In principle, it would be better to use another tool to do this. For example, have a look at BioPerl and R. Vcftools is designed to parse and filter data, and the options to calculate Fst are still experimental.
Nevertheless, I like challenges :-) so here is a command that will use the GNU/parallel tool to calculate Fst using vcftools, calling it separately for each window of 7 SNPs:
grep -v '^#' ALG11.vcf | gawk '{print $3}' | parallel -N 7 -j1 "echo {} > param_{#}; vcftools --vcf ALG11.vcf --out ALG11_fst_{#} --hapmap-fst-file second_vcf_file.vcf --snps param_{#}
Explanation:
Step1:
grep -v '^#' ALG11.vcf | gawk '{print $3}'
This will extract all the SNP id in the file. Replace "ALG11.vcf" with the name of your file.
Step 2:
| parallel -N 7 -j1
This will call the GNU/parallel tool, which here is used to calculate Fst in parallel. Note that this must be the GNU/parallel tool, and not the standard parallel tool that comes installed with coreutils. If you have trouble installing it, I can rewrite the code for xargs, or for the other parallel tool, but I recommend you to install GNU/parallel.
The -N7 option is used to split into windows of 7 SNPs. If you prefer another window size, just use another number here.
Step 3:
"echo {} > param_{#}; vcftools --vcf ALG11.vcf --out ALG11_fst_{#} --hapmap-fst-file second_vcf_file.vcf --snps param_{#}
This is the argument to GNU/parallel. Basically, it saves the SNPs to a param_ file, and then call vcftools to calculate Fst. You may want to add an option to remove all the params file afterwards.
You can use the --snps option to select only a subset of SNPs. Thus, you can generate a list of all the SNPs in the two files, split them in sliding windows, and then call vcftools multiple times to calculate Fst for each window. It is a bit of overhead, but it should work..