the devs package whichever apps that real research 
indicates will be best suited and/or received by the users. This 
probably wasn't realistic for Warty (due to cost and unfamiliarity) 
so we do the usual bootstrap thing instead - try to make an informed 
prediction and hope you get it right. If you succeed somewhat, tweak 
as necessary, but go and get real data as quick as you can.

-- 
If only you and dead people understand hex, 
how many people understand hex?

Alan McKinnon
alan at linuxholdings dot co dot za
+27 82, double three seven, one nine three five