Closed Bug 142171 Opened 23 years ago Closed 22 years ago

publish to file:// goes on forever

Categories

(SeaMonkey :: Composer, defect)

defect
Not set
normal

Tracking

(Not tracked)

VERIFIED FIXED
mozilla1.0.1

People

(Reporter: lasse, Assigned: cmanske)

References

Details

(Whiteboard: publish [adt2 rtm][fixed in trunk],custrtm-[verified-trunk])

Attachments

(1 file, 2 obsolete files)

When a site is set up to publish to a local or a network drive the publish dialog goes on for ever without succeding or a warning. To reproduce: 1. Set up a site to publish to a local or a network drive, like file://c:/inetpub/wwwroot 2. Create a page, and publish to that site Actual results: The publish dialog goes on for ever (well at least 10 minutes..) Expected results: Publish is successful, OR fails Preferably it should succeed, but I'll file an RFE for that..
Filed bug 142173 for publishing to file://
--> cmanske
Assignee: syd → cmanske
-->brade
Assignee: cmanske → brade
URL: file://
Status: NEW → ASSIGNED
Keywords: nsbeta1
Summary: publish to file:// goes on for ever → publish to file:// goes on forever
Whiteboard: publish
Keywords: nsbeta1nsbeta1+
Whiteboard: publish → publish [adt2 rtm]
Target Milestone: --- → mozilla1.0
Charley--I need your help with this one. The problem I am seeing is that the saving (publishing) is completed before the progress dialog comes up. Could you check if the progress dialog not being up is causing the missed progress notification? On another note, publishing a small html file (no images) to a local file does seem to have worked (my html was saved).
OS: Windows XP → All
Hardware: PC → All
I agree that the file is actually published correctly. I don't think publishing is finished before the progress dialog comes up (there might be some lag in the dialog display, but StartPublishing() is called at the end of the progress dialog startup code, so it is the dialog has been created.) What I see when I turn on the "show debug" flags is that when done, gEditorOutputProgressListener::onStateChange is called with aStatus = STATE_STOP, but the STATE_IS_NETWORK is not set. We test for both bit flags to conclude that publishing is finished, thus it just sits in the progress monitoring dialog if we don't receive that notification. I've seen this problem before with FTP publishing, though not reproduceably :-(
I believe the network flag is not to be set for local files. Perhaps we need to special case this?
*** Bug 142173 has been marked as a duplicate of this bug. ***
Keywords: mozilla1.0.1
Whiteboard: publish [adt2 rtm] → publish [adt2 rtm],custrtm-
I'll look into this.
Assignee: brade → cmanske
Status: ASSIGNED → NEW
Attached patch patch v1 (obsolete) — Splinter Review
Compensate for incomplete onStateChange messages with publishing to "file://" destination
Comment on attachment 87561 [details] [diff] [review] patch v1 r=brade but I would like to see these tiny changes: * combine the first 2 ifs in the first block * expand on comment in 2nd block to explain that it is to handle case of file with images * consistent use of { for gPersistObj assignment (other places in patch put { on new line by itself)
Attachment #87561 - Flags: review+
Attached patch patch v2 (obsolete) — Splinter Review
More comments to explain why we can't combine the 2 sections of the patch as suggested. Move the "{" in the gPersistObj creation.
Comment on attachment 87571 [details] [diff] [review] patch v2 r=brade if you combine the first 2 ifs
Attachment #87571 - Flags: review+
Attached patch patch v3Splinter Review
Better consolidation of "if" clauses and location of comments
Attachment #87561 - Attachment is obsolete: true
Attachment #87571 - Attachment is obsolete: true
Status: NEW → ASSIGNED
Whiteboard: publish [adt2 rtm],custrtm- → publish [adt2 rtm][FIX IN HAND][need sr=],custrtm-
Target Milestone: mozilla1.0 → mozilla1.0.1
Comment on attachment 87577 [details] [diff] [review] patch v3 r=brade
Attachment #87577 - Flags: review+
Comment on attachment 87577 [details] [diff] [review] patch v3 sr=alecf, but I want to see two bugs: 1) bug for necko support for writing to file:// urls 2) but for fixing this after that has been implemented, and make 2) dependent on 1)
Attachment #87577 - Flags: superreview+
Bugs filed as requested: bug 151867 to implement network fix and bug 151869 to remove the publishing code here and in nsIWebBrowserPersist.
checked into trunk Verification tests: 1. Create a new page an insert an image or edit a page with an image. Save the file locally before publishing. 2. Click on "Publish" button to launch publish dialog. 3. Use "New" to create a new Site Settings profile. Name it something like "File Protocol test." Fill in "Publishing address" with a file://... URL to enter an existing location on a local or network drive (different location than the current file.) 4. Go to Publish Tab. Fill in filename 4a. Testing just HTML file: Uncheck the "Include images..." checkbox and click Publish. Result: You should see the 'Publishing' progress dialog and the HTML filename should display in the file list with green check if file copying succeeded. Status message in dialog should be "Publishing Completed" and "All files published" 4b. Testing with image: 4b.1 Use File | Recent Pages to lod the original file in another Composer window. 4b.2 Use Publish again, select the same Sitename you created before. 4b.3 Check the "Include images..." checkbox to copy the image as well. Click Publish button. Result: Both the image and HTML filenames should appear in the Progress dialog list with green checks and same status messages as above.
Status: ASSIGNED → RESOLVED
Closed: 22 years ago
Keywords: adt1.0.1
Resolution: --- → FIXED
Whiteboard: publish [adt2 rtm][FIX IN HAND][need sr=],custrtm- → publish [adt2 rtm][fixed in trunk],custrtm-
working great and as desired on trunk 0617. Thx for the lear steps to verify this one, Charlie!
Whiteboard: publish [adt2 rtm][fixed in trunk],custrtm- → publish [adt2 rtm][fixed in trunk],custrtm-[verified-trunk]
adding adt1.0.1+. Please get drivers approval before checking into the branch.
Keywords: adt1.0.1adt1.0.1+
Comment on attachment 87577 [details] [diff] [review] patch v3 Approval granted for 1.0 branch checkin; please remove the mozilla1.0.1+ keyword when this is fixed and add the fixed 1.0.1 keyword.
Attachment #87577 - Flags: approval+
checked into mozilla1.0.1 branch
verified in 7/28 branch build
Status: RESOLVED → VERIFIED
Keywords: verified1.0.1
removing fixed1.0.1
Keywords: fixed1.0.1
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: