Forum OpenACS Development: Massive commit to HEAD about to happen

Collapse
Posted by Dirk Gomez on
noquote" is about to happen in the next days. Commit will be done so that we can revert in case everything falls apart.

The odds of happening are rather small. My noquote dev system can be found here: http://noquote.cph02.collaboraid.net/ and I am able to install from scratch and navigate the system without too much hassle. Please look at this system and post bug reports.

Nonetheless a lot of code needs to be fixed. It may be a good time to ask for package ownership.

Collapse
Posted by Neophytos Demetriou on
Dirk, the package ownership was brought up just recently. So, I suppose that it remains that people who have not claimed there packages yet, do so in the next couple of weeks.
Collapse
Posted by Mohan Pakkurti on
Is it possible that we can get HEAD on OACS and .LRN to a stable point and then tag it, before proceeding further? By stable I mean, that one can install and use OpenACS/.LRN from HEAD.

Thanks so much!

/Mohan

Collapse
Posted by Andrew Piskorski on
Ok, this may be obvious, but note that there is a fundamental difference between Head being "unstable" in the sense of being incomplete, buggy, and needing work, vs. Head being "unstable" in the sense that's it's so broken that you can't even install it at all. The later state makes working on and improving Head rather more difficult...

When Mohan asked for a CVS tag on a "stable point" on Head, I suspect he probably meant the first state described above - still "unstable", needing work, and incomplete, but installable, not broken.

(The rest of what everybody else said sounds fine of course, I just felt like pointing out the above...)

Collapse
Posted by Malte Sussdorff on
The reason we should apply the noquote stuff now before head get's stable is exactly the fact that

a) It is a major thing that needs to be adopted throughout the system, and better now than later.
b) Noone expects head to be stable. Adding the noquote thing will most likely not break pages that haven't been broken before.
c) The people working on noquote are (AFAIK) not people that would be involved in getting the HEAD more stable anyway.

Looking forward to see it implemented.

Collapse
Posted by Dirk Gomez on

Jeff just committed about 900 files to HEAD, please update your HEAD development systems.

Mohan, the "stable" tag is before-noquote.

I'm going to post more details on what-to-do later this night. In the meantime have a look at the terse project description over at https://openacs.org/projects/openacs/4.7/noquote.

Collapse
6: Terse noquote guidelines (response to 1)
Posted by Dirk Gomez on

Here's a terse guideline on what a package owner should done in the next weeks regarding "noquoting" his or her code:

  • Remove ad_quotehtml from tcl files. Quoting is now the default, so calling this function in a tcl file is unnecessary.
  • Quoting is now the default: add the noquote tag sparingly to adp files.
  • If an html widget is built in tcl, consider moving html generation into a include adp/tcl pair.
  • Move html generation to adp.

How do you test? Create objects with html code.

(There's an exception the first rule rule: if you have a multirow datasource where you explicitly want to quote some rows and don't want to quote the rest, you'll need ad_quotehtml in the tcl file and noquote the variable)

You currently need to noquote property and include parameters that you pass on from one adp to another one. The templating system should be changed so that quoting takes place only once. If someone wants to implement that, please speak up.

Post your questions please, I'll be lending a hand