The Old Joel on Software Forum: Part 3 (of 5) - The world from the idiot savants point of view

The world from the idiot savants point of view

"Frustrated managers" response to the savants thread can be a little instructive to senior developer types,

Sometimes a decision in technology is hard to come around, In a lot of smaller development companies ".NET rules, or Java is soooo kewl" decisions aren't driven by even upper management, they have to do with which major OS or application vendor is bundling your software today (sun,HP,MSFT,Siebel, Peoplesoft ....)

But in most cases these decisions have nothing to with the real world. When "Frustrated manager" says we need our senior to embrace the advances of the last 10 years which advancements does he mean? What are the specific ones will the product take advantage of? Why do you have to retool every 3-4 years? Isn't it better to concentrate on improving what already exists, with familiar tools? Also, when you say the senior programmers don't want to learn, are you sure they don't want to change to a new language, or just dont see any reason to do so.

So, "frustrated manager guy", please understand that while to you a gleaming new system maybe a shiny medal on your resume. To a developer it seems like a waste of time and company resources, to reimplement a working system in the language de jour

So I think, if youre a senior developer watch out, recommending not moving to suns or m$fts or (I guess its IBM's turn) latest hype darling can make you look out of touch. So youre new mantra is ... should be I love .NET because its even more Object Oriented and full of NTierry goodness

What about others on this forum, what percentage of technology projects have you worked on:

because there were good technological or marketing or interoperability reasons.

vs.

Projects that were created because a technical journal fell into the wrong hands

The Artist Formerly Known as Prince
Tuesday, January 27, 2004

I have to agree with TAFKAP.

We are currently rewriting our core application in C#/WinForms/WebServices etc from C++/VB/ADO. Prior to this ~8 man years were spent in a failed attempt to write the application as ASP.Net. The main reason given for these attempts are because 'our UI is stale' and 'to create an n-tier solution'. While I agree with both of these statements, I believe that we could have updated the VB UIs fairly easily and broken our current C++ business objects into client and server objects. Needless to say, when I expressed this opinion to management, I was not very popular. I'm sure that they were thinking "Billy does not want to learn new things..." instead of "Billy sees the investment we currently have in our intellectual property"

Billy Boy
Tuesday, January 27, 2004

"I love .NET because its even more Object Oriented and full of NTierry goodness"

I love that line ... <grin>

Canuck
Tuesday, January 27, 2004

"Projects that were created because a technical journal fell into the wrong hands"

This well-known phenomenon is probably the largest reason why the IT industry is the largest fad-driven market (aside from, say, fashion ;)) in the US (if not the world).

People rarely see more than 10 feet from their noses and think XYZ is the latest greatest thing when it poorly and *partially* solves problems people solved long, long ago.

Even worse, they don’t fully understand the problem *nor* the proposed “solutions” so they are unable to recognize their folly. It’s a poor state of affairs.

MR
Tuesday, January 27, 2004

we have a name for those people "BOSS"

The Artist Formerly Known as Prince
Tuesday, January 27, 2004

I thought everyone knew the term "magazine manager"

"What color do you want that database?"

Philo

Philo
Tuesday, January 27, 2004

I agree with the original post

But doesn't this kind of pressure also sometimes come from the other side too? I mean don't you get programmers who want to rewrite everything in the latest whizz bang language or tool, because it's more objecty or whatever?

S. Tanna
Tuesday, January 27, 2004

S Tanna

Perhaps some developers would rewrite because it is more 'objecty' but most that I have met would not want to do basically the same thing over again. Where's the challenge in that?

Billy Boy
Tuesday, January 27, 2004

S.Tanna - yeah, we have a problem like that. There's a club of programmers in our little corner of the world that are pushing lots of new tech with hype.

[The problem is, they're not training anybody. They're just pushing tools and expect the rest of us to learn in our spare time. Of course, the haven't realised fact that their position - their salary - is to investigate new tech. It's a little bit easier when you have nothing else to do but research. They should go and be academics or something.]

Shodan
Tuesday, January 27, 2004

Billy

> Perhaps some developers would rewrite because it is more 'objecty' but most that I have met would not want to do basically the same thing over again. Where's the challenge in that?

Think of it this way:

(a) Most decent size projects have a long history - so the current developers may not be the ones who worked on it before

(b) Most decent size projects are done by multiple people - so current developers want to rewrite may not be talking about their own stuff

(c) Plenty of developers are perfectionists (my wife accuses me of this regularly on *everything*, so I am no innocent), so even their own work - they think or know that they could do better know

(d) Most developers underestimate the amount of work in any project, including completed projects, -- including rewrites -- and especially projects that were done all or part by somebody else.

(e) Double point (d) if the existing stuff was done using "yesterday's technology"

(f) Most developers like working on new projects -- a blank sheet of paper -- rather than incremental improvements to existing projects where you have to fit into an existing infrastructure.

(g) Most developers like working on "today's technology" not "yesterday's technology". Double that when the job market is up.

I'm not saying it ain't ever worth a rewrite with today's stuff, but, I think it's worth it a lot less often than many developers end up arguing for.

S. Tanna
Tuesday, January 27, 2004

"what color do you want that database"?

The best engineering mgr I ever had asked that question
at an offsite, where visioneering "mgmt team" types were
having a bullshit (er, strategy) session. They said "we
need a scheduler in our product, with a release in three
months". He replied, "what color?" They were incredulous;
and even started talking about colors. He excused himself
and me; we ran into the parking lot and laughed ourselves
silly. We left the company two months later...

x
Wednesday, January 28, 2004

mauve philo, its got to be mauve

the artist formerly known as prince
Wednesday, January 28, 2004

But I heard crimson has more RAM.

Joel Spolsky (Fog Creek Software)
Wednesday, January 28, 2004

I was recently asked, "are we using XML?"

John Ridout
Wednesday, January 28, 2004

Funny how we never hear we never hear "are we using ASCII?" which is essentially the identical question, but for the fact that there aren't many magazine article titles touting ASCII these days.

veal
Wednesday, January 28, 2004

So is the correct answer "We are using an ASCII approximation of XML" for straight text and "We are using a byte-wise approximation of XML" for binary? Management wouldn't know the difference, would they?

Billy Boy
Wednesday, January 28, 2004

"what color do you want that database"?

System colors always seem best.

Stephen Jones
Saturday, January 31, 2004