|
|
I just came back from the vacations to discover that my website is now a mess.
All the delicious mindmap data have been deleted, and so did the MySQL data of the wiki.
I am not really sure how it could happen, but I am just not going to investigate.
I am, more pragmatically, moving my website to another host. It will take some days, and a lot of the data will in any case be lost.
Please if you came to my page looking for any of the previous services: the delicious mindmap, the wiki aggregator, have patience.
As soon as the transfer have been made, I shall post a new message.
So if this current message is the last one, you know that we are still using the old host.
Pietro
So, here is the quest. You need to discover the password of your best friend. Although this might seem morally reprehensible you are moved by nobel and deep reasons. You need to check in his email and discover if he is having an affair with your wife. You can’t just confront him if you are not sure. Nor can you run the risk of confronting your wife before you are sure.
We will assume that you have access to your best mate computer. After all he trusts you, right. And with reasons, you would never do anything against him… normally. But this time it’s different. If he has been seducing your angel, he deserves the worse.
We will also assume that your best friend is really into security, and since firefox is a more secure system, he uses firefox. Daily. And as a final assumption we shall assume that since he keeps his mail on the web, and accesses it through the browser. Let’s assume it’s on Yahoo.
This is what you need to do:
- Get his computer, in a moment he is out for launch (with your lady, bastard!), and lock yourself for the work. You will need privacy for this. If you are confronted you can always say Continue reading Reading your mate passwords in Firefox
Did some more debugging. Now any unicode the user used in the tags should be ok. Still there is a big brick wall in terms of memory usage. And some users are not having any luck just out of the fact that their map is taking so much resources that it goes beyond the ISP limit. I could work hard and distribute the whole calculation so that all variables are stored on disk, so the memory would never be hit, but honestly, it is not my top priority at the moment. I am here to help those users run the program on their own machine. And eventually we might solve that problem too. So, what are my top priorities:
- Add an rss feed.I would like to add an rss feed that every time a new map is done, the feed gets updated. It wouldn’t just tell the name but all sorts of data, like the list of the Main Tags. So the users could see if they might be interested in checking the new buddy’s map
- Insert a way for user to delete their own maps. If I am going to go into hosting business, I am not going to be one of those hosts where you can add info, but you cannot delete it. I am aware that users info ultimately is adding value to my site, as such I want users to be happy in having their map here. Not forced.
- Insert a general log of all the maps that are being started, and ended. Right now such a log is absent, and there are about 200 maps completed, and more than twice maps that have been started. So about 300 have been dropped. I bet many of those users would have success, if they tried right now, after those 3 deubugging session. Still I want something that tells me: Warning warning warning, map dropped. Bug? OutOfMemoryError?
- Add the number of posts inside a tag. Just obvious
- Probably add some of the MainTags as keywords to each single map. The problem is: which? All is too much. All the ones that contain more than x posts, y subtags is not flexible enough. The solution should be: if a MainTag is part of a ParetoFront of Delicious than the keyword should be there. The fact that this means writing a whole program that stores in a database the latest ParetoFront is just a small detail
 . And before you ask: no, I will not need anybody’s password to do that, and the data will all be public.
- Add a bookmarklet to save a map in your own delicious, with the keywords as tags
- Change the map, so that it can run on a single tag. Useful for big complex maps like mine, and others.
- Make it change the Title of the Map Page, to show the owner of the map. Useful if people want to add the maps to their delicious pages.
And then there are some tests I would like to make, like:
- Check if it would make sense to show all the tags that appear with a single tag, and not the subtags.
There is more? If you can think of other modifications , please drop a line in the comment section. Also if you tried to run the map maker and it is not giving you satisfaction let me know. I’ll whip it appropriatly. HarHarHar. (I’ve always wanted to say that!)
Some people (few) were in the unfortunate situation that the tool would calculate their map, and would correctly add it to the make map page, but then the map could not be open. If you were one of those people, I have good news. I tracked down the bug (this time only derived by my stupidity) and nailed it. So, please try again. Insert again your data, calculate it again, and then open it. With this I ended debugging the obvious big errors. If you try now and the tool does not work, please drop me a line with your username, maybe send me by email (available via my homepage) your complete list of all the posts. And I will see what I can do for you. If you don’t contact me I have no way to know that the tool failed, and I will not be able to help you.
Things are progressing, more and more people are using the tool. Unfortunately not for all was a succesful experience. I could spot two separate bugs. In the first case the map would not be created at all, and the program would stop just after making the poststotag dictionary. In the second the map would be created but it was unreadable from the user. Yesterday evening (in my camper van!) I debugged the first issue. Essentially the program was downloading from delicious two different files, the (don’t click on it) list of all posts, and the list of all tags. Well, the two files were not coherent one with the other, and the list of all tags would in some rare cases list tags that had no post associated with them. Of course as soon as the dictionary would start being created the program would protest, and quite correctly so. I think the problem has something to do with how del.icio.us is at the moment handling the change tag name function. Maybe the problem has been solved by now, and what I got into were some users that had used the function while it was still not completely bug free.
In any case I circumvented the problem by not downloading the tag list file at all, but recovering the list of tags directly from the posts/all page. It is obviously slower (by big map moved from 377 sec to 460 sec.) but more secure.
So, if you tried to use the map before, and you did not had luch luck. If it did not create the map at all, then try again, and now it should work. And if it doesn’t please contact me, and maybe send me an email with your all.xml file.
If instead the map was created, your name was added but the map wan unopenable, then keep having patience, and this evening I hopefully will kill that bug too.
And thanks to all who are using the tool, is such an interesting project for me!
Pietro
I am getting really annoyed. I use Firefox, and at the moment it is quite useless on a number of sites. As they never stop reloading.
In particular (it seem to me) on all the sites that have a reload. Two among all: bloglines (the left bar). And the home page of La Repubblica. In the first page, according to its source code, it should reload every 3600 seconds. That is every hour. Quite an honest command for a feed reader. On the second every 300 seconds, 5 minutes. Right for a news site.
Instead they download, then the first waits 7 seconds (I checked it) and start download again in a loop. And since blogline is quite slow, this makes everything slow down. The second downloads the page, and then does it again and again without ever stopping.
It is as if he counts the tenth of second instead of the seconds. Then 5 minutes would be 1 second, and 3600 would be 6 seconds.
Now, if I open IE all this does not happen.
Strangely enough this ‘bug’ does not happen only with firefox. I remember a cd burner software that use to reload the page on and on. SO much that I had to disinstal it and find a new one.
I am using Microsoft windows 2000 with service pack 4. I installed every possible service pack and extension available from microsoft.
Any suggestion on what could it be?
I would like not to have to reinstall everything (windows), although I suspect if I do I would solve the problem.
Addendum, June 2005, 6 month later: I showed my problem to my friend who is visiting, he adressed me to adSubtract, which in the PRO version gives the possibility of stripping html code of incoming pages from the reload command. Now everything works fine, and is quite funny to see those commercial pages, reduced to their bare essential. Thanks to both.
|
|
Recent Comments