Hi Tom,
ttweed wrote:galis wrote:My point is, the request is not unreasonable in principal or execution. The last couple notes on this represent 450% of the effort to make the change.
I appreciate your learned input on the issue, but how about the problem that even if the patches are applied, they would not convert existing links on the forum, but only apply to new postings? This seems to me to be an inconsistent, confusing and unacceptable outcome for all users, regardless of how they may have voted on the poll. I dare say with all the facts in the open now, many (including myself) would have voted differently.
How would you propose overcoming this obstacle? Yes, eventually, new postings would be added and come to dominate the postings, but old topics would occasionally be dredged up by a new comment being posted to an old thread, and users would have to be cognizant of the dates on each one to be able to be able to predict how a link is going to react to being clicked. Sometimes they would open a new tab, and sometimes they wouldn't, according to whether they were "pre" or "post" patch. To me, this would be extremely frustrating and initiate a worse situation than our current one--definitely NOT an improvement in the user experience.
(I wouldn't rule out programmatically changing the historical links but I don't think it is worth finding out how hard a retroactive change would be.)
For the inconsistency of new posts opening in new tab/windows while old posts do not, the worst case scenario I see is a user expecting the old behavior but experiencing the new. Another scenario might be a new user not expecting the old behavior and landing on an old post. I use the cmd/ctrl-click feature when I want to open several links from the page I'm on before leaving it. Another benefit of new windows/tabs on "3rd party link click" is the new page opens up on top---the new window/tab is in the foreground. (But you can still use the cmd/ctrl-click feature to open multiple tabs in the background.)
So, scenario one:
- Old user cmd/ctrl-clicks 3rd party link on a new page. Works as expected.
- Old user cmd/ctrl-clicks 3rd party link on an old page. Works as expected.
- Old user clicks 3rd party link on a new page. A new window/tab opens in the foreground, as expected, however the back button doesn't work. It would be bad if the old user turned off the computer at this point, but more likely they will close the window/tab. The former page comes into the foreground as if the back button where used.
- Old user clicks on 3rd party link on an old page. Might be disappointed that he's experiencing the old behavior but this is NOT a reason to avoid the new behavior. The old user might close the third party window/tab only to find he has to open a new window/tab and load up the forum again---typical of the old behavior.
In short a new user might experience the old behavior with the new system, if they reach an old page. The only inconsistency an old user will experience is a non working back button, in which case closing the window will reach the intended (back) location.
There may be some other regressions but do they have a consequence of confounded users? Or, anything of consequence?
If the new system is better than the old, it does not mean the old should not be changed because it would become inconsistant! Sure there *is* an inconsistency, but is the inconsistency being avoided because it is inconsistant? Or is the consequence being considered in context of the benefit of the change?
-George