Reasons for sticking with OmniGraffle 5


#1

If, by “updates”, you mean updates to OG 5, no they have stopped producing them. They produce only new bug versions, which you have to debug yourself, which are labelled product upgrades (down-grades, really), and cost mucho mullah.

I wouldn’t mind paying for a bug-free OG 5, but I am told there is no chance of that.

Cheers


Downgrading from 7 to 6 - any known issues
Text in Object Auto Resizing
Orthagonal Line Issues
#2

It sounds like there’s some confusion here about our testing process! We always start out with internal testing (including both automated tests, and manual hands-on testing) and our internal test team has gotten much stronger through the years.

But we also do public testing of our apps, of course, just like we’ve done with every version of OmniGraffle starting with version 1 in 2001. There’s nothing like getting feedback from actual customers as they put the app through its paces to find out what we’ve overlooked, or misunderstood about their workflow!

Speaking of which, based on your frustration it certainly sounds like we’ve overlooked or misunderstood something about the way you use the app. Would you care to elaborate on anything specific, so we can try to address those concerns?

We have a lot of customers who use OmniGraffle in very different ways, and one area where I know we struggle is that we can overweigh the feedback we’re hearing: in trying to address that feedback and make those customers happy, we sometimes change things that a silent majority were already happy with and end up making them unhappy. Ideally, we would hear all that feedback during our public test cycle, but customers who were happy with the previous version often don’t participate in that testing so we end up hearing about it after the next version ships and end up having to correct it during our .x releases.

For example, in OmniGraffle 6.0 we moved inspectors into the window sidebar based on feedback from laptop users who struggled with having many separate inspector windows to manage. 6.0 made many of those customers much happier, but when 6.0 shipped we quickly learned how many people preferred things the way they were: they loved being able to place those inspectors wherever they want, such as on their second monitor. We addressed that feedback in 6.1 by giving people both options.

We made a similar mistake in version 7 with the placement of the stencils window: many people found stencils to be too buried where they were living in the Inspectors in v6, so we promoted them to their own top-level window with its own toolbar item (much like they were in v5 and earlier versions of the app) and added a sidebar to make those stencils easier to navigate. But as we’ve heard here on the forums, some people really preferred the v6 layout because they love having stencils in the same window as their content. We hear you, and will be bringing back that placement option soon.

What are some of the specific frustrations you have with OmniGraffle 7 that we haven’t yet addressed?


#3

04 May 17 - Several Typos Corrected. Further Detail Added.

Ken

First thanks for being interested in what heavy duty long-term (decades) customers do with OG. From my various email to the “support team”, which is where the frustration begins, the new OG development team seemed to have no interest. They understand “like” and “dislike” but they do not understand objective truth (the single and one Reality, with a capital R).

Second, sorry for the delay.

Third, I cannot, and will not answer re OG 7. OG 6 was too awful to get past. And I do not have the time to do your testing for you (evidently my simple testing far surpasses your “both automated tests, and manual hands-on testing” that has “gotten much stronger through the years”.) I will answer re the product in general, all the releases, and certainly with specifics.

There is a lot in your post, I might not get through all of it today.

Background. I am a s/w Architect, I use OG for all my architectural diagrams. S/w architecture in many different areas of s/w. Eg I have my own set of stencils, and I am known mostly for my database work. Which mean, OG is the app that I use most, above all other apps. My working docs and models are quite large in PageSize. Many finished docs are in A4, and a large no of pages.

Not in my experience. They have gotten weaker through the years. OG 5 is buggier than OG 4; OG 6 is buggier than OG 5; and OG 7 is not worth the time to find out. If and when there is a OG 7 that is for purchase, rather than public testing, I will try it. When I tried OG 6, which was years after the general (not testing) public release, it had bugs that caused it to crash and burn on small complex files or large simple files.

I don’t have time for user testing of s/w that I purchase. I purchase s/w that does something that I want. So my statements above exclude the “public testing” versions.

Yes. Definitely. But there are too many, so it will probably be one major item per post, over a few days. Let’s start with this item.

I find that a very short-sighted, immature, and reactive approach for a s/w company to take. It is the FarceBook, anti-social media presented and called “social”, approach to life.

I am a database geek, using Sybase exclusively since 1993. If, in any progressive release, Sybase removed functionality that they had in the previous release, they would be out of business.

If they added things that new users “liked” that would be fine, but short-sighted and stupid, because new users are generally clueless about the in-depth use of the product, what they want may be supplied some other way, or may be inappropriate. And mature users are the ones whom they should be concerned with, because they know the product in-depth, and they know what is required to make it even better. New users come and go. Old users hang around. New features that new users want are likewise irrelevant, they will not be used by the silent majority.

Corporations live or die based on the majority that they serve, be it silent or not. Corporations that react to the vociferous minority do not last long. Listening to the minority, and failing to canvas the majority is not reasonable if survival is the goal.

Listening to the minority on FarceBook, counting their “likes” and “dislikes”, is short-term, reactionary, and most important, you will be forever reversing and re-reversing features.

Sometimes ??? Excuse me, but you do that all the time, with every single release. Personally, I have seen many useful features removed (corporate suicide), and useless features added (Farcical “likes” scored).

Unhappy is the wrong word. Think about how you would feel if the train station you use to commute to work every day was removed. Without notice. And you had to pay money to catch the new train that does not stop at your station.

Correct.

But that is not how the real world works (it may be how the virtual world, with all its unreality works, but even that is fleeting, ever-changing). It is unreasonable to expect the silent majority to (a) participate in testing s/w that does not work, let alone (b) tell you about what they think of the changes. Most companies tell the user base about the impending changes (eg. we have email) and that gives the user base a chance to respond. © the changes you mention are actually removal of functionality and features. I have difficulty accepting that a s/w company with more than three employees actually removes functions and features from their product (and I am not talking about placement of a stencil in this or that or the other window … that is a relocation, not a removal of a function, sure it affects efficiency as I have stated, but that is nothing compared to removal).

OG gets zero for “listening”. Listening is not important. Listening to things from valid people is what is important. If I listened to every man in the street, I would go mad, because now that we have opened the doors of the asylums, mad people are out there infecting normal people.

What was frustrating about the emails to support was that their attitude was one of “gee whiz, we didn’t know customers used OG that way”. Which is idiotic, because OG is programmed that way. The only way a customer can use it that way is because OG programmed it that way. The feature was programmed. The product was sold with the feature. The customers (especially the long term ones) purchased the product for those features, and use those feature. it is stupefying to me, to have to deal with a support team that not only accepts the removal of a feature, as if it were not the horrendous problem that it is, but will not do anything about it on the basis that “we didn’t know that customers used the program that way”, the way that was programmed by the same programmers.

Now there is a second level of frustration because they (and for the most part, the OG employees on this new discourse forum, since the wealth of information in the old forum has been suppressed) have a dishonest way of making a removal sound like an added feature. One cannot respond to such dishonesty without being rude, and then rudeness gets the post suppressed. So basically it is censoring, such that negative feedback is eliminated, and the kindergarten kids get to hear only positive feedback.

Let’s say we have features ABCDEF in one release. If the next release has ABCEFGHI, it does not matter that the new kids on the block “like” GHI. It matters that you have removed D, that customers use D, because you programmed D, you sold D, and therefore now you cannot, you must not, ever remove D (without replacing it with a better more functional D).

A • Scale per Canvas Lost

Let’s take a simple, easily proved/disproved item to start with, and depending on how you respond to it, we can handle the bigger, more important items that make OG 4 more that 5, 5 run circles around 6, and 6 better than 7, etc. I hope you don’t mind, I have already invested a fair amount of time explaining things in detail to the support bots, and I am not about to repeat that, with each and every feature that you have removed (servicing the thread title), unless there is going to be some fruit.

In OG 4, we have the Scale in the Inspector/Canvas, which allows the scale to be set per canvas.

I use Templates, and I have an entire set of them, for each PageSize, and for each Scale. Eg. PageSize A4 at Scale 80%; 90%; 100%.

For the final doc, which is usually PDF, and one PageSize, eg. A4, eg. 50 pages, I might use eg. 90% scale on most pages (canvasses), say 40 pages, and 80% scale on the remaining 10 pages (canvasses). I have hundreds of docs in OG 4. PDF handles that perfectly. It shows up in Windead (Adobe) systems and Apples (Preview) perfectly. It prints (I use print shops a lot, in many different cities) perfectly. The headers; footers; and logos etc, show up perfectly, as if they were all one size, because the templates are mature and correct (eg. Times 9 at 100% Scale and Times 10 at 90% Scale).

In OG 5 the darling people who work for OG, being evidently ignorant about what OG is actually used for, what OG can actually do, and being unconcerned with the consequences of removal of features, removed Scale from the Canvas Inspector.

No amount of serious complaint could get the support bots to think past “gee, we didn’t know the program that we programmed was used the way we programmed it”.

As long as my hundreds of docs do not change, which is the mindset of newbies, who have only simple docs that can be replaced easily, that is fine. But my docs are mature, and they must be changed with every release of Sybase, or the next release of a Data Model, or the next big contract, etc. And no, I am not about to create a new OG doc from scratch. So now, whenever I change an existing doc, if it has different scales per canvas, I have to choose one scale (for the doc) and reproduce every page that is not that single scale. Therefore OG 5 is a step backwards in time, a functional regression.

OG 6 has not been corrected, the feature removal has not been reversed. As you are probably aware, this issue does come up in the forum.

Response please.

Cheers


#4

B • Diagram Styles Lost

I have detailed this particular lost feature in another thread. I hope I have conveyed therein the point that is relevant in this thread, which is Reasons for sticking to OG 5. Which is what I use, despite having fully-paid licences for OG 6 and 7.


#5

This post was flagged by the community and is temporarily hidden.


#6

This post was flagged by the community and is temporarily hidden.


#7

This post was flagged by the community and is temporarily hidden.


#8

###[Moved here by the moderator, from another thread]

OG 5, 6, 7 can live (be installed) on the same machine, no problem.

OG 5 and 6 should not be open (launched, running) at the same time. They use the same workspaces, so if you have the same doc open (accidentally), it gets confused, and it will confuse you even more. So, they can be open at the same time, but I advise against it for sanity reasons.

I have gotten into the habit of closing all docs in version X before opening version
Y. Because the last thing I want when opening version Y is for it to automatically (it does) open the docs I had open with version X.

FYI, I am sticking to OG 5, the GUI is better; the context handling is better, and it has less bugs. There is another thread Reasons for Sticking with OG5 , that Ken has opened, for that purpose. I have not provided a complete list yet.

To be perfectly honest, I would stick to OG 4, except for one nasty problem: it does not render on Retinas. OG 5 has a couple of major problems, serious dramas, that OG 4 does not have. But the OG 5 GUI is much better than 6 and 7, and I have gotten used to OG 5 and its quirks.

Yessiree bob. I could not have said it better myself.

Years ago, I was responsible for getting OG into every customer site that I consulted at. Now I cannot recommend it, because it is so broken.

It appears that the different GUI components are written by different people, who do not speak to each other.

There is no overall GUI design, no cohesion.

When you switch from one tool or palette to another, the interaction required is quite different. ESC means one here, another thing there, and a third thing somewhere else. That is just one example.

They do not seem to understand that all these changes between versions actually affect productivity. Each version has substantial changes to the GUI, which either removes features or (as in your example) splits functionality across two tools.

They have also lost the sense, the meaning, of context. The various effects are labelled “contextual”, but that “contextual” is in name only, it has no context. V6 has somewhat more understanding of context, and V5 has somewhat more again. The point is, as you say, in every progressive version, the product is losing functionality, specifically functionality that affects productivity. Loss of context means having to enter more keystrokes in the later version, to perform the same action that one did in the previous version.

I too, welcome improvements. But not removal of features. And not loss of functionality.

They’ve killed the goose that laid the golden egg, for sure. They are going after the whiz bang market, which is only interested in the look, the image. Not serious work.

Yes. It is telling, when the fix is a return to the operation of the previous version. Perhaps they can just save the expense of breaking it in the first place.

That point manipulation thing (I will refrain from calling it a tool) is a train wreck, it has to be fixed. But more important, they must not subtract anything the Selection tool currently does.

Cheers


#9

My attention was directed to a post in this thread, which I flagged as a violation of our community guidelines. These forums are for discussing software, not attacking people, and everyone needs to play by the same rules. If the author edits the post to bring it into alignment with the guidelines, it will automatically become visible again.

On a similar note, posts that rely on inflammatory terms like ‘idiotic’ and ‘daft’ are flag-worthy as well as ones that speculate about people. Keep it to the software.


#10

I totally can understand GraffleGuru’s frustration. I’m too have the impression, that OG is on a fast paced feature train and cares little about trusted stuff going overboard. My mayor annoyance with v7 and v6: hyphenation is broken. It does work while you edit text but the hyphens disappear when you deselect the shape. My use case: I design a lot of pages for (scientific) surveys. The question items often contain quite some text and horizontal space is limited. I’m annoyed every time I have to insert hypens by hand (like it’s 1985) and have to remove them when the text changes (my main language is german which is infamous for long words but as OG supports this language). I mailed that twice to support (the first time early after I upgraded to v6, so not “just yesterday”) just to get a polite “thanksforwritingingwewilllookintoitsomeday”-reply. Hyphenation worked perfectly fine in v5 (as it did in v4). To me handling text should be thought a core competence of a visualization software that calls itself “Professional”. Hyphenation may be boring, but it is a bread-and-butter feature and I can not understand why such basic bugs don’t get any attention while shiny stuff like “infinite canvas” (which I’ll probably never need - I know, ymmv) is thrown in.

I could add more examples. As for the GUI, there are a lot of quirks: The quickbar disappears sometimes without apparent cause; in v7 i have “ghost selections”: I select shape A, then shape B but I still see the compound box around A; the over way round sometimes I select an item in the sidebar and no selection box appears on the canvas (“invisible selection”); switching between inspectors is sloooow; I define a page height of 15 mm but a 15 mm tall rectangle overlapps the displayed page boundaries; the first shipped version of v7 had an erroneous (and unusable) export sheet … Versions 4 and 5 were bullet proof in that regard. You could trust it on the day-to-day-stuff. And you could trust it from day 1. To me, with v6 and v7 OG has taken a questionable turn and acquired the stench of banana software that ripens at the customer. As a pro user (at least in terms of paying pro-upgrade prices - and I’m one of that loyal users) I’m not pleased with that trend. Thanks to GraffleGuru for the extensive and reasonable exhibits.

I’m not pleading for ceasing innovation on OmniGraffle but I wish it would remember it’s heritage and would again interpret “Professional” more terms of software quality than in terms of marketing babble.


#11

Thanks to both of you for sharing your feedback: we are listening, and I feel the best way to respond to your concerns are to address the issues you’ve reported.

To that end:

I’m very sorry this bug slipped through the cracks for so long! It’s now fixed in the latest test builds of OmniGraffle 7.5.


#12

@kcase:
Thank you for the fix. This is really a relief and very much appreciated.

To “reward” this, a little praise: the new automation capabilities are really promising :-)


#13

Honestly, I wish I could downgrade back to 5 pro. I had to purchase a new mac because my previous one was stolen and I just sort of upgraded apps as a matter of course.

the UI in 7 is just so intolerably bad.