@aa - the reason it was designed this way was that when we first wrote it, we didn't have a way to overinstall on a live, loaded extension (we didn't have in place disabling or upgrading working). Now that we do, I think we can fix this properly.
On Fri, Oct 30, 2009 at 9:51 AM, Robert Billingslea <[email protected]> wrote: > Here is what happened to us, maybe this will convince you to reconsider: > We defined an empty page_actions array in the manifest file initially. Then > a dev build of chrome was released that did not allow this anymore (it was > previously accepted). So this broke our extension for all of our users who > upgraded. What is worse, is that apparently it no longer listed our > extension on the Extensions page (I assume because the manifest no longer > parsed correctly?), so users could not uninstall. > We didn't know of this problem with versions initially so our advice to > reinstall would not fix the problem either. Our users were basically stuck > and we had to tell them to go to the Extensions directory and manually > remove our folder. The whole saga is discussed here: > http://forums.lastpass.com/viewtopic.php?f=12&t=18409&hilit=chrome > I ran into this problem again today when working with a user that was having > a problem. I made a test build for him to fix it but he reported no change > after installing the latest version. Needless to say, the extension never > was updated. > If you do not change the behavior, you should at least report a warning to > the user. Otherwise, I guarantee this will cause problems for others in the > future. Thanks for the detailed report, and sorry for the problems here. I've filed two bugs about what you ran into: http://code.google.com/p/chromium/issues/detail?id=26537 (uninstall) http://code.google.com/p/chromium/issues/detail?id=26538 (overinstall) Erik > On Fri, Oct 30, 2009 at 1:33 PM, Aaron Boodman <[email protected]> wrote: >> >> It was designed this way intentionally, though I agree with you that >> in the case of manual installation, it is counter-intuitive. >> >> There is an edge case that I can't remember right now that we were >> concerned about that made us do it this way. Erik will remember... >> >> - a >> >> On Fri, Oct 30, 2009 at 9:27 AM, Antony Sargent <[email protected]> >> wrote: >> > I'm pretty sure we intentionally designed it this way but I can't >> > remember >> > off the top of my head what the concern was. Perhaps one of the other >> > extension devs will recall. >> > >> > On Fri, Oct 30, 2009 at 8:44 AM, Robert Billingslea >> > <[email protected]> wrote: >> >> >> >> I took a quick look through crbug.com for this issue but I didn't see >> >> an >> >> existing change request that matched. >> >> I am not sure if this was intentional, but if an extension exists with >> >> a >> >> particular version and you manually try to reinstall an updated crx >> >> (with >> >> the same version), it doesn't appear to update any of the files. >> >> There have been times (such as when the chrome dev channel updates and >> >> breaks our extension or when debugging a problem with a user) that we >> >> may >> >> want to push out a quick chrome build without a version bump (we have >> >> many >> >> products that share the same version and do not want to make new builds >> >> every week because the new chrome dev build has issues, it is much >> >> easier to >> >> tell the early chrome early adopters to redownload the crx). >> >> I think most people would expect it to replace the current version. >> >> Thanks, >> >> Bob >> >> >> > >> > >> > > >> > > > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Chromium-extensions" 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/chromium-extensions?hl=en -~----------~----~----~----~------~----~------~--~---
