User talk:Gnosygnu/Archives/2018/October

Xowa issues

Hello,

Let me start by saying that I really like Xowa.

Issues with Download Central: 1) In the 9/18 English image downloads there are a number of cases where the download is larger than the saved file. By this I mean a download of 1.9GB becomes 1,250,712KB (say) after decompression. Seems backwards. 2) I suspect that there is a problem with the bottom line of the dialog when downloading. I noticed this on the prior version of Xowa. In the line "download: 0% · 18:02 m · 6.9 mb / 1.13 gb" the "6.9 mb" field would be something like 1.1GB, 1.2GB, ..., 1.10GB of <blah>. I think that 1.01GB, 1.02GB (note leading zero) was meant. 3) Also on prior version I noticed that more than one download at a time was possible. More than one subfolder in tmp/bldr. Not repeatable, only noticed when connectivity was really bad. 4)Also on prior version I noticed that Xowa would restart on a different Wiki. More than one subfolder in tmp/bldr. Not repeatable, only noticed when connectivity was really bad. Might be some metadata issue.

Enhancements: 1) It would be nice if the last field in "English Wikipedia - Images (2018-07) [107.35 GB]" was updated after each download. The part between the [ and ]. 2) It would be nice if there an option to not unzip during the download. I suspect that most Xowa users are moving the data to another location (PC, flash, etc) anyway. The rest of the process (verify, move) should remain unchanged. The end user would then unzip manually on the target. It would be helpful if the file names were numbered. Maybe a prefix like "0001", "0002".

Thanks for the great tool!

73.109.60.217 (talk) 00:26, 8 October 2018 (UTC)

Xowa issues

Hello,

Let me start by saying that I really like Xowa.

Issues with Download Central:

1. In the 9/18 English image downloads there are a number of cases where the download is larger than the saved file. By this I mean a download of 1.9GB becomes 1,250,712KB (say) after decompression. Seems backwards.
2. I suspect that there is a problem with the bottom line of the dialog when downloading. I noticed this on the prior version of Xowa. In the line "download: 0% · 18:02 m · 6.9 mb / 1.13 gb" the "6.9 mb" field would be something like 1.1GB, 1.2GB, ..., 1.10GB of <blah>. I think that 1.01GB, 1.02GB (note leading zero) was meant.
3. Also on prior version I noticed that more than one download at a time was possible. More than one subfolder in tmp/bldr. Not repeatable, only noticed when connectivity was really bad.
4. Also on prior version I noticed that Xowa would restart on a different Wiki. More than one subfolder in tmp/bldr. Not repeatable, only noticed when connectivity was really bad. Might be some metadata issue.

Enhancements:

1. It would be nice if the last field in "English Wikipedia - Images (2018-07) [107.35 GB]" was updated after each download. The part between the [ and ].
2. It would be nice if there an option to not unzip during the download. I suspect that most Xowa users are moving the data to another location (PC, flash, etc) anyway. The rest of the process (verify, move) should remain unchanged. The end user would then unzip manually on the target. It would be helpful if the file names were numbered. Maybe a prefix like "0001", "0002".

Thanks for the great tool!

73.109.60.217 (talk) 00:26, 8 October 2018 (UTC)

Hi! Thanks for the interest in XOWA.

1. In the 9/18 English image downloads there are a number of cases where the download is larger than the saved file. By this I mean a download of 1.9GB becomes 1,250,712KB (say) after decompression. Seems backwards.

By any chance, do you know which files were reported incorrectly? I did a quick check now and everything seems smaller:

SELECT * FROM import_step WHERE wiki_abrv = 'en.w' AND wiki_date = '2018.07' AND import_size_raw < import_size_zip;

 2. I suspect that there is a problem with the bottom line of the dialog when downloading. I noticed this on the prior version of Xowa. In the line "download: 0% · 18:02 m · 6.9 mb / 1.13 gb" the "6.9 mb" field would be something like 1.1GB, 1.2GB, ..., 1.10GB of <blah>. I think that 1.01GB, 1.02GB (note leading zero) was meant.

The status bar shouldn't report bad sizes. I don't know why there's a 1.2 GB being reported for a 1.13 GB file. Would you happen to remember which file? Otherwise, I'll keep a look out for it later

 3. Also on prior version I noticed that more than one download at a time was possible. More than one subfolder in tmp/bldr. Not repeatable, only noticed when connectivity was really bad.

Yeah, this shouldn't happen. Let me know if you can reproduce!

4. Also on prior version I noticed that Xowa would restart on a different Wiki.

Yeah, again, this shouldn't happen. XOWA saves the order of tasks to a temp location and should replay the same steps. Reproduction steps would be helpful. :)

1. It would be nice if the last field in "English Wikipedia - Images (2018-07) [107.35 GB]" was updated after each download. The part between the [ and ].

Download Central keeps track of how many GB are left. I'd like to keep the 107.35 GB label intact as it represents the true size (while Download Central reports the size in progress)

2. It would be nice if there an option to not unzip during the download. I suspect that most Xowa users are moving the data to another location (PC, flash, etc) anyway. The rest of the process (verify, move) should remain unchanged. The end user would then unzip manually on the target. It would be helpful if the file names were numbered. Maybe a prefix like "0001", "0002".

Unfortunately, the unzipping is part of the verification. If it fails, it triggers the download again. I do agree that there are use cases where a user may want to skip unzipping in order to save disk space and just move to another machine. Unfortunately, that's not a use case that XOWA supports. Sorry.

Anyway, thanks for the helpful feedback!

gnosygnu 03:46, 8 October 2018 (UTC)

Download center missing Japanese Import

home/wiki/Special:XowaDownloadCentral

There is no selection for importing Japanese wikipedia, dictionary.... etc. Please add this function. Thanks. — Preceding unsigned comment added by 218.103.134.121 (talk) 01:31, 21 October 2018 (UTC)

Hi! Thanks for the interest in XOWA. I've been a bit short on time recently due to a change in jobs. I can try to put in some time to generate the Japanese wikis, but I could really use a tester to go through them. Let me know if you're interested, and I'll make an attempt to go through them. Thanks! gnosygnu 03:50, 23 October 2018 (UTC)