Technology Bites recently brought my attention to two different online PDF creation tools: PrimoOnline and Acrobat.com.  To choose between them I took them for a quick test drive.

Acrobat.com is by far the more appealing site of the two.  The site usability is as polished as you would expect from an Adobe product, despite its beta status.  It also provides storage space for your documents and the ability to share them with others.  Its achilles heel would appear to be the 5 PDF conversion limit quoted in both the Technology Bites post and the Acrobat.com user guide.  Whether by accident or design this limit is not enforced at this time.  During testing I converted 6 documents in 15 minutes with no issues.

By comparison the PrimoOnline interface is spartan but functional.  It provides a basic web form to specify the file to convert and the email address to send it to, and quickly sends the result to your inbox.  Its claim to fame is having 300 formats it can convert into PDF.  Upon reviewing the list most would be readable by MS Office applications so I’m not sure how useful this will be for most people in practice.

The proof of the pudding is in the tasting, as they say.  I ran the same test document through both services (Acrobat.com output and PrimoOnline output) and here are my observations about the results:

• PrimoOnline does not honour the fonts used in the source document.  If you compare the test documents you will find that Acrobat.com’s output is near identical to the input document, while PrimoOnline has substituted fonts and looks noticeably different.
• As expected neither service coped with the Enhanced MetaFile (EMF) included in the document.  For some reason the most space efficient way to copy Visio objects into Word documents is also the most difficult to convert into PDF form.  The Device Independent Bitmap version of the diagram converted successfully on both services and demonstrates how the EMF version should have looked.
• Acrobat.com consistently output smaller files than PrimoOnline.  It was also the only service to include a Bookmarks list, though both services do generate clickable hyperlinks in their output.  This feature is one of the main advantages these services have over a PDF printer based solution.

Based upon the test results Acrobat.com is a clear winner, provided that Adobe continue to ignore (or remove) their quoted 5 conversion limit.  Given its difficulties maintaining document formatting I’d find it hard to recommend PrimoOnline unless you were purely looking for a universal document converter so aesthetics were less critical.

Via Technology Bites

Availion runs a service to help you share your availability information – not your calendar – with others.  This is quite a powerful concept because its often more appropriate for others to know when you have time in your schedule rather than what you have in your schedule.  e.g. If I want to talk to two vendors about the same piece of work, if they can see each other’s meetings in my schedule they might focus on highlighting the other vendor’s weaknesses instead of addressing the request I’ve placed before them.

The service has quite a promising list of features and calendar support.  The key ones that caught my attention:

• Automatic – No Synchronization
• Email Your Availability
• Use With Multiple Calendars
• Supports Microsoft Outlook
• Supports Lotus Notes
• Supports Google Calendar
• Supports iCalendar

The two most compelling items on this list are Lotus Notes and Google Calendar.  This is the only tool I’ve found that states it supports publishing Lotus Notes free/busy information, and in the past Google Calendar has ignored any attempts to import internet free/busy information so I was curious whether it had resolved this issue.

The site provides only limited information about the product capabilities so I dived right in, signed up for a free account and downloaded the ~5Mb installer to try it out.

After installation I opened the program and saw a fairly basic looking calendaring application.  Stepping through the menu options I quickly came to the realisation this would be a proprietary service requiring others to download the GUI  if they wanted to see my availability.  Either that or I use the email facility to send snapshots of my schedule to others, which negates the "schedule meetings without my intervention" benefit that makes a free/busy service valuable in the first place.  Strike number one: I had been hoping for a service that would publish an ifb or vfb data file so calendar clients could directly lookup my availability when scheduling meetings without installing extra software.

Next I tried to add my Lotus Notes 6.5 calendar to the GUI.  Strike number 2: it didn’t matter whether I used the local replica or server options when adding my calendar, it always returned the same "unable to read availability from Lotus Notes" error.  If I shutdown the Notes client I saw a slightly different message telling me to make sure Notes was running, so I made sure the client was running and I had my File\Security\User Security "Don’t prompt for a password from other Notes-based programs (reduces security)" preference checked, but that didn’t help either.

Strike number 3?  The complete set of Lotus Notes configuration documentation is a single page with a few screenshots and scant commentary.  It describes only part of the configuration process and didn’t help with my particular situation.  No indication of Lotus Notes client versions supported.  Using the application’s help facility referred me back to the same page.  No general help information, FAQ or forum is available.  Any help with this problem requires me to talk to their support staff but strangely enough they have yet to respond to my request given its Sunday today.

This service promised a lot but failed to deliver.  Even if the Lotus Notes functionality had worked the proprietary nature of the service would have kept me looking for a better, more open option that didn’t require my clients to install yet another piece of software on their PCs.  My search for schedule synchronisation nirvana continues.

(Via CalendarReview)

In Can I Get My Blog Noticed Without Digg’s Help? I asked whether social sites were indeed the "silver bullet" to quickly gain a large and lasting audience for a fledgling blog.  To test the theory I proposed using my own blog to measure the impact produced by different techniques.  As part of this experiment I have outlined how I plan to track changes in a blog’s popularity and how I’m planning to use active social bookmarking to promote my blog.  Unfortunately I’ve discovered that services for tracking a social site popularity metrics aren’t quite as reliable as I had hoped.

What is Wrong With This Picture?

To illustrate my point I’ll analyse results from services I have been using to track blog performance.

popuri.us provides an overview of a website’s traffic, search engine position, and social site position.  Most of the results seem plausible at first glance, but upon further investigation:

• Live Search backlinks don’t match up with results obtained by the Webmaster Live
• Technorati Links appears to be broken, when it should be returning a value of 0 instead.  A number of other entries also return N/A values instead of the 0, and may indicate this metric can no longer be retrieved.
• Overall the results are for a single URL within a site, where as the results would be more useful if they could be returned for the site as a whole.

Socialmeter  provides visibility of the social site uptake of your blog, with some additional search engine metrics to back those up.  The results are problematic:

• Delicious has not reflected the bookmark I know exists for the main website and individual posts within it.  The same comment also applies to Furl and Spurl.
• At least one post on the site has been Dugg before so the figures must only reflect a single URL not the whole site.  Metrics for the whole site would be more useful for my purposes.
• As previously mentioned the Sphere metric is not an indicator of popularity because it counts posts that might be considered to compete with yours, rather than reflecting metrics for your content.  It is almost be an inverse indicator of popularity as it shows how difficult it will be to find your site amongst all of the related content on the web.

Blog Influence provides a subset of the metrics available via popuri.us.  As you can see the Technorati metrics are no longer operational which doesn’t inspire a lot of confidence.  The Yahoo result is interesting but its hard to ascertain exactly which information it is retrieving from the site.  It has been steadily rising over the course of the experiment so I’d like to think it is accurate.

How Will I Track Popularity Now?

As you can see none of the services I’ve been using are reliable.  So I have to revisit how I will track popularity for the remainder of the experiment.

A new search for blog statistics sites returned iBlogstats.  It seems to provide fairly comprehensive information for blog owners, but are the results reliable?

• Delicious has not reflected the bookmark I know exists for the main website or those for individual posts within it.
• I know an individual post has been Dugg, not the main site, so the figure seems to reflect the whole site.  This is the way I’d prefer all statistics to be reported.
• Mister Wong metric doesn’t seem right.  I would have expected something within the 10 to 20 range..
• Yahoo and Live indexed page values aren’t accurate.  I also believe the Live backlink count is inaccurate.

So once again I find myself with no reliable social metrics to work with.

I had originally hoped to look for relationships between social site feedback, site metrics and site analytics.  Now that it would be extremely difficult to measure social feedback I plan to compensate for this by extracting site analytics for referrals from social sites. This is a change in tack from measuring the social site links to my blog, to measuring the effectiveness of those links (i.e. how many people wanted to visit after seeing them).  This seems entirely consistent with the experiment so I’m happy to proceed on that basis.

Next Steps

Next steps are to continue as I had planned last week.  i.e. continue active social bookmarking, monitor any progress and devise an approach for submitting posts to social news sites.  Hopefully next week I’ll have some interesting (and reliable) results to share with you!

The internet is an incredibly convenient way to communicate.  Unfortunately this convenience makes it easy to forget these communiqués are not inherently secure.  A malicious user has many opportunities to tap into or modify this information stream as it goes whizzing past.

Thankfully most modern email clients, and some IM clients, have capabilities built in to allow us to guarantee communications remain private and free from interference.  S/MIME (and PKI) are the foundation for this and provide a way to ensure what is received is exactly what was sent (via a digital signature), and keep the conversation private (via encryption).  Before we can use this functionality we first need to obtain a digital certificate.  To quote wikipedia:

Before S/MIME can be used in any of the above applications, one must obtain and install an individual key/certificate either from one’s in-house certificate authority (CA) or from a public CA. … Encryption requires having the destination party’s certificate on store (which is typically automatic upon receiving a message from the party with a valid signing certificate).

A typical basic personal certificate verifies the owner’s identity only in terms of binding them to an email address and does not verify the person’s name or business. The latter, if needed (e.g. for signing contracts), can be obtained through CAs that offer further verification (digital notary) services or managed PKI service.

Digital certificates can also have other uses such as verifying your (client) identity to a web site, or digitally signing code.

There are a number of sources of free digital certificates available but they can be difficult to locate and near impossible to compare features without ordering and trying each certificate.  To resolve this impasse I’ve performed a comparative review of 9 different certificate authorities to see which is offering the best free digital certificates for typical needs.

The services were picked from those mentioned on the certifi.ca OpenID provider‘s website, Mozilla’s knowledge base, DMOZ and the first 50 hits from a search on Google. For each of the CAs I used Firefox 2 to request, download and install the available certificate.  The certificate was examined within Firefox before using the certifi.ca site to determine whether it could also be used for client authentication.

After reviewing the options I’d recommend:

• Signing up for a CACert certificate because of the range of features available, including a verified Class 3 certificate.  If (when?) it is included in Firefox 3 adoption of this CA should pickup markedly.  Your support can only help with the CA’s mass market adoption.
• Also signing up for a thawte certificate.  Verifying this identity will help you obtain a verified CACert certificate, and this can also be a backup for any sites that are yet to be configured to accept CACert’s root certificate.Update: Unfortunately Thawte have ceased to offer their Freemail certificates as at November 2009.

None of the other options provided any benefits beyond those available through CACert and thawte, so there is no need to resort to using the lesser or unknown CA’s also providing free certificates.  I also don’t see any need to pay for a verified digital identity certificate (e.g. from Verisign) when quality options such as thawte and CACert are available at no cost.  Detailed reviews are provided below, in order of preference, so you can verify these recommendations for yourself. 

So what are you waiting for?  Install a free digital certificate today, and start enjoying a more secure internet!