Search wise you can use GTSpotter to search help topic, I even added that myself as a plugin to GTSpotter at some point with even the ability to preview help inside GTSpotter.
Look wise, sure pharo help isnt the prettiest thing but I would not it say it sucks, it does its job okish, it displays help text. If you want docs that are looking very good, Pillar is doing already that and it outputs in the most portable formats, html and pdf , that makes it easy to take documentation where you cannot use pharo, on tablets and smartphones. As a pharo developer I am very much against of cramming things that already exist outside pharo image and do an incredible good job that would take years if not decades for pharo to catch up , instead of putting things inside to make things harder with close to zero benefit. Lets face it the only reason to put things inside the pharo image is with the promise that we will be modifying , improving and extending them. Documentation however is a large effort by itself , so why not instead of worrying about a Pharo Documentation API , worry about the actual documentation ? Personally I am more interested in reducing the complexity of Pharo and increasing its documentation than the opposite. Funny fact: I used to be a big proponent of onboard/in-image documentation but the great work of the Pillar devs made me change my mind completely. Pillar for me documentation wise is the way forward and it already works incredible well. I would not even complain if the help tool was completely replaced with the a link to the pillar documentation , it does not contain much documentation anyway, nowhere close to the amount and quality of documentation produced by Pillar. On Mon, May 9, 2016 at 10:45 AM Sven Van Caekenberghe <[email protected]> wrote: > > > On 09 May 2016, at 09:37, Peter Uhnák <[email protected]> wrote: > > > > How about WebDoc (or similar) serving up the API with class comments to > > make it searchable on the web > > > > Although the in-image help is really bad, I would argue that for the API > itself it's quite the opposite. > > Pharo is centered around code, so there is a lot of different and > amazing ways to search the API and easily find what you are looking for > (including trivially scripted searching). > > I don't think that any web interface can compete with that (unless it's > powered by Pharo itself). > > > > In fact there used to be WebDoc for Pharo API, however as there are no > comments and examples it wasn't useful at all. You can do better search and > actually try the code directly in the image, or search for examples. > > See http://files.pharo.org/doc/ for WebDoc. > > I do totally agree with the fact that Pharo is first and foremost an > in-image IDE, we should concentrate on that. The web based documentation is > mostly relevant for search indexing and newcomers. > > > The wiki would also act as a place where the documentation can evolve > > until it is considered stable enough to include in the image. > > > > For this we have books imho. Simple documentation should be added to > image. > > > > also have links to other blogs, etc. that may not be appropriate for > > in-image documentation > > > > I don't see why not, but it would be better to just copy the content > without having to click (depends on the content). > > >
