Wow!  I feel a lot better now.  I said "this is a talented and generous group 
of people" and it just proved itself again.

(a) my apologies Pascal .. of all the people's names I could screw up, yours is 
most inappropriate.

(b) Chuck was 100% right on the spilt install suggestion (I claim the small 
boast that I found it before reading his email!)

(c) There is a lot to read in the three replies so far.  I will read, digest 
and respond as necessary.

(d) I'm back running W6 !!

(e) I will contribute to the project.

Thank you .. very much, Gavin

On Jan 14, 2013, at 5:44 PM, Johann Werner <j...@oyosys.de> wrote:

> 
> Hi Gavin,
> 
> Am 14.01.2013 um 20:55 schrieb Gavin Eadie <ga...@umich.edu>:
> 
>> On Jan 14, 2013, at 12:25 PM, James Cicenia <ja...@jimijon.com> wrote:
>> 
>>> Make sure in your Eclipse Preferences --> WOLips--> Build Preferences 
>>> settings you don't have  "Generate Bundle" checked.
>>> 
>>> That has bit me twice now and consumed many frustrating hours.
>> 
>> A diversion from a grumpy Gavin using "Generate Bundle" as an example of my 
>> trouble.
>> 
>> 
>> I've been programming using WebObjects, off and on, for ten years (and using 
>> Wonder for the last six/seven), and still haven't a clue what "Generate 
>> Bundle" actually means or does.  The most cogent description found by 
>> searching for the phrase in the WOCommunity web site is, "A bundleless build 
>> means that you have unchecked the build option Generate bundles within the 
>> WOLips preferences" .. yippee!  I would like to fix things like that for the 
>> next confused person.
> 
> the problem lies in the complexity of a full fledged WO development: you need 
> WO, Eclipse, WOLips, Wonder, Java, property files, developer tools, … that 
> makes it easy to get something wrong and difficult to understand all the 
> inner workings and interactions. For most people things like bundleless 
> builds should be an unimportant detail and just a free enhancement for those 
> who know (or even need) that feature. That it was needed for James to fix his 
> problem seems not right but is probably due to the aforementioned complexity 
> and the resulting multitude of different installations/settings that are 
> possible.
> 
>> I applaud Robert's (and everyone's) work on Wonder 6 without reservation, 
>> and I sincerely wish I could contribute to it but I feel just enough below 
>> the "expert" level to doubt my abilities to do that.
> 
> Of course many things are complicated (or at least seem so) and feeling 
> daunted to change anything in Wonder is a normal reaction–keep in mind that 
> Wonder is the collected knowledge of so many developers contributing over so 
> many years not to count all the effort and ideas NeXT put into WO and its 
> underlying design! But then we all have begun with no knowledge at all and 
> are now nonetheless creating applications and frameworks with it. Don't think 
> that you have to be an expert to be able to add anything to Wonder. Sure 
> there are parts in Wonder you need to get your head wrapped around–even twice 
> or thrice ;)–but that's only a part of it.
> 
> There are so many things that can be done (and have to be done) in Wonder 
> that doesn't need you to have a doctor's degree in computer science. Let it 
> be simple things like the addition/correction of Javadocs, renaming of 
> cryptic variable names to more meaningful ones, adding generics or like you 
> said "to fix things like that for the next confused person" by contributing 
> to the wiki. Everyone can take an active part in the community. Don't think 
> that that type of contribution is worth less than a complete framework that 
> will cut the development time of everyone's next big project by half. In sum 
> even those little enhancements will add value, make it easier for the next 
> person to understand / use the code and to contribute himself. And with every 
> commit and change you make you gain more insight in Wonder's code and results 
> in you creating better apps or develop them faster (by knowing where to look 
> for a specific method, by using existing code instead of reinventing the 
> wheel, …) and finally being able to make more complex changes in Wonder.
> 
> I hope this doesn't sound too pathetic but I think the community as a whole 
> has a lot more potential to make Wonder even more… wonderful. You think you 
> aren't good enough to create the Next Big Thing? Then make your Next Less Big 
> Thing, even then there will be many people appreciating it. You are not sure 
> if your patch is good / correct / appropriate? Don't hesitate to make a pull 
> request, github has such great features to discuss pull requests and annotate 
> the code with comments and questions waiting to be used. You are saying that 
> you don't have so much time left for such "unpaid work"? In my experience the 
> most difficult part is to know what is actually in Wonder. Wonder is such a 
> big tile of code and ideas that sometimes it is literally a search for a 
> needle in a haystack to find a class or method that does exactly what you 
> need just in the moment you need it. Knowing what Wonder already offers you 
> for free makes it easier to tell your client what you can do for him and how 
> much time it will take to create his application.
> 
> Ok, have to stop here now ;-)
> 
>> For another example, I've no idea why *.api files are included in a 
>> deployment build when they are a purely build-time artifact (aren't they?).  
>> My version of build.xml hasn't copied them to a build *.woa (or *.war) for 
>> many years so I think I'm right -- but the WOLips build.xml does copy them 
>> to Resources and that's so engrained in WOLips that it must be intentional 
>> (yes/no?).  Should I submit an update via git to improve this build.xml is 
>> this, and other, ways?  My feeling is, "No Way!" -- there's hardly a more 
>> critical file in W6 -- I'd be crazy to touch it!
> 
> Uhm, keep those api files for now, there are some ideas I have in mind, just 
> have to find… some… spare… time :-O
> 
>> FWIW: I'm grumpy because I'm caught, again, with a scenario that baffles me 
>> -- an app that works correctly on my development Mac that when, when rsync'd 
>> to Amazon EC2, doesn't (clicking in an Ajax.framework component doesn't fire 
>> the "action" method).  This behavior started when I moved to Wonder 6 so I 
>> need to reverse that decision, and stop using W6, to stay productive.
> 
> That's no good. But it must be something wrong in your deployment. Did you 
> check your browser if it gets all resources, are there any javascript errors? 
> Does it work with Wonder 5.8.2?
> 
> jw
> 
>> That puzzle plays into this message too.  I'm good enough at this stuff that 
>> I could eventually find the cause, most likely something I missed, less 
>> likely a bug, version skew, etc, by diverting time from my paid work and 
>> digging into this, but it's not worth it (at least not this month) because I 
>> have a product to ship.  That doesn't improve my mood -- I'm a devotee of 
>> this technology, I'm a believer in open software, this is a talented and 
>> generous group of people -- I'd love to contribute, but don't feel that I 
>> can, that feels wrong in so many ways.
>> 
>> 
>> Does this dilemma strike an accord with anyone?  Actually, I know the answer 
>> to that is "Yes" (if you have personal copies of Wonder or WOLips that you 
>> use on a regular basis, it should be familiar).
>> 
>> I feel quite uncomfortable sending this message -- it was generated out of a 
>> long felt frustration and triggered by James' ".. has bit me twice now and 
>> consumed many frustrating hours", and would be much better as part of a 
>> face-to-face conversation but that's not an option.  In no way is it my 
>> intent to brush anyone's fur the wrong way.
> 
> 


 _______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list      (Webobjects-dev@lists.apple.com)
Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/webobjects-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Reply via email to