Why software is so bad ... Very interesting!

Dobero · 06-19-2002, 07:58 AM

Thanks to http://www.xp-erience.org/

A complete and very interesting analysis about the present situation in the software making community. It's a too long article because it's complete :-) so check it at http://www.msnbc.com/news/768401.asp?0si=-&cp1=1
On the same subject there's an interesting white paper (thanks to Paolo Attivissimo @ http://attivissimo.net/) "Open Source Software May Offer Target for Terrorists", so "interesting" and rumorouse that they've made a purchased version of it at http://www.adti.net/html_files/defen...5_30_2002.html

But ;-)

being part of a cool fellow cow's community you've been selected to read it for free, so don't disappointe them and read it!

http://x.emdx.org/Slashdot/old_opens...whitepaper.pdf

As you may expected, this institution ( Alexis de Tocqueville Institution - http://www.adti.net/) is sponsored by Micro$oft (http://www.wired.com/news/business/0,1367,52973,00.html)

So, reflect and add comments.
See yall.

Doobie · 06-19-2002, 08:16 PM

Most programmers, like most people, are simply idiots. Sometimes, it's not the programmer's fault but their anti-competitive boss (e.g. Gates) who instead of keeping the OS simple as it should be wants to bloat it out with features to hurt other software companies (the integration of IE is a prime example). And, MS hiding APIs doesn't help the quality of software either.

Software is much more complex than cars. Indeed, the only thing that really compares are microprocessor chips. Chips aren't as complex as their transistor count might seem to indicate. Chips are also designed by people who are more professional than most programmers and much more effort is made to debug chips than software. Still chips have bugs and programmers must work around those bugs. And, there was a big effort a few years ago to reduce the complexity of chips (remember RISC).

Anyway, look how popular the auto repair industry is. There are only two kinds of baby toys and equipment, those that have been recalled and those that will be recalled (okay, that's a bit cynical). Most of the veggies imported into the US have to be tossed even before they make it to the store. That last salesman who tried to get info from probably wasn't right even half the time. Look how buggy our government's foreign policy is.

igalan · 06-20-2002, 04:16 AM

Quote:

Originally posted by Doobie
Most programmers, like most people, are simply idiots. Sometimes, it's not the programmer's fault but their anti-competitive boss (e.g. Gates) who instead of keeping the OS simple as it should be wants to bloat it out with features to hurt other software companies (the integration of IE is a prime example).[...]

I agree with you 100%. Also, very often programmers must finish a product within a time limit, whether the software is properly tested or not. Not only this, but also bad analisys or even no analisys at all before programming; bad planning or no planning at all. This is no way to work. Good programming practices include an extensive analisys which sometimes lasts for as long as programming. Many people think that analisys is a waste of time because no code is programmed so no progress is made. Big mistake!

06-23-2002, 04:20 PM

The reason we don't own cars that serve us for a lifetime is the car industry wants to have something to do tomorrow as well. Benefit is that cars come relatively cheap, otherwise they would be much more expensive. No I can own 10 cars in my life, and have a recent model with all its features. And millions of people are employed in the car repair industry worldwide, and they account for quite some purchasing power.

When a TV tube is built it is sucked vacuum, filled with an inert gas to prevent the phosphor from 'burning up'. Not a lot of people know on top of the inert gas a predefined amount of pure oxygen is put in as well. Why ? To make sure the tube won't last much longer than the intended lifespan of say 5 to 8 years.
Is that bad practice ? It may seem unfair to intentionally make a product worse than possible, especially when this actually costs money to do. Again, there is an upside: tv's come cheap, and I bet they wouldn't when you'd buy one in for a lifetime, for not production cost sets the price, but company turnover and profit margin.

I find it very hard to judge in the above cases. You ?
Software has many similarities I think. It is possible to avoid all the nasty things mentioned in the article, and start with a proper foundation, a plan and structural design. Of course M$ arguably should have abstained from building its OSes dependent on diverse functionally separate modules (like Explorer etc). But would there be a M$ if they had done all that ?

On the second paper, I found the most interesting part the appendices on software sales and employment. Noticed how the birth and boom of (p2p) filesharing hasn't caused a setback for either category ? Could it be its influence on the industry is highly overrated and/or unevenly lit on the negative side ????

Keep on sharing the grass in your pasture!

igalan · 06-25-2002, 04:08 AM

Quote:

Originally posted by Suredeath
[...]
When a TV tube is built it is sucked vacuum, filled with an inert gas to prevent the phosphor from 'burning up'. Not a lot of people know on top of the inert gas a predefined amount of pure oxygen is put in as well. Why ? To make sure the tube won't last much longer than the intended lifespan of say 5 to 8 years.
Is that bad practice ? It may seem unfair to intentionally make a product worse than possible, especially when this actually costs money to do. Again, there is an upside: tv's come cheap, and I bet they wouldn't when you'd buy one in for a lifetime, for not production cost sets the price, but company turnover and profit margin.

I find it very hard to judge in the above cases. You ?
Software has many similarities I think. It is possible to avoid all the nasty things mentioned in the article, and start with a proper foundation, a plan and structural design. Of course M$ arguably should have abstained from building its OSes dependent on diverse functionally separate modules (like Explorer etc). But would there be a M$ if they had done all that ?

IMO this is not the case in Software. The reason for this is that software is quickly evolving, so tomorrow you will want to do more than you do today. For this reason the software of today, even it has no programming defects, it's not a product you'll use a lifetime. Not to mention that hardware has a low period of obsolescency, so the sotware that is run will suffer from this and eventually will not be supported for the hardware.

Also most of the time, a software not working properly will produce lost of work, time and money (not to mention frustration of the user). While a TV that has to be repaired after 5 or 8 years is not that big problem...

The Software companies won't benefit from flawed software in any way, the users get upset and they will look for alternatives. And since it's not that hard to get GPL software that does almost the same as most commercial software, the options are getting wider making it more difficult for software companies.

The objective of any software company, no matter which product they sell, is to release the software free of defects. There will always exist something that can be improved or some new technologies that allow software to evolve, so users will buy new versions sooner or later.

06-25-2002, 01:49 PM

I have to disagree on your statement that the objective of any software company (...) is to release the software free of defects.
At best it is a secondary or lower-in-line objective. The objective of a (software) company is to make profit. And for a software toko that grosso modo means: get it on the market first, never mind the flaws now, we'll fix that later. Even the in-house development team at my work, who actually only develop for our own specialistic company have the same adagium, and they have to face their customers each day! Don't believe they really care about flaws and instability issues, or maintainability.

Yes I am dissapointed in that aspect, and promised myself I would never stoop to that attitude, but I am convinced software is far more a marketing business than perhaps any more physical-product-making line of business. If it looks good at the presentation, people will buy it.

The test & feedback phase of the normal production cycle is somply postponed to the production stage, and the customer becomes the tester. Have you seen the big masses run away from M$ after they all found out yet another patch is required ? I reckon' people even praise a company for being so supportive and giving them all these goodies after their purchase.

Do you think people buy a car if they know to have to bring it in 10 times after they bought it ? Guess not.

People know they are being fooled and they want to be!

gbread49 · 06-25-2002, 02:49 PM

The sad part about software marketing, is that it is probably the only product that can get away with that kind of presentation to the public. With the competition as intense as it is in software distribution, thorough testing is of low priority.

Even though it might be buggy and the developers/programmers know that, it is still made available for consumer use/purchase. As Suredeath stated, the consumers are the "Beta" testors.

When was the last time you saw a software developer "recall" a product! They just make a patch or upgrade available. What really turns me off is that when they want you to purchase the upgrade to fix problems!

bryce223 · 06-25-2002, 03:21 PM

I think "feature creep" is an issue( yes, I base all my observations on Dilbert

). I think most programmers would be happy to focus on the main purpose of a project and develop it to perfection. But they are forced to meet the demands of the target consumer who believes text editors need IM integration and they'd rather have an annoying popup "help wizard" than RTFM.

I'd take several small functional programs over one behemoth ram-sucking mother. For most of the Bloat-ware there is an alternative: http://www.tinyapps.org/

igalan · 06-26-2002, 05:10 AM

Quote:

Originally posted by Suredeath
[...]The objective of a (software) company is to make profit. And for a software toko that grosso modo means: get it on the market first, never mind the flaws now, we'll fix that later.

Of course that 1st objective is making money, but users demand software free of defects, and companies should work with this in mind. We all know how things really work, but this is no way of working, and in the near future many companies will have to take more care of the products they release. There is a growing movement demanding better software, and probably companies will start to be asked for quality plans (which our company is asked for each development), quality certifications (ISO)...

Currently users don't have many options besides Windows (Linux is not for the average user, right)? If you want a car you have many options that will satisfy your needs; if one car has known problems, you'll buy another. With OS is not the same. But also here there are some people who want to have Windows OS made by different companies besides MS, this way you'll have alternatives (Lindows for instance).

Quote:

People know they are being fooled and they want to be!

Not for a long time

!

06-19-2002, 07:58 AM	#1 (permalink)
Dobero Registered User Join Date: Apr 2002 Posts: 54	Why software is so bad ... Very interesting! Thanks to http://www.xp-erience.org/ A complete and very interesting analysis about the present situation in the software making community. It's a too long article because it's complete :-) so check it at http://www.msnbc.com/news/768401.asp?0si=-&cp1=1 On the same subject there's an interesting white paper (thanks to Paolo Attivissimo @ http://attivissimo.net/) "Open Source Software May Offer Target for Terrorists", so "interesting" and rumorouse that they've made a purchased version of it at http://www.adti.net/html_files/defen...5_30_2002.html But ;-) being part of a cool fellow cow's community you've been selected to read it for free, so don't disappointe them and read it! http://x.emdx.org/Slashdot/old_opens...whitepaper.pdf As you may expected, this institution ( Alexis de Tocqueville Institution - http://www.adti.net/) is sponsored by Micro$oft (http://www.wired.com/news/business/0,1367,52973,00.html) So, reflect and add comments. See yall.

06-25-2002, 02:49 PM	#7 (permalink)
gbread49 Registered User Join Date: Apr 2002 Location: Tennessee Posts: 744	The sad part about software marketing, is that it is probably the only product that can get away with that kind of presentation to the public. With the competition as intense as it is in software distribution, thorough testing is of low priority. Even though it might be buggy and the developers/programmers know that, it is still made available for consumer use/purchase. As Suredeath stated, the consumers are the "Beta" testors. When was the last time you saw a software developer "recall" a product! They just make a patch or upgrade available. What really turns me off is that when they want you to purchase the upgrade to fix problems! __________________

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

06-19-2002, 08:16 PM	#2 (permalink)
Doobie Registered User Join Date: May 2002 Posts: 19	Most programmers, like most people, are simply idiots. Sometimes, it's not the programmer's fault but their anti-competitive boss (e.g. Gates) who instead of keeping the OS simple as it should be wants to bloat it out with features to hurt other software companies (the integration of IE is a prime example). And, MS hiding APIs doesn't help the quality of software either. Software is much more complex than cars. Indeed, the only thing that really compares are microprocessor chips. Chips aren't as complex as their transistor count might seem to indicate. Chips are also designed by people who are more professional than most programmers and much more effort is made to debug chips than software. Still chips have bugs and programmers must work around those bugs. And, there was a big effort a few years ago to reduce the complexity of chips (remember RISC). Anyway, look how popular the auto repair industry is. There are only two kinds of baby toys and equipment, those that have been recalled and those that will be recalled (okay, that's a bit cynical). Most of the veggies imported into the US have to be tossed even before they make it to the store. That last salesman who tried to get info from probably wasn't right even half the time. Look how buggy our government's foreign policy is.

06-23-2002, 04:20 PM	#4 (permalink)
Suredeath Guest Posts: n/a	The reason we don't own cars that serve us for a lifetime is the car industry wants to have something to do tomorrow as well. Benefit is that cars come relatively cheap, otherwise they would be much more expensive. No I can own 10 cars in my life, and have a recent model with all its features. And millions of people are employed in the car repair industry worldwide, and they account for quite some purchasing power. When a TV tube is built it is sucked vacuum, filled with an inert gas to prevent the phosphor from 'burning up'. Not a lot of people know on top of the inert gas a predefined amount of pure oxygen is put in as well. Why ? To make sure the tube won't last much longer than the intended lifespan of say 5 to 8 years. Is that bad practice ? It may seem unfair to intentionally make a product worse than possible, especially when this actually costs money to do. Again, there is an upside: tv's come cheap, and I bet they wouldn't when you'd buy one in for a lifetime, for not production cost sets the price, but company turnover and profit margin. I find it very hard to judge in the above cases. You ? Software has many similarities I think. It is possible to avoid all the nasty things mentioned in the article, and start with a proper foundation, a plan and structural design. Of course M$ arguably should have abstained from building its OSes dependent on diverse functionally separate modules (like Explorer etc). But would there be a M$ if they had done all that ? On the second paper, I found the most interesting part the appendices on software sales and employment. Noticed how the birth and boom of (p2p) filesharing hasn't caused a setback for either category ? Could it be its influence on the industry is highly overrated and/or unevenly lit on the negative side ???? Keep on sharing the grass in your pasture!

06-25-2002, 01:49 PM	#6 (permalink)
Suredeath Guest Posts: n/a	I have to disagree on your statement that the objective of any software company (...) is to release the software free of defects. At best it is a secondary or lower-in-line objective. The objective of a (software) company is to make profit. And for a software toko that grosso modo means: get it on the market first, never mind the flaws now, we'll fix that later. Even the in-house development team at my work, who actually only develop for our own specialistic company have the same adagium, and they have to face their customers each day! Don't believe they really care about flaws and instability issues, or maintainability. Yes I am dissapointed in that aspect, and promised myself I would never stoop to that attitude, but I am convinced software is far more a marketing business than perhaps any more physical-product-making line of business. If it looks good at the presentation, people will buy it. The test & feedback phase of the normal production cycle is somply postponed to the production stage, and the customer becomes the tester. Have you seen the big masses run away from M$ after they all found out yet another patch is required ? I reckon' people even praise a company for being so supportive and giving them all these goodies after their purchase. Do you think people buy a car if they know to have to bring it in 10 times after they bought it ? Guess not. People know they are being fooled and they want to be!

06-25-2002, 03:21 PM	#8 (permalink)
bryce223 Registered User Join Date: Jan 2002 Posts: 160	I think "feature creep" is an issue( yes, I base all my observations on Dilbert ). I think most programmers would be happy to focus on the main purpose of a project and develop it to perfection. But they are forced to meet the demands of the target consumer who believes text editors need IM integration and they'd rather have an annoying popup "help wizard" than RTFM. I'd take several small functional programs over one behemoth ram-sucking mother. For most of the Bloat-ware there is an alternative: http://www.tinyapps.org/