Disclaimer: This is a personal web page. Contents written here do not represent the position of my employer.
Friday, December 26, 2008
On open source software and technical support: the reality
I was curious about some discussion threads posted in LinkedIn's Open Source group. One had the following content:
Is 2009 The year of "supported open source" Do you agree or disagree ?
I have to admit having worked on both sides of the fence the biggest trend we are now experiencing on web related project requirements is the shift to "supported open source technology"...
The days of 6 figure plus licence fees are now over "phew" our clients thankfully quote...
Old versus New
Interwoven / Vignette versus Alfresco / MySource Matrix ?
Salesforce versus Sugar CRM ?
Documentum / Trim versus Alfresco / Sharepoint Integrated CMS's
What are your thoughts fellow CM Professionals ? Don't just take my word for it... http://cmswatch.com/Feature/189
I read some interesting opinions below, such as this one from Tony Wasserman:
To me, the issue here is adoption and use of open source software by companies and governments for activities that are essential to their businesses. One of the key corporate criteria for that type of software acquisition is availability of long-term 24x7 support, including service level agreements for problem resolution. These CIO's and other decision-makers just don't accept the idea that they can post their issues on a forum and that someone will post a reply. They want a toll-free telephone number to call with the assurance that someone knowledgeable will answer and address the issue.
For example, while small companies and the technically knowledgeable will download and install the Drupal CMS (and its underlying components), the corporate buyer, by contrast, will go to Acquia, gladly paying for support. As a better example, many of the installations of open source software in corporate environments are being done through consultants and system integrators, such as Wipro, Infosys, Accenture, and IBM.
So now the next question is whether the demand for open source software will increase in 2009. For that, I would give a qualified "yes". Startups have been building their companies on open source software for most of this decade, and will continue to do so, but they are not spending a lot of money on support, consulting, and training services. Developers in companies of all sizes have also long used open source software packages and libraries, including Eclipse, NetBeans, JBoss, MySQL, and hundreds of others, again without much money changing hands.
For enterprise applications, companies have already made substantial investments in existing proprietary software from Oracle, SAP, Microsoft, and other vendors, so it is hard for them to justify replacing it. The enterprise opportunities for open source come with new applications, and with growing companies that need these applications.
So I think that there will be growing revenue in 2009 for the system integrators providing open source support and services, as well as some money for the vendors of commercialized open source products, such as Zenoss, Groundwork, and other infrastructure packages. I think that the growth for the application packages is probably a little further out, though I would make an exception for OpenOffice.org , which has already seen a lot of acceptance.
Or this one from Glenn A. Curry:
Tony brings up an interesting issue. That of access to tech support. But our experience has been that the tech support available from commercial software manufacturers is in many cases worse than nonexistent. Not all have 24x7. And even if there is someone that answers the phone, there typically is little value in talking to them.
In trying to resolve software issues for our customers, we are most often given answers like "just keep reinstalling it till it works". Or they want security settings reduced because their interface is sloppy. Or they just flat give bad answers. The extremely poor level of phone support so often found is incredible. The battle required to move up the chain to someone that actually has a clue is frustrating and costly to the customer as they have to pay for our time while on the phone.
e.g. we spent days dealing with Sage on a Peachtree upgrade. Their solution (besides the before mentioned "just keep reinstalling it, that usually makes it work eventually") was to run as administrator all the time to get around a security problem. Eventually we found the problem and solved it with some port mapping. It was a common problem based on discussions on techie sites. AFAWK it is still a problem for them. We sure are not going to tell them how to fix THEIR bad software design for free. And noe one there had a clue.
And I won't even begin to list the massive Microsoft tech support nightmares.
Just because a company claims 24x7 tech support, does not mean it has an actual usable value.
So, without trying to be devil's attorney, I gave my opinion:
I think Tony and Glenn raise good points about support, but I would like to arise a third point, that maybe contrasts with them. The issue is not about support itself, but about what does "support" mean in open source software:
1) Yes, in open source, there's no monopoly about technical support, and this is always a good thing, except for the cases in which it can be seen wrongly as bad quality, because:
a) It may happen that there's no company behind an open source project, just one or some developers, and it's a bit awkward to ask them for support if they're not professionally gathered in some way. Besides, if the developer base is so small, it's going to be difficult to find another company which gives you support if they don't know much about the product.
b) It may happen that the developer-base is spread around many companies (for example: OpenOffice, the Linux kernel), so the company that needs support is normally confused as to which one is the best one to request support. This is awkward to a company which normally considers a product to come from "just one company".
2) Propietary software may mean a monopoly in regards to technical support but, let's face it: many incompetent people working inside companies just prefer some provider to blame if something goes wrong: "we're stuck in this issue because Oracle (or put you're favorite propietary software company here) technical support sucks", and it's not their fault anymore. If they had chosen an open source software solution, they may be accused of:
a) Not choosing the best provider for technical support.
b) Not fixing the issue themselves, since, if you have the code, it's just a matter of time to solve it, right?
Well, just trying to be realistic. Of course I advocate open source (I work on it on a daily basis), but just wanted to give some clues about why it's not getting the success it should have.
In general, the (2) point and (2a) in particular, is very similar to the reason why still incompetent but huge companies survive on the market: I've heard many cases of people just buying Oracle or Accenture because they're the most expensive. If all goes wrong, it's impossible they would be accused of selecting the wrong provider if they selected the most expensive one, right?
UPDATE 16-FEB-2008: Even though the above given problems about choosing open source solutions, here's a nice guide to try to overcome them: Convincing Management To Approve Free Software.
Is 2009 The year of "supported open source" Do you agree or disagree ?
I have to admit having worked on both sides of the fence the biggest trend we are now experiencing on web related project requirements is the shift to "supported open source technology"...
The days of 6 figure plus licence fees are now over "phew" our clients thankfully quote...
Old versus New
Interwoven / Vignette versus Alfresco / MySource Matrix ?
Salesforce versus Sugar CRM ?
Documentum / Trim versus Alfresco / Sharepoint Integrated CMS's
What are your thoughts fellow CM Professionals ? Don't just take my word for it... http://cmswatch.com/Feature/189
I read some interesting opinions below, such as this one from Tony Wasserman:
To me, the issue here is adoption and use of open source software by companies and governments for activities that are essential to their businesses. One of the key corporate criteria for that type of software acquisition is availability of long-term 24x7 support, including service level agreements for problem resolution. These CIO's and other decision-makers just don't accept the idea that they can post their issues on a forum and that someone will post a reply. They want a toll-free telephone number to call with the assurance that someone knowledgeable will answer and address the issue.
For example, while small companies and the technically knowledgeable will download and install the Drupal CMS (and its underlying components), the corporate buyer, by contrast, will go to Acquia, gladly paying for support. As a better example, many of the installations of open source software in corporate environments are being done through consultants and system integrators, such as Wipro, Infosys, Accenture, and IBM.
So now the next question is whether the demand for open source software will increase in 2009. For that, I would give a qualified "yes". Startups have been building their companies on open source software for most of this decade, and will continue to do so, but they are not spending a lot of money on support, consulting, and training services. Developers in companies of all sizes have also long used open source software packages and libraries, including Eclipse, NetBeans, JBoss, MySQL, and hundreds of others, again without much money changing hands.
For enterprise applications, companies have already made substantial investments in existing proprietary software from Oracle, SAP, Microsoft, and other vendors, so it is hard for them to justify replacing it. The enterprise opportunities for open source come with new applications, and with growing companies that need these applications.
So I think that there will be growing revenue in 2009 for the system integrators providing open source support and services, as well as some money for the vendors of commercialized open source products, such as Zenoss, Groundwork, and other infrastructure packages. I think that the growth for the application packages is probably a little further out, though I would make an exception for OpenOffice.org , which has already seen a lot of acceptance.
Or this one from Glenn A. Curry:
Tony brings up an interesting issue. That of access to tech support. But our experience has been that the tech support available from commercial software manufacturers is in many cases worse than nonexistent. Not all have 24x7. And even if there is someone that answers the phone, there typically is little value in talking to them.
In trying to resolve software issues for our customers, we are most often given answers like "just keep reinstalling it till it works". Or they want security settings reduced because their interface is sloppy. Or they just flat give bad answers. The extremely poor level of phone support so often found is incredible. The battle required to move up the chain to someone that actually has a clue is frustrating and costly to the customer as they have to pay for our time while on the phone.
e.g. we spent days dealing with Sage on a Peachtree upgrade. Their solution (besides the before mentioned "just keep reinstalling it, that usually makes it work eventually") was to run as administrator all the time to get around a security problem. Eventually we found the problem and solved it with some port mapping. It was a common problem based on discussions on techie sites. AFAWK it is still a problem for them. We sure are not going to tell them how to fix THEIR bad software design for free. And noe one there had a clue.
And I won't even begin to list the massive Microsoft tech support nightmares.
Just because a company claims 24x7 tech support, does not mean it has an actual usable value.
So, without trying to be devil's attorney, I gave my opinion:
I think Tony and Glenn raise good points about support, but I would like to arise a third point, that maybe contrasts with them. The issue is not about support itself, but about what does "support" mean in open source software:
1) Yes, in open source, there's no monopoly about technical support, and this is always a good thing, except for the cases in which it can be seen wrongly as bad quality, because:
a) It may happen that there's no company behind an open source project, just one or some developers, and it's a bit awkward to ask them for support if they're not professionally gathered in some way. Besides, if the developer base is so small, it's going to be difficult to find another company which gives you support if they don't know much about the product.
b) It may happen that the developer-base is spread around many companies (for example: OpenOffice, the Linux kernel), so the company that needs support is normally confused as to which one is the best one to request support. This is awkward to a company which normally considers a product to come from "just one company".
2) Propietary software may mean a monopoly in regards to technical support but, let's face it: many incompetent people working inside companies just prefer some provider to blame if something goes wrong: "we're stuck in this issue because Oracle (or put you're favorite propietary software company here) technical support sucks", and it's not their fault anymore. If they had chosen an open source software solution, they may be accused of:
a) Not choosing the best provider for technical support.
b) Not fixing the issue themselves, since, if you have the code, it's just a matter of time to solve it, right?
Well, just trying to be realistic. Of course I advocate open source (I work on it on a daily basis), but just wanted to give some clues about why it's not getting the success it should have.
In general, the (2) point and (2a) in particular, is very similar to the reason why still incompetent but huge companies survive on the market: I've heard many cases of people just buying Oracle or Accenture because they're the most expensive. If all goes wrong, it's impossible they would be accused of selecting the wrong provider if they selected the most expensive one, right?
UPDATE 16-FEB-2008: Even though the above given problems about choosing open source solutions, here's a nice guide to try to overcome them: Convincing Management To Approve Free Software.
Labels: General, SoftwareLibre
