Fog Creek Software
Discussion Board




Will 75% ever be reached ?

Hi Joel,

Do you believe that your 75% .NET uptake criteria will ever be satisfied ?

If we assume that it will take a long time, then versions 1.0 and 1.1 will certainly be out of date and replaced by version 2.0. Perhaps even later versions after that.

It would be nice to think that you could use VS.NET 2005 to target any version of .NET (1.0, 1.1, 2.0, etc), but this is likely to be difficult.

In a way, this is similar to designing web sites/applications for various browser versions. Do you target IE6 and accept that IE4 users will not work, or do you target IE4 and leave out any potential gains to be had by using a later browser.

Don't get me wrong, I agree with your position, I just wonder if we will ever be able to ship .NET applications, apart from when we can control the desktop.

Steve Jones (UK)
Monday, March 29, 2004

Good question. It's not looking likely, because the rate of adoption seems to be around 1% a year, and every year and a half there's a new version of the runtime.

I'm starting to think we're more likely to see mass adoption of broadband before we see mass adoption of the .NET framework, and the distribution problem will be solved that way :/

Joel Spolsky
Fog Creek Software
Monday, March 29, 2004

Where does that 1% per year number come from?  We're logging browser info from visits to our web site and we're seeing that around 30% have a .NET verison installed.

Walt
Monday, March 29, 2004

You got weird users. I get low single digits from the main Joel on Software site.

Joel Spolsky
Fog Creek Software
Monday, March 29, 2004

I have to admit we don't have normal users.  Our software is oriented to market researchers, so we end up with more corporate installation scenarios.

But those are the numbers we're seeing.  We used that info to help us determine our .NET position.

In January we rolled out a new version of an app that used a new .NET gui with the rest in C++.  The UI isn't very complicated, so if we got many complaints, we could back away from it.

So far, we've had no complaints.  Not even from download distribution.

But again, our users are not normal.

Walt
Monday, March 29, 2004

30% is consistent with the results I've gotten from our logs.  Doing a quick check now of the past two weeks, I'm seeing 17% with .NET 1.0 and 34% with .NET 1.1. 

We do about 50% business and 50% consumer sales and I'd guess our customer base is a decent random sampling of the general software buying population. 

SomeBody
Monday, March 29, 2004

Yeah, Joel.  I have to say my numbers don't jive with yours either:  http://www.quickfixengine.org/statshistory/BROWSERREP.html

I have a very different user base.  Mostly large financial institutions, which likely has a lot to do with it.  Despite the large percentage of installed CLI runtimes, of the three API's we provide in the product (C++, java, .NET), the .NET one is by far the least used.

Oren Miller
Monday, March 29, 2004

Joel, what numbers do you get on the JoS discussion forum home page?

Philo

Philo
Monday, March 29, 2004

I would've thought Joel would have higher numbers than anybody, since he attracts a lot of visits from developers.

mwv
Tuesday, March 30, 2004

I use Firefox as my primary browser.  Does it report the fact that I also have.NET 1.1 installed.  Perhaps a lot of JoS readers also do not use IE which could be skewing his numbers?

Paul Lefebvre
Tuesday, March 30, 2004

We see 25% with the CLR installed. Our software is consumer focused.

Jack Reeves
Tuesday, March 30, 2004

Joel, is that 1% gathered from people using IE (Windows)? I would suspect your number is diluted by non-windows visitors.

In which case, you don't need to worry about it, 'cause those folks don't like to pay for software anyway.

<lights match>

Rick
Tuesday, March 30, 2004

"I would've thought Joel would have higher numbers than anybody, since he attracts a lot of visits from developers."

Developers are more likely to be using a non-IE browser than the general population.

NoName
Tuesday, March 30, 2004

We're at the botto part of an S curve. Wasn't the adoption of IE about this slow in the beginning? And then you woke up one day and it was 90% of the browser market?

Dan Maas
Tuesday, March 30, 2004

I was going to say the same as Paul. The people who are likely to have .Net installed are unlikely to be using IE.

Stephen Jones
Tuesday, March 30, 2004

It looks to me as if some people are reading "1%" and interpreting it as "1% of people have .NET" when what Joel actually said is that the percentage of people with .NET is going up by about 1 (%) per year.

Gareth McCaughan
Wednesday, March 31, 2004

"I'm starting to think we're more likely to see mass adoption of broadband..."

It's definitely happening now.  We're seeing $29.95 for DSL in some places, so unless you're really strapped for cash it's almost a no brainer for most dial up subscribers.

Cable is a few bucks more but higher bandwith.

Jim Rankin
Wednesday, March 31, 2004

"It looks to me as if some people are reading "1%" and interpreting it as "1% of people have .NET" when what Joel actually said is that the percentage of people with .NET is going up by about 1 (%) per year."

So .Net has been out for 25 years?

(25% or more is what posters before you are claiming to see.)

Jim Rankin
Wednesday, March 31, 2004

Gareth: Joel didn't claim seeing 1% adoption as of now but he did claim a "low single-digit" percentage so the gap to the 25-50% the rest are seeing is still pretty wide.

Chris Nahr
Thursday, April 01, 2004

Jim, price has never been an obstacle with respect to broadband - ADSL has been available at $40/month for years.

*If* you can get it.

There are two problems:
1) DSL has a limited range and must be over a direct copper line. More than 3 miles from the phone company and you're SOL on DSL.
2) More importantly, in suburban areas the telcos adopted a distributed star topology - trunks fan out from the Central Office to substations called Digital Loop Carriers (DLCs), then smaller trunks fan out from there. DSL can't go from a CO to a home through a DLC.
When the FCC forced the telcos to open Central Offices to third-party DSL providers (ca. 1998) you saw the Covad & Co. explosion. However, due to a lack of foresight on the part of the FCC and tarriff drafters, the telcos weren't forced to open their DLC's.
DSL can be implemented at a DLC, but since you're serving a smaller population from a smaller building with a lower-scale card, the profit margin is lower. Ergo, the telco's have zero interest in bothering with it.
So - telco's don't feel like provisioning DLC's, and don't have to let third parties into the DLC's to provision them, so if there's a DLC between you and the central office, odds are you won't be seeing DSL any time soon.

Meanwhile, my cable company seems to be intent on proving why monopolies can in fact be quite evil. ($50/month for broadband with VPN filters. $90/month to get 5 DHCP'd IP's and have the VPN filter removed, $190/month to get real static IP's)

Philo

Philo
Thursday, April 01, 2004

> why monopolies can in fact be quite evil

"can in fact"  -> "are usually, if they can get away with it"

Portabella
Saturday, April 03, 2004

Just another data point. We see about 30-35% penetration on our public site. Our users are by and large non-technical marketing users.

Brad Wilson (dotnetguy.techieswithcats.com)
Sunday, April 04, 2004

*  Recent Topics

*  Fog Creek Home