August 10, 2008

GMail on Thunderbird: what you might not know

Filed under: google,tech — SiKing @ 12:25 pm

Google Mail is one of the most popular free (as well as paid) e-mail providers. Thunderbird is a very popular free e-mail client. In this article I will discuss what are some of the features of this duo that you might not expect or know about. I will not be discussing why you would want to chose either of these over any of their competition, nor will I be discussing why you would want to use Thunderbird over the GMail web interface. The discussion here will focus on features, not pros and cons.

The nicest thing about standards is that there are so many of them to choose from.” Ken Olsen

E-mail protocols

Let me start with some background: what you need to know in order to set everything up and to be able to follow the rest of the discussion as well as some of the provided links.

There are three main mail protocols, at least two of which must be configured so you to be able to access a mail server (such as GMail) from a desktop client (such as Thunderbird). I will give only a very simplified description, a quick overview, of each protocol; for a more detailed discussion see any number of websites on the Internet, Wikipedia being a very good starting point.

The Post Office Protocol 3rd version – POP3 – is a simple protocol used for retrieving mail to your machine. A connection between your machine and the server holding your mail is established only for the actual retrieval of mail and then disconnected. Mail is always downloaded and stored in plain-text on the local machine. This may be a potential security / privacy risk (this risk is normally reduced with user-level file access and can be even further reduced with encrypted file systems, but both of these topics are very much outside of the scope of this discussion). POP3 is a one way communication protocol, specifically from the server to your machine; mail cannot be sent from your machine to the server using the POP3 mechanism. Besides retrieving mail, the protocol also offers some simple server management commands like: delete mail after download, mark mail as read after download, etc. GMail ignores most of these commands from the client. They have instead created their own tasks that are automatically performed after the mail is retrieved. The three tasks are: keep mail in the Inbox (effectively do nothing), archive the mail, delete the mail. This is usually the protocol of choice for portable computers.

The Internet Message Access Protocol – IMAP – is a more modern protocol, compared to POP3, and certainly more feature rich. The main difference from POP3 is that your mail is always stored on the mail server, and is downloaded to your machine only for reading. Your local machine keeps only metadata about the mail – things such as message headers, message count, messages read, message location, etc. The mail is physically present on your machine only in memory and only while you are reading it, which lowers the potential security / privacy risk. However, the server and the client must be in constant contact. For this reason, IMAP is not normally a choice for portable computers. IMAP allows for two way communication; you can use the IMAP protocol to upload a mail message from your client to the server, but only for storage purposes and not for mailing! Obviously to accomplish all of this, the IMAP protocol defines significantly more commands. GMail also traps some of these commands and does their own thing – complete discussion is further below. This is often the protocol of choice for desktop computers, and especially for users that want to be able to access their mail from different computers (such as Internet cafes).

Only one protocol is available for sending mail – SMTP – the Simple Mail Transfer Protocol. It is the counterpart to POP3 in that it was developed at about the same time. Just as POP3 it only establishes a connection for the transfer of the mail, and then disconnects. This protocol is what is used on all servers all over the Internet to pass your mail from one point to the next.

Each of these protocols is handled by a piece of software on your mail provider’s computer, and is often referred to as a server. The mail provider, depending on their needs, can have any number of hardware servers (often and unfortunately also referred to as just a server) from one (a small personal mail server) to possibly even hundreds (for somebody like GMail). Each of these will usually have a unique server name. In order to keep each of the services from colliding with each other, this is especially important in the case of one server systems, each service is assigned a unique port number. This is a virtual software identifier that allows multiple services to live on one physical machine. Normally the end user is only concerned with the machine name and port number for each protocol, and this must be provided by the mail provider.

A simple world

The easiest setup the user can have is POP3 / SMTP. In this case the user will configure their mail client to receive mail from their provider from a server, for convenience, often named something like, and the default port is 110. The user then defines that outgoing mail should be sent to their provider to a server called, again for simplicity, something like, and the default port is 25. Normally all the default setting are predefined in the client software. Historically the users were given small storage quotas, usually enough to keep a few hundred mails on the server. Once the quota was reached, the user had to either delete mail or downloaded and remove it from the server – all this was managed from the mail client. Today storage space is no longer an issue and users are often given extremely high, effectively unlimited storage quotas.

The other combination that a user can have is IMAP / SMTP. In this case, the only difference is the incoming server will be probably called something like, and the port that it will be expecting by default will be 143. There is actually one more slight difference, but this is completely transparent to the user. Remember that in this case, all the mail actually lives on the server; even copies of the mail that the user sends out. When the user sends a mail, it will actually go out twice: once to the SMTP server to be mailed off to the intended party, and a second time over to the IMAP server for storage so that the user can have a local copy of the mail.

Of course the user is free to use all three servers, assuming the provider is offering all three services.

A more complicated world

In today’s world, due to privacy and security issues, almost all communication is done over secure encrypted channels. This is accomplished quite simply by another software service, installed on the same machine as each of the above discussed protocols, but using a different port. Secured POP3 connection is normally on port 995, and 465 for secure SMTP. Secure IMAP normally runs on port 993. All of the defaults, secured or not, can be changed by the provider if they chose to do so. Again, all the details must be given to you by the mail provider; they are mentioned here however so that you have some idea of what to look for when you are searching for the setting on your provider’s help site.

Thunderbird features

Thunderbird logoA list of Thunderbird features is online; the ones that are relevant to the discussion here are:

  • Support for all the mail protocols discussed above.
  • Allows grouping of mail into effectively unlimited number of folders and subfolders.
  • Support for “staring”, which is a special mark used to highlight a mail message. Staring is supported by the IMAP protocol, and thereby transferable to other machines. Remember that POP3 is a one-way protocol; if you connect to your mail server using POP3 and want to mark a message with a star, Thunderbird will allow it but the mark will not be available on the server and therefore not visible on any other machine that you use to access your mail.
  • Support for labels, which are additional marks you can assign to mails that can subsequently be used in display filters or searches. Labels do not have any kind of hierarchy. It is important to note that mail labels are not part of any of the standard protocols. The labels that you define in Thunderbird will become part of the metadata and will be stored only on the local machine. This means that even if you use the IMAP protocol (where all the mail is stored on the server), the labels will not be available there and therefore not visible on any other machine that you use to access your mail.

Updated 2008/12/21: I am beginning to suspect that I am wrong about the Thunderbird labels. I just reinstalled my kompy, and somewhere in the process I managed to screw up my backups. Once I got everything set up in TBird and downloaded all my mail through IMAP, all my old labels – mails that were marked in TBird with a particular label – were all still there! However, I still do not see any way to get at these labels through the GMail online interface.

GMail offerings

Gmail logoGMail’s list of features can be found here; again, only the ones that are relevant to this discussion are:

  • Support for all the mail protocols discussed above, even with the free accounts. Each protocol provides some additional functionality.
  • GMail has a different definition of folders. Their philosophy is that they give the user effectively unlimited disk space, they remove the ability (hassle?) for the user to sort their mail, and thereby force the user to use their search engine to find past mails. Each account comes with several “views” (as GMail calls them), each of which have a special function, that are presented to the user as folders. The user is not able to create additional views or to remove existing views. The views are:
    • Inbox – holds all incoming mail. Using only the GMail web interface, the user does not get any copy or move buttons as they might expect. Mail is moved or copied into the other views through various actions. To remove mail from the Inbox, the user can Archive the mail (see All Mail below) or Delete the mail, effectively move it to the Bin view. If you connect via POP3, you will be able to access only mail that is in this view; all the others are effectively invisible to the POP3 protocol.
    • Starred – message that have been marked with a star. As mentioned above, this mark is supported by the IMAP protocol and therefore visible to the client. The star cannot be transfered over the POP3 protocol.
    • Chats – history of discussions from Google Talk.
    • Sent Mail – stores copies of all mail sent by the user. Special note: any mail sent through is automatically copied to this view. This is a feature that significant majority of other providers do not have. If you configure your client to place a copy of each outgoing message here, GMail automatically will delete the duplicates.
    • Drafts – stores a copy of mails that you started to create but have not yet sent out.
    • All Mail – holds an apparent copy of every single mail from every single view above; the two views below are not duplicated here. I use the word apparent for several reasons. If, for example, a mail arrives into your Inbox it will also be replicated in All Mail. However, when you Delete the mail in the Inbox, the copy here is deleted as well – you do not have to explicitly go and delete this copy also. The Archive button, mentioned above, is used to remove the mail from the Inbox only and preserve the copy in the All Mail view. The link works the other way as well: if you delete a mail in the All Mail view, it will be deleted in all other views where it may exist, except for the Drafts view – it is not clear to me if this is a feature or a bug.
    • Spam – any mail that GMail determines to be a scam of some kind. It is possible to turn this feature off, although not advised; the GMail spam filter is very good. The user is also able to mark any mail in any view as Spam, in which case that mail will be moved into this view and automatically removed from all other views. The user can subsequently unmark any mail in this view, whether marked by GMail automatically or by the user, thereby moving the mail from this view into the All Mail view.
    • Bin – any mail that is scheduled for deletion; this view is emptied periodically by GMail or can be emptied by the user. It is possible for a user to subsequently undelete any mail in this view, thereby moving from this view into the All Mail view.
  • Labels are also handled differently at GMail. As was already mentioned labels are not supported by any of the protocols; GMail has an interesting workaround for this: they expose all labels as folders to the client. Also, any folders created by the client will show up as labels at GMail.

Putting it all together

When you first create a GMail account on the web, all the protocols are disabled. You need to go into your account’s Settings > Forwarding and POP3/IMAP and enable POP3 and/or IMAP; SMTP will get enabled automatically if either of the previous two is turned on. While you are there, I already mentioned that GMail ignores the commands from your client regarding what should be done with the mail on the server after it is retrieved. By default GMail will do nothing, that is, it will leave the mail in your Inbox. In all likelihood, you will want to change this to Archive all mail; this is done on the same Settings page where you activate POP3.

Thunderbird has a wizard to help you set up a GMail account. Answering all the questions will configure Thunderbird for POP3 and SMTP. If you want to also set up IMAP, you will have to do that manually. When you connect for the first time with each of these (receive messages for POP3 or IMAP, send messages for SMTP), you will be asked to enter your password each time. As I already mentioned, each of the protocols is actually a service running on some server somewhere. It is theoretically possible, although not common, that you can have a different username / password for each server. That is why Thunderbird will ask you for the password three times, even though on GMail they are actually one and the same.

When you connect to GMail using IMAP, the first thing Thunderbird does is to retrieve the list of folders. First time, you will notice a new folder appear in Thunderbird called “[Google Mail]” (without the quotes, but with the square brackets). All the views discussed above will now show up as subfolders under this, except for Inbox which will be correctly mapped to your Thunderbird Inbox. By default, Thunderbird creates a local folder called Trash. If you look at your GMail account on the web, you will see a corresponding label “[IMAP]/Trash”.

Ready to go? Not quite. Deleting mail now becomes a little tricky. All mail you retrieve from POP3 will get archived – moved to the All Mail view. If you delete anything in your POP3 account in Thunderbird, GMail will not be notified of it, even if you turned on the “Leave messages on server Until I delete them” options – remember GMail ignores all the POP3 management commands. With IMAP, it’s even worse. If you click on the Delete button in Thunderbird (remember none of the mail in IMAP is actually on your machine) the mail will just get moved to the label [IMAP]/Trash but will still remain in the All Mail view on GMail. In order to delete it, you would actually have drag the mail from [Google Mail]/All Mail to [Google Mail]/Bin. Like I said: GMail is actually pushing the fact that you never have to delete mail as a feature, and therefore they did not put a lot of thought to the delete process.

One more small item; this is independent of GMail. When you delete a mail in a POP3 account on Thunderbird, the mail is not actually deleted but only marked for deletion. This means that the mail is no longer displayed for you, but it is still stored on the disk. This has two implications. First, if you just leave it there thinking nobody will find it, you are wrong. Go to $HOME/.mozilla-thunderbird/*.default/Mail/ (on Windows machines, $HOME is going to be something like C:\Documents and Settings\<login name>\Application Data) and have a look at any of the files in a plain text editor. You will still see mails that have been apparently deleted. This could be a security / privacy issue. Second, even if you delete a mail it is not gone, you can still change your mind. You should be able to retrieve if from the disk with a plain text editor. To remove deleted mail for good, select File > Compact Folders.

Personal habits

On my laptop I have both POP3 and IMAP setup. POP3 is configured to retrieve mail on startup and every 15 minutes, IMAP is does not retrieve mail automatically at all. I use the POP3 account to retrieve / reply to mail, and the IMAP account to do any kind of maintenance – deleting mail (err: moving mail to the Bin view), labelling mail (moving mail to folders), marking mail with stars, etc.

As I already mentioned, any mail that is sent through is automatically copied into the Sent Mail view. Therefore, under Account Settings > Copies and Folders, you can turn off the “When sending messages …” thing for POP3. Under IMAP, first I set a copy of the sent mail to be placed in [Google Mail]/Sent Mail on IMAP, and then I turn this feature off. This may seem strange, but what actually happens is that Thunderbird will start to treat this folder as a proper sent message folder: the icon of the folder will change, but the bigger point is that when you are viewing the list of mails, the Sender column will be changed to the Recipient. Next, the Drafts and the Templates: I leave these as they are for POP3, however under my IMAP account I redirect both of these to their equivalents under the POP3 account. That way if I am creating either I do not have to be online (remember IMAP needs a constant connection), as my POP3 folders are actually physical folders on the machine’s disk.

Also, I turn off the junk mail feature in Thunderbird for the IMAP account, but leave it for the POP3 account. Any junk mail filtering logic will not be passed from Thunderbird to GMail. Having Thunderbird move junk mail round in the IMAP account will only replicate the junk. But if anything gets past the GMail filters (which are quite good), it will get caught by Thunderbird in my POP3 account, without polluting the server with copies of the junk mail. To mark this mail properly as junk at GMail, unfortunately can only be done using the web interface.

If you have more than one IMAP server defined and you try to move messages between the servers, GMail will allow you to only copy not move (that is delete) from GMail to somewhere else. This applies even if you explicitly use the Move function of Thunderbird, or if you use a mouse drag and drop. However, if you have two different GMail accounts, both setup on your machine as IMAP, then in this case GMail will allow a move from one account to another.

There is one last item of interest. I use to keep an archive of all my mail from all past sources on a CD. Over the years I changed my mind a few times on exactly how best to accomplish this. When creating the archives, every once in a while I accidentally duplicated some mails (and once in a while delete some, but GMail will not help me with that now). When I opened my GMail account, which has over 6GB of disk space, I decided to upload all of my old mail archives. Import the mails into Thunderbird as fictitious POP3 account, create the appropriate folders on GMail, and Move the mail. I discovered two interesting things. First that GMail actually filters out (deletes) duplicate mails in any view. Second is that if try to drag and drop the mails to GMail, and something happens with the connection, Thunderbird will not retry it. However, if you use the Move feature in Thunderbird, it actually has a retry, and I was able to move much larger blocks of mail that way.


May 15, 2008

Tell me I’m wrong

Filed under: google,windows,wp,yahoo — SiKing @ 10:52 pm

There are people who would describe me as a Microsoft-hater. I don’t actually hate their technology, and I believe there’s even a place for it. I am a software tester by trade, and all software is flawed and buggy, it’s only a matter of degrees. What I don’t like about MS is their marketing strategy and their stance on many other things that matter. Right from the start little Bill let the world know that he is in it for the money! MS has had its share of dumb luck over the years. But when they realized that the free-loading, commie loving, open sourcie guys actually had something, they opted for all out war. They used some down right evil strategies over the years and are still doing it today, and in doing so they made the world a worse place to be. In the end, when they finally discover they can’t beat them they decide to join them? Gimme a break!

So why a WP blog, why not a Yahoo! 360 blog? I mean I have been a long-time user of Y! Whenever I needed anything, I would usually turn to Y! as my first choice. My first blog was a 360 blog. I have been a fan of Y! search, even when everyone else around me was using HotBot and AltaVista. I have had GeoCities websites before (and after) the buyout. I was the first in line to sign up for Y! e-mail. And there were other services that I tried and even actively used over the years. The problem is that Y!, IMHO, has a nasty habit of taking a good idea and totally crippling it to the point of uselessness. Are you allowed to have non-Y! friends? What about something slightly simpler, like any visitor (even a non-Y! one) being able to leave a comment on your 360 blog? Once they realize the execution of the idea is shit they completely abandon the product and just let it bit-rot. Has anyone today heard of Y! briefcase? Unfortunately, the most useful (and still documented) feature is now turned off and has been for a very long time. How long has Y! 360 been in beta? The list is endless. I’m not even going to get into the recent public affair between Y! and MS. The one single thing that speaks for itself though, is the Yodel Anecdotal: the Yahoo! Corporate blog. Once the page loads up, check the source (somewhere under the View menu of your browser), and look at line 9 – the one that starts with <meta name="generator". Hey, what’s good enough for Jerry is good enough for me! 🙂

Blogger is in a different category … for now. I will admit that my choice between Blogger and WordPress was based solely on this Slashdot article – admittely, not a very informed choice. I already use some of the Google’s other services (mail, search), and there are some that I am afraid suffer from bit-rot (browser sync). But overall, I get the impression that Google is everything the above two are not.

August 31, 2007

Personal Computing, served up BSD Style!

Filed under: bsd,google,thetao — SiKing @ 10:17 am

First up in my BSD review I tackled PC-BSD, version 1.3.01. Actually, it was an article specifically about this distro that got me thinking about this whole BSD trial run thing in the first place.

First impressions: serious looking website geared toward the user, no torrent download, several download mirrors, derived from FreeBSD, discussion fora seem to be active.

The installer

The base distro comes on one CD, with at least one optional CD full of additional packages. Download, burn, boot: the installer was a breeze – just a few, straightforward, obvious questions. Auto-detect correctly detected my video resolution; my video card is slightly odd – it needs a lower refresh rate to get the maximum resolution that I prefer – which causes problems for some auto-detecting programs. The installer even has an option for advanced stuff, none of which was overwhelming. One of the advanced options is to turn off the firewall; since I am already behind a firewall, I disabled it, but again, very straightforward regarding what you want to allow or not.

5 minutes worth of questions, 30 minutes worth of install, one reboot, and I am greeted with a familiar looking KDE desktop. Hey, is it just me or does the default window dressing look like Windows Media Player series 10 – naw, p’wably just my imagination acting up.

I did encounter one small issue. I have two komps hooked up through a KVM switch. If I started the boot switched to the other komp while the auto-detect was going on, it had problems finding my mouse. This required me to reboot the computer, and watch the auto-detection in order to have a fully functional mouse. Not a big deal, but since this problem persists even after the install, it is more than a minor nuisance.

Anyway, top marks for the installer!

Package management

PC-BSD has three (documented) options for installing new applications: PBI Installer (the easy and recommended way), Packages (the quick FreeBSD way), and Ports (the traditional FreeBSD way).

You download a PBI from their website. You then run it on your end, and a wizard guides you through the whole install process. According the PC-BSD website: “Since PBI programs are created from traditional FreeBSD ports and packages, they are fewer and less up to date than ports.” I could not get an exact count, but there are certainly several hundred PBIs, maybe even a thousand available. Most of the stuff that a casual user would want is available through the PBIs. There is also a graphical manager that keeps track of all your installed PBIs, through which you can easily uninstall (but not update) any of them.

Installing packages is done through the command line, and the packages come from the FreeBSD Ports server, which (at the time of this writing) holds over 17,500 packages. FreeBSD Ports is what inspired the Gentoo/Portage. They are both systems where you can tell it what you want, and it will download all the sources it needs (keeping track of dependencies) from the Internet, and build the requested application from scratch right on your system.

The Ports and Packages are two different ways of getting at the same thing. Ports keeps track of all available packages right on your system; the synchronization of this takes quite a while, especially the first time. With Packages, you have to know a little more of what you are doing, as it does not keep track of dependencies for you. All software that comes pre-installed with PC-BSD is managed as Ports.

A casual user should be scared away from the system by this. A geek (like me) will probably like the Ports system.

Software support

PC-BSD does not come with either of the Mozilla packages, relying instead on the KDE equivalent: Konqueror and Kmail. I downloaded the Firefox and Thunderbird PBIs and both installed without a problem. Installation required the root password, as it should. I even installed the Browser Sync plugin (which I now somehow coaxed into working on the first try). Everything seems to be working fine. All good stuff so far. Oddly enough, I noticed some differences in rendering of pages between FF/BSD and FF/anything else. Not sure what to make of this.

While looking for the Firefox PBI, I noticed there is also a Flash plugin PBI. While not part of my stated requirements, I thought I would try it out. The package reported that it installed properly, however it failed to work. Afterwards I found out there is a post-install procedure that apparently fixes this. Negative point as a home desktop. PBI also installed without a problem, and runs just fine – I am writing this entry in it. For some reason (I am sure there must be a reason), the “Install new dictionaries” wizard is missing?!?! I know how to hack up OOo to manually accept new dictionaries, but I anticipate problems. Negative point as a home desktop.

As for my other required applications, I do not have a preference for those two, so I will go with the available default: Kopete, and Kaffeine or Kaboodle. Again, Kaffeine requires a post-install procedure to get all the media codecs into the system; unfortunately, the codecs PBI that is linked from the docs is not available. So Kaffeine disappointed, but Kaboodle saved the day. When downloading an audio file, FF crashes every time it tries to launch Kaffeine; haven’t bothered to try and convince it to launch Kaboodle. More negative points as a home desktop.

The distribution was missing the GNU-awk. Now admittedly the casual user will never, never, ever notice this or even need this. Unfortunately, I need specifically this flavour of the tool for my work. No big deal, I can test run the Ports. I used the graphical front-end to sync my Ports tree, which first time takes quite a while. After that, it’s command line only with no problems.

Even thought I was able to write this blog entry on the system, I was not able to publish it from this system! Neither Konqueror nor Firefox were able to display the Yahoo! blog composition page correctly. For comparison’s sake, the page displays perfectly fine under both FF/Windows and FF/Linux.

Software support is good enough for a workstation, but I would have definite reservations for a home machine.

Browsing shares

Browsing Windows shares was accomplished using Samba in Konqueror, without any problems. There was no setup required, it worked out of the box – major points! Mental note: still have to find where to store passwords in Samba, like I could in LISa.

External devices

When I connected my Seagate USB drive only the first (ext3) partition was found, however, the system failed to mount this partition.

My Nokia faired a little better. The system did find it, and it did mount it properly. I was able to both browse the (fat32) file system, and store / retrieve files to / from it. Unfortunately, the system was not able to unmount it properly. When I tried the “Safely remove” option, the icon quickly flashed as if it did get unmounted but it immediately came back.

As I stated before, this is no a KO criteria, but negative points in this area.


Added my printer as a Network TCP printer, using the KDE wizard. Afterwards I could print, however, single-sided only. The printer does have a double-sided capability, which is normally accessible from Linux. However, good enough.

Brownie points

I have to admit that I am slightly biased toward the KDE desktop, so that choice by the PC-BSD team definitely resonated well with me. The default window theme is nice, but I would be making personal changes to the dressing as well as the behaviour if this becomes my first choice…

PC-BSD does not do that whole sudo thing. I don’t know if this is a BSD-wide choice (will see with the other distros), or just a PC- choice. This means that if you want to perform tasks with elevated privileges, you actually have to become root. This is neither positive or negative for me, I am just mentioning it for the sake of completeness.

Overall, I could see myself getting work done on this system, but it will not be able obsolete my Windows box at work, and I will not be letting it anywhere near my home machine.

April 20, 2007

Browser Sync FF plugin

Filed under: google — SiKing @ 11:28 am

Dear Google,

I thought I would pass on my experience with your Browser Sync plugin … in case someone is actually listening.

When I first heard about the plugin here, I thought my wishes have been answered. I have been dreaming about something like this since like Mozilla 4.0!

I installed the plugin on my work machine (which I consider to be my primary in this case) first. Everything seemed to have gone fine. Next I installed it at home, where I am absolutely certain the bookmarks do not match with my work machine. Nothing happened. Well, I thought the plugin must somehow keep track of dates, or whatever. So I exited FF, deleted the bookmarks, and opened up FF again. Nothing. Since I had no idea _when_ the plugin synchronizes (continuously, upon startup, upon exit, …) I tried exiting and restarting FF several times. Still nothing. Over the next several weeks I tried various other things (both at work and at home) to coax the plugin to synchronize the bookmarks, with no apparent results of any kind.

Well today, after approximately a month, the plugin suddenly “woke up” and completely messed up my bookmarks on the primary machine – I have no idea where it got this bookmarks file from. Most (but not all!) of my bookmarks seem to be there, but scattered in random folders, plus a few extra bookmarks that I have never bookmarked before. Thank ghod that for some unknown reason reason just yesterday I decided to backup all my bookmarks.

Apparently I am not the only one with issues with this plugin.

This could have been, in my opinion, the second most useful plugin for FF, if only it would work. Today I am uninstalling it.

In my opinion, the plugin needs to have two separate user-accessible functions: “push” to force complete overwrite of the bookmarks file on the server, and “pull” to force a complete overwrite of the bookmarks on the client side. You now have one button that forces overwrite, but it is not apparent (at lest to me) if this is overwrite of the server or client.

Just thought I would mention another one that I have been dreaming of: “Mailer Sync” – synchronize my address book from Thunderbird. But, again, only if it works…


Updated 2008/06/10: I mailed this in to Google. I now honestly cannot remember if I also posted this on their discussion group, well it showed up there anyway.

Create a free website or blog at