@Kaine it will of course have to scrape the data and proceed it (analyzing structures, sentences and so on). However no data is deleted here, it will get that, generate content and thats it.
I understand but only English can be auto-spin. If i scrape Fr article, i can't use again same sentence. That want says only use generated content, and content used must be deleted for not be used again.
I think we say the same thing but do not know how to implement it at the moment.
Oh, yes thats right...the beta was set to expire today. I totally forgot about it sorry. I should have noticed it here. However, I might extent the beta phase some more with a next update. But for all the people asking to participate in the beta-tests, not many have returned any feedback. I am however thankful for those who did. Especially for the stuff I can never test as the different languages that will get added.
Running a long time (10 minutes) - 99% CPU utilization after scrape - log updating shows things like unrelevant content, unwanted unicode, unable to extract content. What additional info do you need? Don't see a way to capture entire log.
default sources # articles = 20 no images or videos no spinning
Problem seems to be related to using too many sources.
In one case, for at least 5 minutes, CG was processing (99% cpu) with source = My Art Submit shown in window. In all test cases I ran with many sources selected CG hanged and showed Scraping 17%.
When I choose just a couple of sources, CG works. When I selected just one source = My Art Submit, got message that there was not enough content.
Do current running list of log entries pertain only to the source being processed? If true, then that might indicate a problem because in current test log, entries are rolling for several minutes so far with source = article content directory.
When the only source = article content directory, CG finishes in just a couple of seconds.
@fng in case of high cpu usage, please click help->create bugreport. keywords with quotes must exist exactly in that order on a page, no other word between. If you use no quotes, it checks for "bunk" and then for "bed"...though both words must be present.
0.43 - fix: one 100% cpu freez bug
- new: option to only use fresh source content
- new: russian language added
0.42 - new: ability to add any language you want for content generation - just click
Tools->Add new language (can take a while though)
- fix: many issues on languages fixed
- fix: generic anchors fixed
0.41 - new: added Danish as language
- new: inproved Frensh spin syntax (less but more correct, thanks to Kaine)
0.40 - new: improved some language content/spinners
- new: improved RankerX export template (thanks to Anton)
0.39 - fix: import of LibreOffice Thesaurus
- new: added support for Frnch as language (including basic spin support)
0.38 - new: spinDB tool improved
- new: improved internal spin databases for all languages
0.37 - fix: some issues in internal spin database
- new: improved SpinDB Tool
- fix: issue in RankerX export (untested)
0.36 - fix: bugs with custom local sources
0.35 - new: SpinDB Editor via Tools button (not flly working though)
- new: improved read flow when mixing sentences
- new: ability to use local files/folders as sources
0.34 - new: added Swedish as supported langauge (no spin sytax...feel free to provide it)
- fix: some problems with Italian content scraping fixed (unicode mismatch)
- new: improved some scrapers
- fix: authority link scraping with %domain% was broken
- fix: number of articles was not changable for some algorithms
@kaine please explain a bit better...the log is not showing characters correctly, but the generated content will...the number of articles are not shown when scraping as many might be removed again or detected as not useful at the later steps.
I downloaded the 5 day trial version last night and tried it out. I started it this morning and it said the trial has expired. Am i doing something wrong? Thanks
0.45 - new: added Spanish, Arabic and others as supported language
- new: added 3 new spinning services (Spanish/Portuguese only)
- new: added Indonesian as new language (no spin support though)
- new: added some more (strange) export templates
- new: improved GUI a bit
- new: improved speed for Markov algorithms
- fix: minor issues on language processing
Comments
However no data is deleted here, it will get that, generate content and thats it.
However, I might extent the beta phase some more with a next update. But for all the people asking to participate in the beta-tests, not many have returned any feedback.
I am however thankful for those who did. Especially for the stuff I can never test as the different languages that will get added.
default sources
# articles = 20
no images or videos
no spinning
keywords with quotes must exist exactly in that order on a page, no other word between. If you use no quotes, it checks for "bunk" and then for "bed"...though both words must be present.
This would help decrease duplicate content.
@710fla i will add an option for that.
At french scraping, "é, è, à ....", make problem. Like that must be exact search, seem to be not grab article.