> 2) I see potential risks that you could crash a whole site if there is
> some problem with the upgrade. I personally prefer manually overseeing
> the process. Easier to fix things fast.

Looking at some IDE systems, I would recommend a system where a
testing environment is automatically created, so the result of the
upgrade can be previewed.
I imagine the following steps:

* An upgrade button is pressed
The update is downloaded into the barns folder, where we have,say, the
folder "barn-current" and the folder "barn-4.0.0" is added.
* a link to the preview site is provided
http://your.host.com/index.php?p=main&barn=4.0.0
with this link, the system uses the new barn and the site is
previewed. To be completely secure a copy of the site pages/folders
may be neccesary, but it's doable. Alternativly, one could simply
disable the write-to-disk functions such that no changes are in the
upgrade-preview mode are possible.
* a button with "accept upgrade" is pressed.
The older barn is deleted and the newer barn is renamed to barn-
current.

>
> I think some kind of external installer script is the better
> solution--probably not php. Though of course that installer script
> could be called from within BoltWire. Anyone with programming
> experience in a more appropriate language is welcome to tinker here.
> I'd be happy to help where needed.
>
> The bigger reason is our upgrade process is sometimes a bit rough. If
> we got to the point where we were really going from one stable jump to
> another stable jump, this would make more sense. Maybe when we get to
> 4.xx we can explore this some more. I really hope to slow or even stop
> development of BoltWire at that point (esp BoltWire lite). So
> occasional, more secure patches might be more appropriate for an
> automatic upgrade process.
>
> Great suggestion, I'll file it away. When BoltWire is ready I'd like
> to move that direction.
>
> Cheers,
> Dan
>
> On Fri, Mar 12, 2010 at 8:44 AM, Erlend Sogge Heggen <[email protected]> 
> wrote:
>
> > My favorite feature in Wordpress is the upgrade functionality of both
> > the core bundle as well as additional plugins. I know FTP overwriting
> > is rather straight forward and possibly safer, but if there's any 'one
> > button click' that I adore it's this one.
>
> > On Mar 7, 11:45 pm, The Editor <[email protected]> wrote:
> >> We've been knocking out some good stuff lately, and there's only one
> >> or two things left on my immediate agenda. So I took some time to
> >> revise my roadmap for the immediate future. Here's an outline of
> >> things to come:
>
> >> 1. Finish up the current 3.3.x series and put out a stable 3.4.
> >> Basically, as soon as we feel confident the currently release seems to
> >> be working.
>
> >> 2. Add a few things to the 3.4 series and finish it off with a final,
> >> stable 3.5. And make this the end of the entire 3.xx round of
> >> development. The todo's for 3.4 include:
>
> >> * Improved automatic paragraph making...  We really need an entirely
> >> different system for doing this...  Have some ideas, but haven't
> >> tinkered much with them yet. Soon...
> >> * Improved performance of searching by caching index, and a new data
> >> querying technique... I have a custom query script working on my site,
> >> that gives impressive results speed wise, and with many enhanced
> >> capabilities. I just need to integrate it into the core somehow.
>
> >> 3. For 4.xx I plan to finally do the major overhaul I've been planning
> >> for how forms work. Namely, forcing all commands to be session
> >> variables and allowing formats like this:
>
> >> [session mail to=... from=... subject=... body=... etc]
>
> >> These were listed as a 3.5 goal, but I've renumbered things, and it
> >> seems more appropriate to make this a 4.xx goal, as it could be quite
> >> disruptive, affecting virtually all existing sites and many plugins...
> >> Yet I'm convinced this will be a major improvement to things. Can't
> >> wait to do it...
>
> >> The other goal for 4.xx is to go through the code very carefully and
> >> try to simplify everything possible:  strip out all absolutely
> >> unnecessary code. Features we don't use. Options we don't need.
> >> Anything complicated. That kind of stuff. Just general house cleaning.
> >> Perhaps even get us back down to under 100k. I've been beginning to
> >> feel BoltWire is not quite as spry as it once was...
>
> >> To be honest, I'm inclined to have a BoltWire lite and a BoltWire pro.
> >> With the lite version being a simpler feature set that is rock solid
> >> stable, easy to use, and rarely updated. Just works. And then a
> >> BoltWire pro that is more full featured, with a several choice plugins
> >> built in, more actions and common stuff built in. Everything ready to
> >> go as an all-in-one CMS system. Just toying with the idea actually.
> >> Not sure I like maintaining two versions, but I like the idea of
> >> giving people a really positive, first impression. And of having
> >> something bigger with everything in it. Just thinking out loud.
>
> >> Well, open to feedback and opinions, and of course feature requests
> >> not listed above. All the usual stuff.
>
> >> Thanks to everyone for being a part of BoltWire's development.
>
> >> Cheers,
> >> Dan
>
> > --
> > You received this message because you are subscribed to the Google Groups 
> > "BoltWire" 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 
> > athttp://groups.google.com/group/boltwire?hl=en.

-- 
You received this message because you are subscribed to the Google Groups 
"BoltWire" 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/boltwire?hl=en.

Reply via email to