Hi All,
I just updated an image that had been lying dormant for a few months. There were several package versions on a branch I didn't have in my package-cache. As I like my package-cacher to contain all intermediate versions I opened the repository browser and started selecting all the versions I was missing. As I selected each the package browser reported "downloading xxx" as expected. However, when I did an lds of my package-cache directory for the branched package nothing showed up. So I again selected each version, again got the "downloading report", again didn't see the version show up in my package-cache until I used the Browse button to open a package browser on the version. Only then did the package version show up in the ls of t6he directory. I don't understand this at all. What's going on?
_,,,^..^,,,_ best, Eliot
On Fri, Dec 02, 2022 at 12:38:09PM -0800, Eliot Miranda wrote:
Hi All,
I just updated an image that had been lying dormant for a few months.
There were several package versions on a branch I didn't have in my package-cache. As I like my package-cacher to contain all intermediate versions I opened the repository browser and started selecting all the versions I was missing. As I selected each the package browser reported "downloading xxx" as expected. However, when I did an lds of my package-cache directory for the branched package nothing showed up. So I again selected each version, again got the "downloading report", again didn't see the version show up in my package-cache until I used the Browse button to open a package browser on the version. Only then did the package version show up in the ls of t6he directory. I don't understand this at all. What's going on?
I can't guess what is going on in this case, but in my own experience I run into issues related to the Monticello caching fairly often, and as a result I am in the habit of doing "MCFileBasedRepository flushAllCaches" frequently.
I have not looked at this in any detail, but I notice that cacheing causes problems, flushing the caches fixes the problems, and I do not notice any performance difference either way.
Maybe it is time to get rid of the caching? The VM (and Squeak) are very fast these days, so I suspect that the MC caching may have outlived its usefulness.
Dave
squeak-dev@lists.squeakfoundation.org