Over the last 5 years of working at ZC, one of the major things I've learned is to resist, and resist with good reason is TTW development. TTW coding in any shape or form is wrong IMNSO unless one has the mechanism for bringing the result out into the file system where 'normal' tools can manipulate, examine, and modify the results of TTW code. I find TTW models such as ZClasses abhorrible. Like many others who have had to pay the price of their predecessors of have developed in this model, I quickly came away knowing this model is inherently wrong. If the tools were inplace to make this transparent to the file system development I do today, I might not care. They aren't - and IMNSO, they're a long ways off. The cost of writting them solves what I would ask? Lowering the cost of admission?
I would argue further on Chris' note that Marketing cares whether non-technical people can change the behavior of the system. I would argue they are more concerned with changing the 'view' of the system, rather then the underlying technical implementation. To achieve this we need robust applications which run the web sites they would care about. Widgets, portlets, etc. are they key here IMNSHO. Developers/Programmers write these components on the file system: Non-technical site users/admins use these to (possibly substantially) change the 'view' or behavior of their sites.
To go back to cost of entry, because I've heard Paul so many times backlash against the 'mode' of Zope Coders and how 'they' would tend to go off on it's own design paths, not reusing or reusing existing patterns or models. In doing so, I will agree 'we' often made things more difficult then they should have been, but I believe this is an entirely different issue. The cost of entry, what is it? What should it be? It shouldn't be such that it leads new users down the WRONG paradigm to get them to 'drink the punch'. That quickly results in people getting ill and moving on, having the wrong impression of the capabilities and real power of Zope/Python. I would ask Paul, 'what is a content developer' anyhow?
Having been an early PHP adopter back in the first alpha and beta versions, I was apart of it's explosion. I do not think PHP got the fame it did solely because of it's cost of admission. It quickly won a user base due to the applications which were written on it. Users would download those, and then decide they wanted to change, enhance, or redo the application. They either joined an existing project or started their own. The sad fact IMNSHO is Zope never got that 'ready to use' application subset out there and accessible to the general masses. Most of us using ZOpe were writting bigger, complex applications for internal or customers.
This may have nothing at all to do with what prompted Chris to 'rant' about Paul to start this blog. I do agree that Zope 3 is extermely exciting and I for one am glad they are not going down the TTW paradigm (and pray they do not). If someone needs this, they should step back and look at the applications that should be written to allow the Marketing Depts. use cases to be a reality. Having paid the painful price of rewritting the ZClass product that is referenced by Chris that replaced Vignette, I would prefer that individual had the 'right' documentation and models to have 'drank the punch' more correclty. It was painful. The resulting ZClass application was crap. He personally didn't write the app I (and Jens) rewrote, but that was the model being pushed by ZC at the time, and it was wrong. Period. It might be quick; yet as Chris points out, for business the severity of the impact cannot be an option.
This leads me to my final point, which is that of 'there is no template designer'. I think there is. I think though, in reality most successfull ones are quickly turned into 'scripters' - which is by my definition, a programmer (of sorts). Chris and I were talking on Friday about some issues I'm confronting at my new job, and I'm very exicted about the model which Meld has for these people. I think there is the possibility that they 'guy' who writes my sites' look and feel has 0 knowledge of the logic that actually turns that into a end user facing web page. Meld looks bad ass for this.
I had to go off on meld, because I'm working on some ideas and was going to blog on it myself. :)
Oh, by the way....sup Paul - you owe me beers for pushing this insanity!
:)
I would argue further on Chris' note that Marketing cares whether non-technical people can change the behavior of the system. I would argue they are more concerned with changing the 'view' of the system, rather then the underlying technical implementation. To achieve this we need robust applications which run the web sites they would care about. Widgets, portlets, etc. are they key here IMNSHO. Developers/Programmers write these components on the file system: Non-technical site users/admins use these to (possibly substantially) change the 'view' or behavior of their sites.
To go back to cost of entry, because I've heard Paul so many times backlash against the 'mode' of Zope Coders and how 'they' would tend to go off on it's own design paths, not reusing or reusing existing patterns or models. In doing so, I will agree 'we' often made things more difficult then they should have been, but I believe this is an entirely different issue. The cost of entry, what is it? What should it be? It shouldn't be such that it leads new users down the WRONG paradigm to get them to 'drink the punch'. That quickly results in people getting ill and moving on, having the wrong impression of the capabilities and real power of Zope/Python. I would ask Paul, 'what is a content developer' anyhow?
Having been an early PHP adopter back in the first alpha and beta versions, I was apart of it's explosion. I do not think PHP got the fame it did solely because of it's cost of admission. It quickly won a user base due to the applications which were written on it. Users would download those, and then decide they wanted to change, enhance, or redo the application. They either joined an existing project or started their own. The sad fact IMNSHO is Zope never got that 'ready to use' application subset out there and accessible to the general masses. Most of us using ZOpe were writting bigger, complex applications for internal or customers.
This may have nothing at all to do with what prompted Chris to 'rant' about Paul to start this blog. I do agree that Zope 3 is extermely exciting and I for one am glad they are not going down the TTW paradigm (and pray they do not). If someone needs this, they should step back and look at the applications that should be written to allow the Marketing Depts. use cases to be a reality. Having paid the painful price of rewritting the ZClass product that is referenced by Chris that replaced Vignette, I would prefer that individual had the 'right' documentation and models to have 'drank the punch' more correclty. It was painful. The resulting ZClass application was crap. He personally didn't write the app I (and Jens) rewrote, but that was the model being pushed by ZC at the time, and it was wrong. Period. It might be quick; yet as Chris points out, for business the severity of the impact cannot be an option.
This leads me to my final point, which is that of 'there is no template designer'. I think there is. I think though, in reality most successfull ones are quickly turned into 'scripters' - which is by my definition, a programmer (of sorts). Chris and I were talking on Friday about some issues I'm confronting at my new job, and I'm very exicted about the model which Meld has for these people. I think there is the possibility that they 'guy' who writes my sites' look and feel has 0 knowledge of the logic that actually turns that into a end user facing web page. Meld looks bad ass for this.
I had to go off on meld, because I'm working on some ideas and was going to blog on it myself. :)
Oh, by the way....sup Paul - you owe me beers for pushing this insanity!
:)
Andrew