I just went through the install of regular modx-2.0.0-beta3 and it connected to the db ok and all tests were ok but after clicking INSTALL button I just get a blank page. My install was on Fedora 10 x86_64 under xampp 1.6.3. Can someone tell me what is wrong? Did the install work?
Regards,
Gerry Reno
Ok, I had to do several things. I had to edit php.ini and raise the memory_limit to 128M and the timeout to 120. Then I had to restart xampp. Then I had to manually go into mysql and drop the database and recreate it. Then I reran the setup and all tests then succeeded and now it succeeded with the installation.
Regards,
Gerry Reno
Quote from: greno at Oct 13, 2009, 10:30 PM
Ok, I had to do several things. I had to edit php.ini and raise the memory_limit to 128M and the timeout to 120. Then I had to restart xampp. Then I had to manually go into mysql and drop the database and recreate it. Then I reran the setup and all tests then succeeded and now it succeeded with the installation.
Regards,
Gerry Reno
I’m glad you got it working.
BTW, what was memory_limit set to before you changed it?
FYI, this is mostly fixed in the upcoming beta4 - we’ve got memory_limit tests hitting around 18M max or so on installs now.
shaun mccormick | bigcommerce mgr of software engineering, former modx co-architect |
github |
splittingred.com
Quote from: splittingred at Oct 13, 2009, 11:14 PM
FYI, this is mostly fixed in the upcoming beta4 - we’ve got memory_limit tests hitting around 18M max or so on installs now.
Do you mean that it will install with memory_limit at 32M now? That would be great, especially if the setup memory tests work reliably -- they don’t seem to now. I’ve seen quite a few cases here where the install appeared to complete normally, but the person obviously had memory_limit set too low for the full install -- leading to all kinds of weird behavior. It’s a serious problem because most of them are having their first experience with MODx.
BTW, I was just going to file something on Jira because in the current rev, the comment says test for < 32M, the test is for < 64M, then we set it to 128M. People with more than 64 but less thatn 128 fall though the cracks and ultimately get less memory than the people with < 64. Is that resolved?
My memory_limit was set at 8M! And timeout was 30.
If we need to have 128M then why don’t you just set the test as: if <128M set to 128M?
Regards,
Gerry Reno