Bugzilla

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both need to be pre-warmed.

Please see http://bitsup.blogspot.com/2008/11/dns-prefetching-for-firefox.html . The same argument is true for the OCSP request.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know exactly who uses Office365 and who uses outlook.com and how many use on-premise Exchange. This does not explain the numbers.

Your argument makes no sense. Why would I fight so hard to keep this simple bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate fighting with you. I argue because your proposed changes will actually concretely **break** stuff.

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both need to be pre-warmed.

Please see http://bitsup.blogspot.com/2008/11/dns-prefetching-for-firefox.html . The same argument is true for the OCSP request.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know exactly who uses Office365 and who uses outlook.com and how many use on-premise Exchange. This does not explain the numbers.

Your argument makes no sense. Why would I fight so hard to keep this simple bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate arguing with you. I argue because your proposed changes will actually concretely **break** stuff.

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both are slow, see comment 20. Both need to be pre-warmed.

Please see http://bitsup.blogspot.com/2008/11/dns-prefetching-for-firefox.html . The same argument is true for the OCSP request.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know exactly who uses Office365 and who uses outlook.com and how many use on-premise Exchange. This does not explain the numbers.

Your argument makes no sense. Why would I fight so hard to keep this simple bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate arguing with you. I argue because your proposed changes will actually concretely **break** stuff.

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both are slow, see comment 20. Both need to be pre-warmed. The actual bug 1572418 was caused by the OSCP request. DNS is one component of that, but not everything. The OSCP request was the other part. Please do not remove bug fixes.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know exactly who uses Office365 and who uses outlook.com and how many use on-premise Exchange. This does not explain the numbers.

Your argument makes no sense. Why would I fight so hard to keep this simple bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate arguing with you. I argue because your proposed changes will actually concretely **break** stuff.

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both are slow, see comment 20. Both need to be pre-warmed. The actual bug 1572418 was caused by the OSCP request. DNS is one component of that, but not everything. The OSCP request was the other part. The pre-warming of the OSCP is what fixed the bug.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know exactly who uses Office365 and who uses outlook.com and how many use on-premise Exchange. This does not explain the numbers.

Your argument makes no sense. Why would I fight so hard to keep this simple bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate arguing with you. I argue because your proposed changes will actually concretely **break** stuff.

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both are slow, see comment 20. Both need to be pre-warmed. The actual bug 1572418 was caused by the OSCP request. DNS is one component of that, but not everything. The OSCP request was the other part. The pre-warming of the OSCP is what fixed the bug.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know who uses Office365 and who uses outlook.com and how many use on-premise Exchange. Your explanation does not explain the numbers.

Your argument makes no sense. Why would I fight so hard to keep this simple bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate arguing with you. I argue because your proposed changes will actually concretely **break** stuff.

> In the context of this it might be useful to investigate whether instead of making a connection to live.thunderbird.net, instead only a connection to the DNS server could be done.

The problem is both the DNS and the OCSP. Both are slow, see comment 20. Both need to be pre-warmed. The actual bug 1572418 was caused by the OSCP request. DNS is one component of that, but not everything. The OSCP request was the other part. The pre-warming of the OSCP is what fixed the bug.

> Seems convenient to "blame" the apparent huge increase in Owl uptake on this change, and not on the fact that you just promoted Owl to being offered (at all), and offered by default for the worlds second largest email server (bug 1592258) which landed at the same time.

I told you that I know who uses Office365 and who uses outlook.com and how many use on-premise Exchange. Your explanation does not explain the numbers.

Why would I fight so hard to keep this trivial bug fix, if I didn't honestly believe this myself? Do you think that I deliberately pick fights with you? No, not at all. I hate arguing with you. I argue because your proposed changes will actually concretely **break** stuff.