Apache OpenOffice (AOO) Bugzilla – Issue 67371
WW8: doc files exported from OOo 2.0.3 are now 68k instead of 8k with older builds
Last modified: 2013-01-07 16:08:02 UTC
OpenOffice.org 2.0.3 at least on Linux/PC creates large .doc files (when compared with OOo 2.0.2 or 1.1.x). Create a new empty text document and save it as .doc (MS Word 97/2000/XP). It is 96256B large when compared with 8KB from 2.0.2 or 1.1.0.
MRU->MBA: exported doc are now much larger than before. an empty doc is now of 68k size. Maybe another side effect of the "custom-doc-properties" implementation by DR?
*** Issue 67476 has been marked as a duplicate of this issue. ***
*** Issue 67642 has been marked as a duplicate of this issue. ***
It seems so. The SummaryInformation stream is huge. Daniel, please have a look
I don't know what exactly goes wrong but possibly the files are broken in some way. Perhaps we create problems for other applications loading those files? We should investigate this. So for the moment 2.0.4 target seems appropriate to me.
The reason is the export of the thumbnail property, this has been added before my reimplementation by SJ with issue 63158 (CWS impress89).
back to QA
OK, it is due to the thumbnail preview which is now also saved in that format.
I believe this might still be a bug. (See attached). * Saved document with single character 'a', with Preview, using Word2003 01.Word2003.doc = 24,064 bytes * Opened above and resaved (immediately- no changes) using OOo 2.0.3 02.ResavedOOo203.doc = 62,976 bytes Both .doc files should have preview images saved, although the one produced by OOo is ~250% larger. The Thumbnail in png format is only 467 bytes! Furthermore, when browsing files using Explorer on WinXP, only the file created by Word actually displays a preview thumbnail- the OOo created .doc *doesn't* display a preview thumbnail... Regards, Andrew
Created attachment 39319 [details] Example of Word and OOo203 output
*** Issue 69789 has been marked as a duplicate of this issue. ***
From Issue 69789 comment 5: "I have tried this as well. On FC4 using OOo Milestone 8, the files I created showed the following. 103936 Sep 22 07:14 File_size_test.doc 6916 Sep 22 07:14 File_size_test.odt A single capital "A" was saved. With a 10 page document of 20,071 characters. No real formatting. 157184 Sep 22 07:23 BadBlockHowTo_test.doc 16662 Sep 22 07:23 BadBlockHowTo_test.odt When I looked at the *.doc files using hexedit, there were a large number of locations just filled with FF's or 00's. Comparing to an original *.doc file created on Windows, there were no large FF's or 00's ranges." I think that's the cause of the problem. Word files saved with OOo have a lot of locations filled with null values. The same files saved with MS Word haven't all those nulls.
The users are right; I think there is some space for optimizing the thumbnail export a little bit.
Adding myself to the CC list. I see Troodon cp/pasted my comments from 69789. Thank you.
This problem would cost me several megabytes of wasted disk space every week if I continued to use OOo 2.0.3. Since the Target Milestone has been set to "OOo later" it could conceivably be years before being fixed. In the meantime, is there any way to disable saving with the preview thumbnail or do I have to go back to OOo 2.0.2? I don't need the preview thumbnail for a text document but I do need the wasted disk space. The preview thumbnails aren't displayed anyway in either system that I use (OOo 2.0.3 on Windows XP and Windows ME).
In ooo 2.0.4 built with ooo-build on linux still persists this problem. Any way to solve this big file ? An empty file saved in .doc results in about 103kb on my system. Rgds Saxa
*** Issue 71099 has been marked as a duplicate of this issue. ***
For 67361 (Word, clicking into "Target milestone" above does not tell us when is "OOo Later" and there is a misspelling in the page "isseus". Since compatibility with Microsoft and its import/export features are an attraction of OpenOffice, we expect this problem to be fixed soon. Similar to the complainer of 67361 (Word), and of 71849 (Excel), and 71915 (PowerPoint), I will be forced to go back to OpenOffice 2.0.1 (the previous official release in Chinese-Traditional-chars) if it will not be fixed "soon". Thanks. Qiyao
*** Issue 65586 has been marked as a duplicate of this issue. ***
*** Issue 71849 has been marked as a duplicate of this issue. ***
Sorry, what I said of 67361 (Word) should be 67371 (Word).
According to the above conversation, this bug is because of this feature: << OK, it is due to the thumbnail preview which is now also saved in that format. >> However, Microsoft Office does not have this problem (super-big file). And this bug causes memory waste (diskspace) and is still marked as "OOo later". This bug already have many duplicates independently found by many netters. So could you tell what is the plan for fixing this bug in a "soon" version of OpenOffice? Thanks. Qiyao
I hope I'll have enough time to fix this issue for OOo 2.3. Be it as it may if someone can supply a patch this problem it will ensure that this issue is solved even faster.
This issue is not restricted to Word files but a general MS Export problem. As it is due to a problem in image generation I change the component to "drawing".
just adding myself as a watcher. I too will be forced back to 2.02 until this is fixed.
Can somebody tell where to start look at ? Rgds Saxa
I do not have enough time to fix this issue for OOo2.3, so I change the target to OOo2.4. @saxa, the code that is creating the preview can be found in sfx2/source/doc/objcont.cxx GDIMetaFile* SfxObjectShell::GetPreviewMetaFile( sal_Bool bFullContent ) const
Sven, please make sure that your plans enable us to integrate the fix for 2.4.
*** Issue 81630 has been marked as a duplicate of this issue. ***
Same here. I am trying to make a file less than 100K and no avail :(
@sj thx for the tip. I will try to look into it. Rgds Saxa
*** Issue 83252 has been marked as a duplicate of this issue. ***
I am sorry, I have to retarget this issue to OOo2.4
*** Issue 88157 has been marked as a duplicate of this issue. ***
This issue has been fixed now in cws[impress141]. Calc & Writer will no longer create preview bitmaps when saving to ms, when saving impress to ppt everything stays unchanged. This is the same behavior as it is done in Office 2007. However, if anybody is unhappy with this change, creating previews can still be enabled/disabled via configuration. In "Office.Common/Filter/Microsoft/Export" following three attributes determine if preview creation is enabled: "EnablePowerPointPreview", "EnableCalcPreview" and "EnableWordPreview". There is no UI planned to enable/disable these settings.
sj->wg: This issue is ready to be verified in cws[impress141]
Reassigned. Please verifiy.
CGU: Verified in cws impress141
CGU: Integrated in dev300m22
Could you tell me where to find "Office.Common/Filter/Microsoft/Export" please?