On Sun, Jul 8, 2012 at 1:04 PM, Leslie P. Polzer <[email protected]>wrote:
> Since interest on Weblocks has picked up recently it's appropriate that I > write about what I have in mind for it. > > There are my opinions after working with Weblocks for a couple of years. > > Weblocks has a lot of bloat and too strong coupling in it, and also many > leftovers. Much of it is just not practical. > > My proposals: > > * Create small components with their own systems, e.g. weblocks-base, > weblocks-stores, weblocks-forms, weblocks-continuations... > I don't object to this, but I don't think it's very important. Sure, there's lots of stuff in the current Weblocks that I'm not using, and I suppose I could care about the memory consumed by these components, but I'm sure it's minuscule by modern standards. (There are lots of other Quicklisp systems that are being loaded as dependencies, and I'm not even sure I'm actually using all of them. This is where I'd start if I were so concerned about image size.) > * Decouple these components so that you don't have to deal with > store-dependent stuff when you want to roll your own data storage > mechanisms. > I'm not familiar with these issues since I'm happy using the CLSQL store. > * Get rid of continuation stuff. It's not a common tool, it's a tool that > has its merit in special situations but is difficult to understand for > beginners, and difficult to debug for experts. > I like using flows. I've used a number of small flows in BountyOSS. I admit, it took me a while to be completely comfortable with them, but I am now and wouldn't want to lose them. I think some of the difficulty could be avoided with better documentation. I'll write more about this soon. > * Provide sane versions of dataform and gridedit that don't depend on > stores and are easily customizable. I already have a good dataform > substitute. > I haven't yet had a chance to study form-widget, but dataform has done everything I've wanted so far. > * Get rid of Prototype and Scriptaculous in favor of JQuery. > This, I think, would be a huge improvement. I have to qualify this a little because I haven't looked closely at what Prototype and Scriptaculous are capable of; but I see how much momentum jQuery has, including a vibrant plugin ecosystem, and it does seem to be the way to go. > * Provide good template support. > Huh, that's a surprise. I thought one of the best things about Weblocks was that it doesn't use templates. Anyway, there's already 'with-html' -- what more does anyone need? > * Get rid of the test suite. Try to write a frontend-based testing > mechanism (I probably won't do this). > A frontend-based testing mechanism would be particularly great if it could be easily adapted to testing sites. I haven't done anything about automated testing for BountyOSS, but I will obviously need to at some point. (I don't think it's my biggest problem quite yet, but it's looming larger.) If anyone has any other useful ideas or is interested in helping, please > chime in. Thanks! > I have more little features and bug fixes in my GitHub fork. i'll be sending more pull requests soon. I think that if Weblocks is to become more popular, its most urgent need is much better documentation. Maybe I can help with this at some point... not that I have much time. -- Scott -- You received this message because you are subscribed to the Google Groups "weblocks" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/weblocks?hl=en.
