Crippleware, Interfaces, Lines...
« previous entry | next entry »
Apr. 17th, 2008 | 01:05 pm
I've been getting pings all morning about my thoughts on the Slashdot article:
http://developers.slashdot.org/article.pl?sid=08/04/16/2337224
Google reveals my thoughts on crippleware. I've never been particularly quiet on the subject of the "Open Source 2.0" branding cycle of trying to say that a "mostly free" open source project is open source. The entire movement of "mostly free" pretty much disturbs me.
I do not believe though that this subject is cut and dry. I am going to use Innodb as an example. I do not consider it crippleware, and I have been amazed and delighted at Oracle's ownership of it. They have made excellent stewards of the source code.
Do they have a closed source backup tool? Yes.
Do I have an issue with this? No.
Why not? We can use Innodb just fine with or without the tool. There are no special versions of Innodb required in order to use innodb hot backup. No ifdefs, no requirements to use a non-gpl version of Innodb.
To me this is key.
It is an extension through an interface. In the last eight years anyone could have decided to re-implement the tool if they had wanted too. It has never happened, but anyone could have done it.
Interfaces to me are what make the difference. Is the interface open to competition, or is it hidden behind proprietary binaries? Proprietary smells of protectionism to me (I'm a fiscal conservative... I dislike the boo hoos of those who want a protected market).
The proposed backup drivers I have not seen. As to which side of the fence they land on at this point, or will land when they are released, is something I am unsure of.
There is a rats nest of questions regarding this strategy.
Will MySQL give out the source to customers?
Do you want the source to them? (aka take on the risk of accidently providing proprietary code to the open market).
Can I still compile code?
Monty lead a talk today on the "Future of MySQL". I am hoping his slides will be published. Having the audience chant "We don't ship crippleware" was a highlight for many I assume.
There really is an open question though, that can't be glossed over, and one that chants do not resolve.
Linbit, makers of DRBD, have close sourced versions of it. Does it invalidate what is great product? How about the guys who did Tripwire? Is Redhat a good open source steward because they do not open source the infrastructure around their Enterprise network? Do Google, Yahoo, or Facebook consume more open source then they produce?
I believe that at the end of the day most everything will be open source. If drivers and backup are important they will be created.
Companies that have invested in open source will steer clear of ambiguous environments, but they will continue to have a heterogenous environment filled with open source and close software which will inter-opt.
We live in a very mixed software world.
http://developers.slashdot.org/article.pl?sid=08/04/16/2337224
Google reveals my thoughts on crippleware. I've never been particularly quiet on the subject of the "Open Source 2.0" branding cycle of trying to say that a "mostly free" open source project is open source. The entire movement of "mostly free" pretty much disturbs me.
I do not believe though that this subject is cut and dry. I am going to use Innodb as an example. I do not consider it crippleware, and I have been amazed and delighted at Oracle's ownership of it. They have made excellent stewards of the source code.
Do they have a closed source backup tool? Yes.
Do I have an issue with this? No.
Why not? We can use Innodb just fine with or without the tool. There are no special versions of Innodb required in order to use innodb hot backup. No ifdefs, no requirements to use a non-gpl version of Innodb.
To me this is key.
It is an extension through an interface. In the last eight years anyone could have decided to re-implement the tool if they had wanted too. It has never happened, but anyone could have done it.
Interfaces to me are what make the difference. Is the interface open to competition, or is it hidden behind proprietary binaries? Proprietary smells of protectionism to me (I'm a fiscal conservative... I dislike the boo hoos of those who want a protected market).
The proposed backup drivers I have not seen. As to which side of the fence they land on at this point, or will land when they are released, is something I am unsure of.
There is a rats nest of questions regarding this strategy.
Will MySQL give out the source to customers?
Do you want the source to them? (aka take on the risk of accidently providing proprietary code to the open market).
Can I still compile code?
Monty lead a talk today on the "Future of MySQL". I am hoping his slides will be published. Having the audience chant "We don't ship crippleware" was a highlight for many I assume.
There really is an open question though, that can't be glossed over, and one that chants do not resolve.
Linbit, makers of DRBD, have close sourced versions of it. Does it invalidate what is great product? How about the guys who did Tripwire? Is Redhat a good open source steward because they do not open source the infrastructure around their Enterprise network? Do Google, Yahoo, or Facebook consume more open source then they produce?
I believe that at the end of the day most everything will be open source. If drivers and backup are important they will be created.
Companies that have invested in open source will steer clear of ambiguous environments, but they will continue to have a heterogenous environment filled with open source and close software which will inter-opt.
We live in a very mixed software world.
(no subject)
from:
dormando
date: Apr. 20th, 2008 03:13 am (UTC)
Link
Personally I'm a bit more extreme and would prefer everything related to a utility be open. Even if someone could reimplement a tool that's just as good, it might not gain usage due to FUD around not knowing just what features the closed source tool has.
I'm trying not to make the slippery slope comment, but at first if you're missing a backup interface, next you're missing everything else. MySQL proxy is missing many basic features at this point. Sure they're all plugins to an existing interface, but the tool is almost useless to the community at large.
Systems which are complete and useful out of the box help companies grow. We have many website startups now. With technologies like OpenID which will help them accrue registrations more easily, they need to also aquire backend technology more easily.
Wouldn't the biggest users of online backup be folks who can't afford having backup slaves around? That's just punishing small business.
Reply | Thread
(no subject)
from:
krow
date: Apr. 20th, 2008 07:03 am (UTC)
Link
Personally? I wouldn't publish any of my projects and hold back "commercial features". It seems like a great way to limit your market. I understand the temptation to, but personally I would not do it with my stuff.
On the topic of backup... its just MyISAM. Archive has had a public backup utility for years. I'd like to see someone create one on Innodb but... frankly I use LVM snapshots, so none of this means a thing to me.
If proxy turns out to be useless, then we will just have to see where yours goes. Hell... you don't have to wear gloves so I'd place bets on your if you push it to completion.
Reply | Parent | Thread
(no subject)
from:
awfief
date: Apr. 20th, 2008 03:57 am (UTC)
Link
Monty lead a talk today on the "Future of MySQL". I am hoping his slides will be published. Having the audience chant "We don't ship crippleware" was a highlight for many I assume.
I am very sad that I wasn't recording that talk. Although that can become the motto of the user community, and we can just spontaneously break out into chanting that, whenever.
Oh, and as you already know but I figured I'd link here, the slides are at: http://www.scribd.com/doc/2575733/The-future-of-MySQL-The-Project
I'm not for a "protected" market, but it is important for a company to be viable so they can continue creating good software. And certainly, if MySQL isn't crippled due to lack of backup now, it wouldn't be in a year from now......although to be clear, there will be more backup in a year from now, than there is now....
but yeah, I insist that Workbench is crippled until you can actually print from it....but online backups without compression or encryption? A fellow traveler I was talking to in the airport said "I was in the partner meeting when the controversy happened, and it really sounded like someone wanted their blog to get more hits."
I don't know if I agree with that, but it definitely was blown out of proportion.
Reply | Thread
(no subject)
from:
krow
date: Apr. 20th, 2008 06:17 am (UTC)
Link
No printing == crippleware though. So if that is the case... then yes it would probably meet that definition.
Reply | Parent | Thread
(no subject)
from:
awfief
date: Apr. 20th, 2008 04:48 pm (UTC)
Link
But yeah, not having encryption nor compression doesn't make a backup solution crippled, and having an online backup solution is just that much better.
Reply | Parent | Thread