SVN access?

[gslender]

As Ext GWT is going through Beta, there are bound to be heaps of bugs... how do we get access to SVN whilst fixes are being developed and tested?

If the answer is commercial support - are all the forums and support features ready for silver support?

[darrellmeyer]

SVN access is available to support license holders. And yes, silver support and its features are available now.

[gslender]

Darrell,

I think when the product is released (ie production ready and not beta) I think that model of restricting access to the next beta release makes sense.

It this case, I'm not encouraged to find bugs (but rather work around them) as I'm not going to get the benefit of instant fixes and no understanding about how long I'll need to wait until the next ZIP release is available.

This is where commercials get in the way of what good products should be - the point of opensource is the continual access and shared development - I guess this isn't the way Ext JS want their products developed? :-?

It's a shame - here's hoping that releases roll quickly and quality remains high.... /:)

Cheers,
Grant

[kindkid]

Hi Darrell,

I use the library under the terms of the GPL. When users like myself create fixes for something or implement some missing functionality, which of the following is the expected course of action?

1. They are to purchase a $200+ silver support license so that they can contribute an up-to-date patch, ready to be merged.
2. They are to send your team a patch against the most recently released zipped source code, and your team will port it manually if/when you have time. Users who wish to use others' contributed fixes/functionality must wait for your next release.
3. They are to create a public fork based on the most recently released zipped source, and apply their patch to the fork. Users who wish to use contributed fixes/functionality (under a GPL license) use (and contribute to) the public fork instead. Your team will occasionally manually pluck the good fruit from the public fork and port it over into your repository.
4. They are to go elsewhere (ie - create a public fork under the GPL). Ext JS will not be interested in their contributions.

Also, just a point of clarification. As far as I know, someone who creates a derivative work under the GPL still maintains copyrights of their specific modifications by default. If Ext JS, LLC then incorporates that work back into their product, Ext JS would then be bound by the GPL with regard to those modifications (as time goes on, perhaps even the whole code base due to the expansive influence of the GPL). This would damage Ext JS's ability to dual-license. Not a situation I'd like to see your lovely company fall into. :( Have I understood the GPL correctly? At a minimum, I'd suggest that you require contributors to place their modifications under a new BSD style license or in the public domain, before accepting the contributions into your code base. Note that this goes against the quid-pro-quo ideal, but at least it would be in Ext JS, LLC's favor, eh?

I'd also like to gently suggest that if Ext JS is interested in incorporating fixes/features from contributors, it may save you some work to encourage, rather than discourage such would-be-contributors to write patches against the latest in SVN trunk. ;)

By the way, Ext GWT is a lovely library indeed! Excellent work!

[gslender]

+1

C'mon ExtJS - wake up and smell the roses

In AUD Silver SVN access is $320 - if SVN only access was more like $50/year (to be an ExtJS Member) I'd consider paying, but $320 is just too steep to pay for the privildge of getting what will be available soon anyway - and mostly it only really helps ExtJS test/bug/patch fix their code.... ???

[brian.moeskau]

kindkid is absolutely correct re: our ability to dual license, which is precisely why we cannot open up SVN to general public contributions. Any contributed code would have to be in the public domain or we would require a CLA assigning copyright to Ext. Again, this is not because we are trying to profit off of anyone -- it's required legally for us to be able to include such code. We generally do not actively solicit contributions for those reasons, although when people are interested in contributing and are willing to accept the terms, we are perfectly willing to accept improvements where appropriate.

Regarding our support subscriptions, I'd like to correct the apparent assumption that the price is for "SVN access." Support includes support credits that can be used for direct email/phone support from the core team, access to post in the premium forums, guaranteed SLAs and more, depending on support level. SVN access is considered a support feature because that's how people get access to the latest bug fixes between releases, but again, it's only one component of what you are paying for.

[gslender]

brian,

any chance of a revised support cost whereby SVN access is the only service? not everyone wants to pay for support (or needs it) and given the costs of providing vs benefit, I'd imagine many more folk like me (who prefer to exist in the OSS side of your commercial model) would jump at the chance...??

ie - I would have thought 10,000x $50 is better than 100x $299 ?

[brian.moeskau]

Can you guarantee us 10,000 non-support, SVN-only users?? :D

It's come up before, and maybe something we can consider. I'll have to get back to you...

[gslender]

I can guarantee not getting 10,000x @USD$299 support users who just want SVN access..... cheeky hey :>

[avrecko]

While ExtGwt is in beta (read as half baked and infested with bugs) you could at least provide daily or at least weekly builds to be available. I find that even better than SVN access.

It makes little sense to pay lots of money for something that is far from complete - yes, double clicking on tree arrow makes the whole nodes dissapear bug is quite a showstopper in my book. Saying fix is in SVN helps very little.

That is like buying a new car which is missing a bumper, the roof is leaking and the tail lights are not working but please do come around to the mechanic he will be happy to fix it for you(for $). That just doesn't add up.

Right now we are helping you with beta testing you could at least provide some daily or weekly builds, it can only help with testing and spotting new bugs or regression bugs.

[Canto]

While ExtGwt is in beta (read as half baked and infested with bugs) you could at least provide daily or at least weekly builds to be available. I find that even better than SVN access.

It makes little sense to pay lots of money for something that is far from complete - yes, double clicking on tree arrow makes the whole nodes dissapear bug is quite a showstopper in my book. Saying fix is in SVN helps very little.

That is like buying a new car which is missing a bumper, the roof is leaking and the tail lights are not working but please do come around to the mechanic he will be happy to fix it for you(for $). That just doesn't add up.

Right now we are helping you with beta testing you could at least provide some daily or weekly builds, it can only help with testing and spotting new bugs or regression bugs.

I agree and vote for DAILY builds.

[kindkid]

Hi Brian,

First, thank you for your response. :)

Now, I can only speak for myself, but I thought it might be worthwhile to share my perspective under the assumption that there may indeed be others like myself. I quite enjoy using Ext GWT so far, and hope to use it for a personal project or two, or anywhere else that the GPL permits. I have already come across a few places that could use some improvements, and my first inclination is to do the following:
1. Browse your bug database to make sure it isn't already fixed (so that I'm not duplicating effort), post a new bug if appropriate.
2. Import Ext GWT into an SVK repository (refreshing first from your SVN trunk).
3. Write the improvement.
4. Commit to my local repo for my own benefit. (SVK does star-merging, so it's no problem)
5. Generate a patch, and attach the patch to the bug (with a "released into the public domain" agreement), for your team to take-or-leave as you see fit.
6. Feel good and continue to enjoy Ext GWT.

At no point would I need any support. However, in order for this to work, I'd need at least read access to the bug tracker and read-only access to SVN. Now, I can only speak for myself here, but if getting access to a bug tracker and SVN to allow me to contribute required me to pay, I would consider it only briefly before realizing that I could just as easily (perhaps even more easily) do the following instead:

1. Search google to see if anyone else has started up a public fork of Ext GWT yet.
2. Import the latest zipped sources into a googlecode project under the GPL.
3. Use the googlecode issue tracker to track my changes.
4. Write the improvement.
5. Commit to the fork.
6. Accept contributions from anyone else willing to accept the terms of the GPL and release their modifications into the public domain.
7. Occasionally port the latest improvements from Ext back into the fork.
8. Feel good and continue to enjoy Ext GWT (perhaps under a different name).

Note that this would be a sad thing though, as ports from the fork to your trunk or vis versa would become more and more painful over time, until finally they become two entirely separate projects, or one side is abandoned and dies of code rot. As long as both sides made sure to limit contributions to the public domain, users would continue to enjoy the ability to migrate to a commercial license from Ext JS, LLC if needed. (However, if the fork began to accept non public domain contributions, this benefit would be lost. Unfortunately, it is not clear that a fork would have any motivation for preventing this from happening)

In fact, aside from licensing flexibility, merely the hassle of porting modifications into (or out of) a fork makes keeping all the changes in one (your) bug tracker and source tree seem the most preferable choice.

Well, anyway, that's my perspective: I'm interested in contributing up-to-date patches for nothing more than a thanks, or at worst, the sound of crickets. But with no other benefit guaranteed than that, I'd sooner add my contributions to a fork than cough up any amount of cash. ;)

So now, I guess the question is: Would your team prefer to offer free unsupported SVN (and bug tracker) access to folks like me who agree to release contributions to the public domain... or would you prefer that we congregate around our own public fork instead? (this assuming that "we" ends up being more than just me :">) I'm happy either way, but naturally, I prefer the former.

Thank you for your thoughtful consideration, and again for your willingness to contribute to (and hopefully benefit from!) the open source community. You have an excellent product going here! :D

[screw]

Is there are some changes since Apr 2008 in Ext JS distribution policy?