No in ScrapeBox when a query fails due to a bad proxy, it will retry the query again with a different proxy 3 times by default. You can raise it up to 20 retries under Settings > Adjust multi-threaded harvester proxy retries.
Also the Custom Harvester will go up to 99 retries for every query, see the "Proxy Retries" setting up the top right.
Allow me to clarify.
GScraper places the queries into a circular que. Using the default setting for Scrapebox; if P1 is dead, and P2 is dead, and P3 is dead, then an error is thrown and query is skipped. In Gscraper, if P1...P3 is dead, then Q is moved to P4...Pn. The only time that a query will be skipped is if all the proxies are dead.
Each method has its pro's and con's. Because of the transitory nature of proxies, I believe that GScrapers method is much better than the stack method that Scrapebox appears to use, especially for long scrapes that may be run over several days.
Both SB and GS suffer from the fault of being unable to pause the program to insert new proxies during the scrape. Scrapebox attempts to resolve this with the custom harvester that scrapes for new proxies every X minutes, where as Hawkes suggested work around is to set the flag to delete the used keywords when the scrapping is stopped and then insert the new proxies. Both methods are cumbersome. With SB this is because the proxy scrape and test is based off the settings and may pull >34K proxies and then test them prior to continuing. With GS this is cumbersome because you may lose the majority of a search for several keywords.
When it comes to posting, at one time GS dusted SB at ~70% to ~40% success. Now due to bugs in GS, the posting success is less than SC.
GS has no learning mode for new platforms, whereas SB does. SB wins here.
GS only searches Google, SB searches many different engines and can be taught new engines (if you have the proper settings for Baidu and Yandex, can you provide them as I can't seem to get them right). Scrapebox wins hands down.
Gscraper allows utf8 encoding so that if I want to search for botnet in Russian and Chinese I can enter ботнет and 僵尸网络 directly into the keywords. Scrapebox has no facility that allows me to do this. This would be especially useful when searching foreign (non US) countries search engines using the native language.
One of my peeves with GS is the use of the registry for the database along with attempting to disable registry tracing. Scrapebox does not do this. However, even though frowned on by MS because of the high probability of corrupting the registry, many commercial programs do this as well. I get around this problem by running Gscrapper in user space rather than administrative space. While I do not run SB in administrative space, I would feel comfortable doing so. I would not run GScraper in administrative space.
Then my biggest peeve with .net programmers (GScraper) is that because Visual Basic (.net) has garbage collection and rudimentary memory management, they think that they do not have to manage the memory. Gee, I don't have to delete an unused object and recover the space because the garbage collector will do it. This leads to excessive memory consumption and disk thrashing. I have had to force shutdown of GScraper more than once behind this. However, GScraper is not as offensive as Proxy Goblin in this regard.
Then there is another issue with GScraper, every search is sent to China. This appears to be validation of the programs authority to run, but I do not really know. This excessively consumes outgoing bandwidth and leads to the familiar no authority for the operation message. While I have not had SB long enough to see if this is the case with SB, on the surface SB appears only to validate at program start up.
Not every program is going to have everything that a person might want in it, for the purpose, at this stage of development, and within the tests I have done and the observations I have made, Scrapebox wins hands down. GScraper has gone from good to bad, and is now working on terrible.
The reason I am bringing this out is that maybe it will motivate Hawke into addressing the problems that GScraper has.
